Skip to content

Reference for ultralytics/utils/plotting.py

Note

This file is available at https://github.com/ultralytics/ultralytics/blob/main/ultralytics/utils/plotting.py. If you spot a problem please help fix it by contributing a Pull Request 🛠️. Thank you 🙏!


ultralytics.utils.plotting.Colors

Colors()

Ultralytics color palette https://docs.ultralytics.com/reference/utils/plotting/#ultralytics.utils.plotting.Colors.

This class provides methods to work with the Ultralytics color palette, including converting hex color codes to RGB values.

Attributes:

Name Type Description
palette list of tuple

List of RGB color values.

n int

The number of colors in the palette.

pose_palette ndarray

A specific color palette array with dtype np.uint8.

Ultralytics Color Palette

Index Color HEX RGB
0 #042aff (4, 42, 255)
1 #0bdbeb (11, 219, 235)
2 #f3f3f3 (243, 243, 243)
3 #00dfb7 (0, 223, 183)
4 #111f68 (17, 31, 104)
5 #ff6fdd (255, 111, 221)
6 #ff444f (255, 68, 79)
7 #cced00 (204, 237, 0)
8 #00f344 (0, 243, 68)
9 #bd00ff (189, 0, 255)
10 #00b4ff (0, 180, 255)
11 #dd00ba (221, 0, 186)
12 #00ffff (0, 255, 255)
13 #26c000 (38, 192, 0)
14 #01ffb3 (1, 255, 179)
15 #7d24ff (125, 36, 255)
16 #7b0068 (123, 0, 104)
17 #ff1b6c (255, 27, 108)
18 #fc6d2f (252, 109, 47)
19 #a2ff0b (162, 255, 11)

Pose Color Palette

Index Color HEX RGB
0 #ff8000 (255, 128, 0)
1 #ff9933 (255, 153, 51)
2 #ffb266 (255, 178, 102)
3 #e6e600 (230, 230, 0)
4 #ff99ff (255, 153, 255)
5 #99ccff (153, 204, 255)
6 #ff66ff (255, 102, 255)
7 #ff33ff (255, 51, 255)
8 #66b2ff (102, 178, 255)
9 #3399ff (51, 153, 255)
10 #ff9999 (255, 153, 153)
11 #ff6666 (255, 102, 102)
12 #ff3333 (255, 51, 51)
13 #99ff99 (153, 255, 153)
14 #66ff66 (102, 255, 102)
15 #33ff33 (51, 255, 51)
16 #00ff00 (0, 255, 0)
17 #0000ff (0, 0, 255)
18 #ff0000 (255, 0, 0)
19 #ffffff (255, 255, 255)

Ultralytics Brand Colors

For Ultralytics brand colors see https://www.ultralytics.com/brand. Please use the official Ultralytics colors for all marketing materials.

Source code in ultralytics/utils/plotting.py
def __init__(self):
    """Initialize colors as hex = matplotlib.colors.TABLEAU_COLORS.values()."""
    hexs = (
        "042AFF",
        "0BDBEB",
        "F3F3F3",
        "00DFB7",
        "111F68",
        "FF6FDD",
        "FF444F",
        "CCED00",
        "00F344",
        "BD00FF",
        "00B4FF",
        "DD00BA",
        "00FFFF",
        "26C000",
        "01FFB3",
        "7D24FF",
        "7B0068",
        "FF1B6C",
        "FC6D2F",
        "A2FF0B",
    )
    self.palette = [self.hex2rgb(f"#{c}") for c in hexs]
    self.n = len(self.palette)
    self.pose_palette = np.array(
        [
            [255, 128, 0],
            [255, 153, 51],
            [255, 178, 102],
            [230, 230, 0],
            [255, 153, 255],
            [153, 204, 255],
            [255, 102, 255],
            [255, 51, 255],
            [102, 178, 255],
            [51, 153, 255],
            [255, 153, 153],
            [255, 102, 102],
            [255, 51, 51],
            [153, 255, 153],
            [102, 255, 102],
            [51, 255, 51],
            [0, 255, 0],
            [0, 0, 255],
            [255, 0, 0],
            [255, 255, 255],
        ],
        dtype=np.uint8,
    )

__call__

__call__(i, bgr=False)

Converts hex color codes to RGB values.

Source code in ultralytics/utils/plotting.py
def __call__(self, i, bgr=False):
    """Converts hex color codes to RGB values."""
    c = self.palette[int(i) % self.n]
    return (c[2], c[1], c[0]) if bgr else c

hex2rgb staticmethod

hex2rgb(h)

Converts hex color codes to RGB values (i.e. default PIL order).

Source code in ultralytics/utils/plotting.py
@staticmethod
def hex2rgb(h):
    """Converts hex color codes to RGB values (i.e. default PIL order)."""
    return tuple(int(h[1 + i : 1 + i + 2], 16) for i in (0, 2, 4))





ultralytics.utils.plotting.Annotator

Annotator(
    im,
    line_width=None,
    font_size=None,
    font="Arial.ttf",
    pil=False,
    example="abc",
)

Ultralytics Annotator for train/val mosaics and JPGs and predictions annotations.

Attributes:

Name Type Description
im Image.Image or numpy array

The image to annotate.

pil bool

Whether to use PIL or cv2 for drawing annotations.

font truetype or load_default

Font used for text annotations.

lw float

Line width for drawing.

skeleton List[List[int]]

Skeleton structure for keypoints.

limb_color List[int]

Color palette for limbs.

kpt_color List[int]

Color palette for keypoints.

Source code in ultralytics/utils/plotting.py
def __init__(self, im, line_width=None, font_size=None, font="Arial.ttf", pil=False, example="abc"):
    """Initialize the Annotator class with image and line width along with color palette for keypoints and limbs."""
    non_ascii = not is_ascii(example)  # non-latin labels, i.e. asian, arabic, cyrillic
    input_is_pil = isinstance(im, Image.Image)
    self.pil = pil or non_ascii or input_is_pil
    self.lw = line_width or max(round(sum(im.size if input_is_pil else im.shape) / 2 * 0.003), 2)
    if self.pil:  # use PIL
        self.im = im if input_is_pil else Image.fromarray(im)
        self.draw = ImageDraw.Draw(self.im)
        try:
            font = check_font("Arial.Unicode.ttf" if non_ascii else font)
            size = font_size or max(round(sum(self.im.size) / 2 * 0.035), 12)
            self.font = ImageFont.truetype(str(font), size)
        except Exception:
            self.font = ImageFont.load_default()
        # Deprecation fix for w, h = getsize(string) -> _, _, w, h = getbox(string)
        if check_version(pil_version, "9.2.0"):
            self.font.getsize = lambda x: self.font.getbbox(x)[2:4]  # text width, height
    else:  # use cv2
        assert im.data.contiguous, "Image not contiguous. Apply np.ascontiguousarray(im) to Annotator input images."
        self.im = im if im.flags.writeable else im.copy()
        self.tf = max(self.lw - 1, 1)  # font thickness
        self.sf = self.lw / 3  # font scale
    # Pose
    self.skeleton = [
        [16, 14],
        [14, 12],
        [17, 15],
        [15, 13],
        [12, 13],
        [6, 12],
        [7, 13],
        [6, 7],
        [6, 8],
        [7, 9],
        [8, 10],
        [9, 11],
        [2, 3],
        [1, 2],
        [1, 3],
        [2, 4],
        [3, 5],
        [4, 6],
        [5, 7],
    ]

    self.limb_color = colors.pose_palette[[9, 9, 9, 9, 7, 7, 7, 0, 0, 0, 0, 0, 16, 16, 16, 16, 16, 16, 16]]
    self.kpt_color = colors.pose_palette[[16, 16, 16, 16, 16, 0, 0, 0, 0, 0, 0, 9, 9, 9, 9, 9, 9]]
    self.dark_colors = {
        (235, 219, 11),
        (243, 243, 243),
        (183, 223, 0),
        (221, 111, 255),
        (0, 237, 204),
        (68, 243, 0),
        (255, 255, 0),
        (179, 255, 1),
        (11, 255, 162),
    }
    self.light_colors = {
        (255, 42, 4),
        (79, 68, 255),
        (255, 0, 189),
        (255, 180, 0),
        (186, 0, 221),
        (0, 192, 38),
        (255, 36, 125),
        (104, 0, 123),
        (108, 27, 255),
        (47, 109, 252),
        (104, 31, 17),
    }

box_label

box_label(
    box,
    label="",
    color=(128, 128, 128),
    txt_color=(255, 255, 255),
    rotated=False,
)

Draws a bounding box to image with label.

Parameters:

Name Type Description Default
box tuple

The bounding box coordinates (x1, y1, x2, y2).

required
label str

The text label to be displayed.

''
color tuple

The background color of the rectangle (B, G, R).

(128, 128, 128)
txt_color tuple

The color of the text (R, G, B).

(255, 255, 255)
rotated bool

Variable used to check if task is OBB

False
Source code in ultralytics/utils/plotting.py
def box_label(self, box, label="", color=(128, 128, 128), txt_color=(255, 255, 255), rotated=False):
    """
    Draws a bounding box to image with label.

    Args:
        box (tuple): The bounding box coordinates (x1, y1, x2, y2).
        label (str): The text label to be displayed.
        color (tuple, optional): The background color of the rectangle (B, G, R).
        txt_color (tuple, optional): The color of the text (R, G, B).
        rotated (bool, optional): Variable used to check if task is OBB
    """
    txt_color = self.get_txt_color(color, txt_color)
    if isinstance(box, torch.Tensor):
        box = box.tolist()
    if self.pil or not is_ascii(label):
        if rotated:
            p1 = box[0]
            self.draw.polygon([tuple(b) for b in box], width=self.lw, outline=color)  # PIL requires tuple box
        else:
            p1 = (box[0], box[1])
            self.draw.rectangle(box, width=self.lw, outline=color)  # box
        if label:
            w, h = self.font.getsize(label)  # text width, height
            outside = p1[1] >= h  # label fits outside box
            if p1[0] > self.im.size[0] - w:  # size is (w, h), check if label extend beyond right side of image
                p1 = self.im.size[0] - w, p1[1]
            self.draw.rectangle(
                (p1[0], p1[1] - h if outside else p1[1], p1[0] + w + 1, p1[1] + 1 if outside else p1[1] + h + 1),
                fill=color,
            )
            # self.draw.text((box[0], box[1]), label, fill=txt_color, font=self.font, anchor='ls')  # for PIL>8.0
            self.draw.text((p1[0], p1[1] - h if outside else p1[1]), label, fill=txt_color, font=self.font)
    else:  # cv2
        if rotated:
            p1 = [int(b) for b in box[0]]
            cv2.polylines(self.im, [np.asarray(box, dtype=int)], True, color, self.lw)  # cv2 requires nparray box
        else:
            p1, p2 = (int(box[0]), int(box[1])), (int(box[2]), int(box[3]))
            cv2.rectangle(self.im, p1, p2, color, thickness=self.lw, lineType=cv2.LINE_AA)
        if label:
            w, h = cv2.getTextSize(label, 0, fontScale=self.sf, thickness=self.tf)[0]  # text width, height
            h += 3  # add pixels to pad text
            outside = p1[1] >= h  # label fits outside box
            if p1[0] > self.im.shape[1] - w:  # shape is (h, w), check if label extend beyond right side of image
                p1 = self.im.shape[1] - w, p1[1]
            p2 = p1[0] + w, p1[1] - h if outside else p1[1] + h
            cv2.rectangle(self.im, p1, p2, color, -1, cv2.LINE_AA)  # filled
            cv2.putText(
                self.im,
                label,
                (p1[0], p1[1] - 2 if outside else p1[1] + h - 1),
                0,
                self.sf,
                txt_color,
                thickness=self.tf,
                lineType=cv2.LINE_AA,
            )

circle_label

circle_label(
    box, label="", color=(128, 128, 128), txt_color=(255, 255, 255), margin=2
)

Draws a label with a background circle centered within a given bounding box.

Parameters:

Name Type Description Default
box tuple

The bounding box coordinates (x1, y1, x2, y2).

required
label str

The text label to be displayed.

''
color tuple

The background color of the rectangle (B, G, R).

(128, 128, 128)
txt_color tuple

The color of the text (R, G, B).

(255, 255, 255)
margin int

The margin between the text and the rectangle border.

2
Source code in ultralytics/utils/plotting.py
def circle_label(self, box, label="", color=(128, 128, 128), txt_color=(255, 255, 255), margin=2):
    """
    Draws a label with a background circle centered within a given bounding box.

    Args:
        box (tuple): The bounding box coordinates (x1, y1, x2, y2).
        label (str): The text label to be displayed.
        color (tuple, optional): The background color of the rectangle (B, G, R).
        txt_color (tuple, optional): The color of the text (R, G, B).
        margin (int, optional): The margin between the text and the rectangle border.
    """
    # If label have more than 3 characters, skip other characters, due to circle size
    if len(label) > 3:
        print(
            f"Length of label is {len(label)}, initial 3 label characters will be considered for circle annotation!"
        )
        label = label[:3]

    # Calculate the center of the box
    x_center, y_center = int((box[0] + box[2]) / 2), int((box[1] + box[3]) / 2)
    # Get the text size
    text_size = cv2.getTextSize(str(label), cv2.FONT_HERSHEY_SIMPLEX, self.sf - 0.15, self.tf)[0]
    # Calculate the required radius to fit the text with the margin
    required_radius = int(((text_size[0] ** 2 + text_size[1] ** 2) ** 0.5) / 2) + margin
    # Draw the circle with the required radius
    cv2.circle(self.im, (x_center, y_center), required_radius, color, -1)
    # Calculate the position for the text
    text_x = x_center - text_size[0] // 2
    text_y = y_center + text_size[1] // 2
    # Draw the text
    cv2.putText(
        self.im,
        str(label),
        (text_x, text_y),
        cv2.FONT_HERSHEY_SIMPLEX,
        self.sf - 0.15,
        self.get_txt_color(color, txt_color),
        self.tf,
        lineType=cv2.LINE_AA,
    )

display_analytics

display_analytics(im0, text, txt_color, bg_color, margin)

Display the overall statistics for parking lots.

Parameters:

Name Type Description Default
im0 ndarray

Inference image.

required
text dict

Labels dictionary.

required
txt_color tuple

Display color for text foreground.

required
bg_color tuple

Display color for text background.

required
margin int

Gap between text and rectangle for better display.

required
Source code in ultralytics/utils/plotting.py
def display_analytics(self, im0, text, txt_color, bg_color, margin):
    """
    Display the overall statistics for parking lots.

    Args:
        im0 (ndarray): Inference image.
        text (dict): Labels dictionary.
        txt_color (tuple): Display color for text foreground.
        bg_color (tuple): Display color for text background.
        margin (int): Gap between text and rectangle for better display.
    """
    horizontal_gap = int(im0.shape[1] * 0.02)
    vertical_gap = int(im0.shape[0] * 0.01)
    text_y_offset = 0
    for label, value in text.items():
        txt = f"{label}: {value}"
        text_size = cv2.getTextSize(txt, 0, self.sf, self.tf)[0]
        if text_size[0] < 5 or text_size[1] < 5:
            text_size = (5, 5)
        text_x = im0.shape[1] - text_size[0] - margin * 2 - horizontal_gap
        text_y = text_y_offset + text_size[1] + margin * 2 + vertical_gap
        rect_x1 = text_x - margin * 2
        rect_y1 = text_y - text_size[1] - margin * 2
        rect_x2 = text_x + text_size[0] + margin * 2
        rect_y2 = text_y + margin * 2
        cv2.rectangle(im0, (rect_x1, rect_y1), (rect_x2, rect_y2), bg_color, -1)
        cv2.putText(im0, txt, (text_x, text_y), 0, self.sf, txt_color, self.tf, lineType=cv2.LINE_AA)
        text_y_offset = rect_y2

display_objects_labels

display_objects_labels(
    im0, text, txt_color, bg_color, x_center, y_center, margin
)

Display the bounding boxes labels in parking management app.

Parameters:

Name Type Description Default
im0 ndarray

Inference image.

required
text str

Object/class name.

required
txt_color tuple

Display color for text foreground.

required
bg_color tuple

Display color for text background.

required
x_center float

The x position center point for bounding box.

required
y_center float

The y position center point for bounding box.

required
margin int

The gap between text and rectangle for better display.

required
Source code in ultralytics/utils/plotting.py
def display_objects_labels(self, im0, text, txt_color, bg_color, x_center, y_center, margin):
    """
    Display the bounding boxes labels in parking management app.

    Args:
        im0 (ndarray): Inference image.
        text (str): Object/class name.
        txt_color (tuple): Display color for text foreground.
        bg_color (tuple): Display color for text background.
        x_center (float): The x position center point for bounding box.
        y_center (float): The y position center point for bounding box.
        margin (int): The gap between text and rectangle for better display.
    """
    text_size = cv2.getTextSize(text, 0, fontScale=self.sf, thickness=self.tf)[0]
    text_x = x_center - text_size[0] // 2
    text_y = y_center + text_size[1] // 2

    rect_x1 = text_x - margin
    rect_y1 = text_y - text_size[1] - margin
    rect_x2 = text_x + text_size[0] + margin
    rect_y2 = text_y + margin
    cv2.rectangle(im0, (rect_x1, rect_y1), (rect_x2, rect_y2), bg_color, -1)
    cv2.putText(im0, text, (text_x, text_y), 0, self.sf, txt_color, self.tf, lineType=cv2.LINE_AA)

draw_centroid_and_tracks

draw_centroid_and_tracks(track, color=(255, 0, 255), track_thickness=2)

Draw centroid point and track trails.

Parameters:

Name Type Description Default
track list

object tracking points for trails display

required
color tuple

tracks line color

(255, 0, 255)
track_thickness int

track line thickness value

2
Source code in ultralytics/utils/plotting.py
def draw_centroid_and_tracks(self, track, color=(255, 0, 255), track_thickness=2):
    """
    Draw centroid point and track trails.

    Args:
        track (list): object tracking points for trails display
        color (tuple): tracks line color
        track_thickness (int): track line thickness value
    """
    points = np.hstack(track).astype(np.int32).reshape((-1, 1, 2))
    cv2.polylines(self.im, [points], isClosed=False, color=color, thickness=track_thickness)
    cv2.circle(self.im, (int(track[-1][0]), int(track[-1][1])), track_thickness * 2, color, -1)

draw_region

draw_region(reg_pts=None, color=(0, 255, 0), thickness=5)

Draw region line.

Parameters:

Name Type Description Default
reg_pts list

Region Points (for line 2 points, for region 4 points)

None
color tuple

Region Color value

(0, 255, 0)
thickness int

Region area thickness value

5
Source code in ultralytics/utils/plotting.py
def draw_region(self, reg_pts=None, color=(0, 255, 0), thickness=5):
    """
    Draw region line.

    Args:
        reg_pts (list): Region Points (for line 2 points, for region 4 points)
        color (tuple): Region Color value
        thickness (int): Region area thickness value
    """
    cv2.polylines(self.im, [np.array(reg_pts, dtype=np.int32)], isClosed=True, color=color, thickness=thickness)

    # Draw small circles at the corner points
    for point in reg_pts:
        cv2.circle(self.im, (point[0], point[1]), thickness * 2, color, -1)  # -1 fills the circle

draw_specific_points

draw_specific_points(keypoints, indices=None, radius=2, conf_thres=0.25)

Draw specific keypoints for gym steps counting.

Parameters:

Name Type Description Default
keypoints list

Keypoints data to be plotted.

required
indices list

Keypoint indices to be plotted. Defaults to [2, 5, 7].

None
radius int

Keypoint radius. Defaults to 2.

2
conf_thres float

Confidence threshold for keypoints. Defaults to 0.25.

0.25

Returns:

Type Description
ndarray

Image with drawn keypoints.

Note

Keypoint format: [x, y] or [x, y, confidence]. Modifies self.im in-place.

Source code in ultralytics/utils/plotting.py
def draw_specific_points(self, keypoints, indices=None, radius=2, conf_thres=0.25):
    """
    Draw specific keypoints for gym steps counting.

    Args:
        keypoints (list): Keypoints data to be plotted.
        indices (list, optional): Keypoint indices to be plotted. Defaults to [2, 5, 7].
        radius (int, optional): Keypoint radius. Defaults to 2.
        conf_thres (float, optional): Confidence threshold for keypoints. Defaults to 0.25.

    Returns:
        (numpy.ndarray): Image with drawn keypoints.

    Note:
        Keypoint format: [x, y] or [x, y, confidence].
        Modifies self.im in-place.
    """
    indices = indices or [2, 5, 7]
    points = [(int(k[0]), int(k[1])) for i, k in enumerate(keypoints) if i in indices and k[2] >= conf_thres]

    # Draw lines between consecutive points
    for start, end in zip(points[:-1], points[1:]):
        cv2.line(self.im, start, end, (0, 255, 0), 2, lineType=cv2.LINE_AA)

    # Draw circles for keypoints
    for pt in points:
        cv2.circle(self.im, pt, radius, (0, 0, 255), -1, lineType=cv2.LINE_AA)

    return self.im

estimate_pose_angle staticmethod

estimate_pose_angle(a, b, c)

Calculate the pose angle for object.

Parameters:

Name Type Description Default
a float)

The value of pose point a

required
b float

The value of pose point b

required
c float

The value o pose point c

required

Returns:

Name Type Description
angle degree

Degree value of angle between three points

Source code in ultralytics/utils/plotting.py
@staticmethod
def estimate_pose_angle(a, b, c):
    """
    Calculate the pose angle for object.

    Args:
        a (float) : The value of pose point a
        b (float): The value of pose point b
        c (float): The value o pose point c

    Returns:
        angle (degree): Degree value of angle between three points
    """
    a, b, c = np.array(a), np.array(b), np.array(c)
    radians = np.arctan2(c[1] - b[1], c[0] - b[0]) - np.arctan2(a[1] - b[1], a[0] - b[0])
    angle = np.abs(radians * 180.0 / np.pi)
    if angle > 180.0:
        angle = 360 - angle
    return angle

fromarray

fromarray(im)

Update self.im from a numpy array.

Source code in ultralytics/utils/plotting.py
def fromarray(self, im):
    """Update self.im from a numpy array."""
    self.im = im if isinstance(im, Image.Image) else Image.fromarray(im)
    self.draw = ImageDraw.Draw(self.im)

get_bbox_dimension

get_bbox_dimension(bbox=None)

Calculate the area of a bounding box.

Parameters:

Name Type Description Default
bbox tuple

Bounding box coordinates in the format (x_min, y_min, x_max, y_max).

None

Returns:

Name Type Description
width float

Width of the bounding box.

height float

Height of the bounding box.

area float

Area enclosed by the bounding box.

Source code in ultralytics/utils/plotting.py
def get_bbox_dimension(self, bbox=None):
    """
    Calculate the area of a bounding box.

    Args:
        bbox (tuple): Bounding box coordinates in the format (x_min, y_min, x_max, y_max).

    Returns:
        width (float): Width of the bounding box.
        height (float): Height of the bounding box.
        area (float): Area enclosed by the bounding box.
    """
    x_min, y_min, x_max, y_max = bbox
    width = x_max - x_min
    height = y_max - y_min
    return width, height, width * height

get_txt_color

get_txt_color(color=(128, 128, 128), txt_color=(255, 255, 255))

Assign text color based on background color.

Parameters:

Name Type Description Default
color tuple

The background color of the rectangle for text (B, G, R).

(128, 128, 128)
txt_color tuple

The color of the text (R, G, B).

(255, 255, 255)

Returns:

Name Type Description
txt_color tuple

Text color for label

Source code in ultralytics/utils/plotting.py
def get_txt_color(self, color=(128, 128, 128), txt_color=(255, 255, 255)):
    """
    Assign text color based on background color.

    Args:
        color (tuple, optional): The background color of the rectangle for text (B, G, R).
        txt_color (tuple, optional): The color of the text (R, G, B).

    Returns:
        txt_color (tuple): Text color for label
    """
    if color in self.dark_colors:
        return 104, 31, 17
    elif color in self.light_colors:
        return 255, 255, 255
    else:
        return txt_color

kpts

kpts(
    kpts,
    shape=(640, 640),
    radius=None,
    kpt_line=True,
    conf_thres=0.25,
    kpt_color=None,
)

Plot keypoints on the image.

Parameters:

Name Type Description Default
kpts Tensor

Keypoints, shape [17, 3] (x, y, confidence).

required
shape tuple

Image shape (h, w). Defaults to (640, 640).

(640, 640)
radius int

Keypoint radius. Defaults to 5.

None
kpt_line bool

Draw lines between keypoints. Defaults to True.

True
conf_thres float

Confidence threshold. Defaults to 0.25.

0.25
kpt_color tuple

Keypoint color (B, G, R). Defaults to None.

None
Note
  • kpt_line=True currently only supports human pose plotting.
  • Modifies self.im in-place.
  • If self.pil is True, converts image to numpy array and back to PIL.
Source code in ultralytics/utils/plotting.py
def kpts(self, kpts, shape=(640, 640), radius=None, kpt_line=True, conf_thres=0.25, kpt_color=None):
    """
    Plot keypoints on the image.

    Args:
        kpts (torch.Tensor): Keypoints, shape [17, 3] (x, y, confidence).
        shape (tuple, optional): Image shape (h, w). Defaults to (640, 640).
        radius (int, optional): Keypoint radius. Defaults to 5.
        kpt_line (bool, optional): Draw lines between keypoints. Defaults to True.
        conf_thres (float, optional): Confidence threshold. Defaults to 0.25.
        kpt_color (tuple, optional): Keypoint color (B, G, R). Defaults to None.

    Note:
        - `kpt_line=True` currently only supports human pose plotting.
        - Modifies self.im in-place.
        - If self.pil is True, converts image to numpy array and back to PIL.
    """
    radius = radius if radius is not None else self.lw
    if self.pil:
        # Convert to numpy first
        self.im = np.asarray(self.im).copy()
    nkpt, ndim = kpts.shape
    is_pose = nkpt == 17 and ndim in {2, 3}
    kpt_line &= is_pose  # `kpt_line=True` for now only supports human pose plotting
    for i, k in enumerate(kpts):
        color_k = kpt_color or (self.kpt_color[i].tolist() if is_pose else colors(i))
        x_coord, y_coord = k[0], k[1]
        if x_coord % shape[1] != 0 and y_coord % shape[0] != 0:
            if len(k) == 3:
                conf = k[2]
                if conf < conf_thres:
                    continue
            cv2.circle(self.im, (int(x_coord), int(y_coord)), radius, color_k, -1, lineType=cv2.LINE_AA)

    if kpt_line:
        ndim = kpts.shape[-1]
        for i, sk in enumerate(self.skeleton):
            pos1 = (int(kpts[(sk[0] - 1), 0]), int(kpts[(sk[0] - 1), 1]))
            pos2 = (int(kpts[(sk[1] - 1), 0]), int(kpts[(sk[1] - 1), 1]))
            if ndim == 3:
                conf1 = kpts[(sk[0] - 1), 2]
                conf2 = kpts[(sk[1] - 1), 2]
                if conf1 < conf_thres or conf2 < conf_thres:
                    continue
            if pos1[0] % shape[1] == 0 or pos1[1] % shape[0] == 0 or pos1[0] < 0 or pos1[1] < 0:
                continue
            if pos2[0] % shape[1] == 0 or pos2[1] % shape[0] == 0 or pos2[0] < 0 or pos2[1] < 0:
                continue
            cv2.line(
                self.im,
                pos1,
                pos2,
                kpt_color or self.limb_color[i].tolist(),
                thickness=int(np.ceil(self.lw / 2)),
                lineType=cv2.LINE_AA,
            )
    if self.pil:
        # Convert im back to PIL and update draw
        self.fromarray(self.im)

masks

masks(masks, colors, im_gpu, alpha=0.5, retina_masks=False)

Plot masks on image.

Parameters:

Name Type Description Default
masks tensor

Predicted masks on cuda, shape: [n, h, w]

required
colors List[List[Int]]

Colors for predicted masks, [[r, g, b] * n]

required
im_gpu tensor

Image is in cuda, shape: [3, h, w], range: [0, 1]

required
alpha float

Mask transparency: 0.0 fully transparent, 1.0 opaque

0.5
retina_masks bool

Whether to use high resolution masks or not. Defaults to False.

False
Source code in ultralytics/utils/plotting.py
def masks(self, masks, colors, im_gpu, alpha=0.5, retina_masks=False):
    """
    Plot masks on image.

    Args:
        masks (tensor): Predicted masks on cuda, shape: [n, h, w]
        colors (List[List[Int]]): Colors for predicted masks, [[r, g, b] * n]
        im_gpu (tensor): Image is in cuda, shape: [3, h, w], range: [0, 1]
        alpha (float): Mask transparency: 0.0 fully transparent, 1.0 opaque
        retina_masks (bool): Whether to use high resolution masks or not. Defaults to False.
    """
    if self.pil:
        # Convert to numpy first
        self.im = np.asarray(self.im).copy()
    if len(masks) == 0:
        self.im[:] = im_gpu.permute(1, 2, 0).contiguous().cpu().numpy() * 255
    if im_gpu.device != masks.device:
        im_gpu = im_gpu.to(masks.device)
    colors = torch.tensor(colors, device=masks.device, dtype=torch.float32) / 255.0  # shape(n,3)
    colors = colors[:, None, None]  # shape(n,1,1,3)
    masks = masks.unsqueeze(3)  # shape(n,h,w,1)
    masks_color = masks * (colors * alpha)  # shape(n,h,w,3)

    inv_alpha_masks = (1 - masks * alpha).cumprod(0)  # shape(n,h,w,1)
    mcs = masks_color.max(dim=0).values  # shape(n,h,w,3)

    im_gpu = im_gpu.flip(dims=[0])  # flip channel
    im_gpu = im_gpu.permute(1, 2, 0).contiguous()  # shape(h,w,3)
    im_gpu = im_gpu * inv_alpha_masks[-1] + mcs
    im_mask = im_gpu * 255
    im_mask_np = im_mask.byte().cpu().numpy()
    self.im[:] = im_mask_np if retina_masks else ops.scale_image(im_mask_np, self.im.shape)
    if self.pil:
        # Convert im back to PIL and update draw
        self.fromarray(self.im)

plot_angle_and_count_and_stage

plot_angle_and_count_and_stage(
    angle_text,
    count_text,
    stage_text,
    center_kpt,
    color=(104, 31, 17),
    txt_color=(255, 255, 255),
)

Plot the pose angle, count value, and step stage.

Parameters:

Name Type Description Default
angle_text str

Angle value for workout monitoring

required
count_text str

Counts value for workout monitoring

required
stage_text str

Stage decision for workout monitoring

required
center_kpt list

Centroid pose index for workout monitoring

required
color tuple

Text background color

(104, 31, 17)
txt_color tuple

Text foreground color

(255, 255, 255)
Source code in ultralytics/utils/plotting.py
def plot_angle_and_count_and_stage(
    self, angle_text, count_text, stage_text, center_kpt, color=(104, 31, 17), txt_color=(255, 255, 255)
):
    """
    Plot the pose angle, count value, and step stage.

    Args:
        angle_text (str): Angle value for workout monitoring
        count_text (str): Counts value for workout monitoring
        stage_text (str): Stage decision for workout monitoring
        center_kpt (list): Centroid pose index for workout monitoring
        color (tuple, optional): Text background color
        txt_color (tuple, optional): Text foreground color
    """
    # Format text
    angle_text, count_text, stage_text = f" {angle_text:.2f}", f"Steps : {count_text}", f" {stage_text}"

    # Draw angle, count and stage text
    angle_height = self.plot_workout_information(
        angle_text, (int(center_kpt[0]), int(center_kpt[1])), color, txt_color
    )
    count_height = self.plot_workout_information(
        count_text, (int(center_kpt[0]), int(center_kpt[1]) + angle_height + 20), color, txt_color
    )
    self.plot_workout_information(
        stage_text, (int(center_kpt[0]), int(center_kpt[1]) + angle_height + count_height + 40), color, txt_color
    )

plot_distance_and_line

plot_distance_and_line(
    pixels_distance,
    centroids,
    line_color=(104, 31, 17),
    centroid_color=(255, 0, 255),
)

Plot the distance and line on frame.

Parameters:

Name Type Description Default
pixels_distance float

Pixels distance between two bbox centroids.

required
centroids list

Bounding box centroids data.

required
line_color tuple

Distance line color.

(104, 31, 17)
centroid_color tuple

Bounding box centroid color.

(255, 0, 255)
Source code in ultralytics/utils/plotting.py
def plot_distance_and_line(
    self, pixels_distance, centroids, line_color=(104, 31, 17), centroid_color=(255, 0, 255)
):
    """
    Plot the distance and line on frame.

    Args:
        pixels_distance (float): Pixels distance between two bbox centroids.
        centroids (list): Bounding box centroids data.
        line_color (tuple, optional): Distance line color.
        centroid_color (tuple, optional): Bounding box centroid color.
    """
    # Get the text size
    text = f"Pixels Distance: {pixels_distance:.2f}"
    (text_width_m, text_height_m), _ = cv2.getTextSize(text, 0, self.sf, self.tf)

    # Define corners with 10-pixel margin and draw rectangle
    cv2.rectangle(self.im, (15, 25), (15 + text_width_m + 20, 25 + text_height_m + 20), line_color, -1)

    # Calculate the position for the text with a 10-pixel margin and draw text
    text_position = (25, 25 + text_height_m + 10)
    cv2.putText(
        self.im,
        text,
        text_position,
        0,
        self.sf,
        (255, 255, 255),
        self.tf,
        cv2.LINE_AA,
    )

    cv2.line(self.im, centroids[0], centroids[1], line_color, 3)
    cv2.circle(self.im, centroids[0], 6, centroid_color, -1)
    cv2.circle(self.im, centroids[1], 6, centroid_color, -1)

plot_workout_information

plot_workout_information(
    display_text, position, color=(104, 31, 17), txt_color=(255, 255, 255)
)

Draw text with a background on the image.

Parameters:

Name Type Description Default
display_text str

The text to be displayed.

required
position tuple

Coordinates (x, y) on the image where the text will be placed.

required
color tuple

Text background color

(104, 31, 17)
txt_color tuple

Text foreground color

(255, 255, 255)
Source code in ultralytics/utils/plotting.py
def plot_workout_information(self, display_text, position, color=(104, 31, 17), txt_color=(255, 255, 255)):
    """
    Draw text with a background on the image.

    Args:
        display_text (str): The text to be displayed.
        position (tuple): Coordinates (x, y) on the image where the text will be placed.
        color (tuple, optional): Text background color
        txt_color (tuple, optional): Text foreground color
    """
    (text_width, text_height), _ = cv2.getTextSize(display_text, 0, self.sf, self.tf)

    # Draw background rectangle
    cv2.rectangle(
        self.im,
        (position[0], position[1] - text_height - 5),
        (position[0] + text_width + 10, position[1] - text_height - 5 + text_height + 10 + self.tf),
        color,
        -1,
    )
    # Draw text
    cv2.putText(self.im, display_text, position, 0, self.sf, txt_color, self.tf)

    return text_height

queue_counts_display

queue_counts_display(
    label, points=None, region_color=(255, 255, 255), txt_color=(0, 0, 0)
)

Displays queue counts on an image centered at the points with customizable font size and colors.

Parameters:

Name Type Description Default
label str

Queue counts label.

required
points tuple

Region points for center point calculation to display text.

None
region_color tuple

RGB queue region color.

(255, 255, 255)
txt_color tuple

RGB text display color.

(0, 0, 0)
Source code in ultralytics/utils/plotting.py
def queue_counts_display(self, label, points=None, region_color=(255, 255, 255), txt_color=(0, 0, 0)):
    """
    Displays queue counts on an image centered at the points with customizable font size and colors.

    Args:
        label (str): Queue counts label.
        points (tuple): Region points for center point calculation to display text.
        region_color (tuple): RGB queue region color.
        txt_color (tuple): RGB text display color.
    """
    x_values = [point[0] for point in points]
    y_values = [point[1] for point in points]
    center_x = sum(x_values) // len(points)
    center_y = sum(y_values) // len(points)

    text_size = cv2.getTextSize(label, 0, fontScale=self.sf, thickness=self.tf)[0]
    text_width = text_size[0]
    text_height = text_size[1]

    rect_width = text_width + 20
    rect_height = text_height + 20
    rect_top_left = (center_x - rect_width // 2, center_y - rect_height // 2)
    rect_bottom_right = (center_x + rect_width // 2, center_y + rect_height // 2)
    cv2.rectangle(self.im, rect_top_left, rect_bottom_right, region_color, -1)

    text_x = center_x - text_width // 2
    text_y = center_y + text_height // 2

    # Draw text
    cv2.putText(
        self.im,
        label,
        (text_x, text_y),
        0,
        fontScale=self.sf,
        color=txt_color,
        thickness=self.tf,
        lineType=cv2.LINE_AA,
    )

rectangle

rectangle(xy, fill=None, outline=None, width=1)

Add rectangle to image (PIL-only).

Source code in ultralytics/utils/plotting.py
def rectangle(self, xy, fill=None, outline=None, width=1):
    """Add rectangle to image (PIL-only)."""
    self.draw.rectangle(xy, fill, outline, width)

result

result()

Return annotated image as array.

Source code in ultralytics/utils/plotting.py
def result(self):
    """Return annotated image as array."""
    return np.asarray(self.im)

save

save(filename='image.jpg')

Save the annotated image to 'filename'.

Source code in ultralytics/utils/plotting.py
def save(self, filename="image.jpg"):
    """Save the annotated image to 'filename'."""
    cv2.imwrite(filename, np.asarray(self.im))

seg_bbox

seg_bbox(mask, mask_color=(255, 0, 255), label=None, txt_color=(255, 255, 255))

Function for drawing segmented object in bounding box shape.

Parameters:

Name Type Description Default
mask ndarray

A 2D array of shape (N, 2) containing the contour points of the segmented object.

required
mask_color tuple

RGB color for the contour and label background.

(255, 0, 255)
label str

Text label for the object. If None, no label is drawn.

None
txt_color tuple

RGB color for the label text.

(255, 255, 255)
Source code in ultralytics/utils/plotting.py
def seg_bbox(self, mask, mask_color=(255, 0, 255), label=None, txt_color=(255, 255, 255)):
    """
    Function for drawing segmented object in bounding box shape.

    Args:
        mask (np.ndarray): A 2D array of shape (N, 2) containing the contour points of the segmented object.
        mask_color (tuple): RGB color for the contour and label background.
        label (str, optional): Text label for the object. If None, no label is drawn.
        txt_color (tuple): RGB color for the label text.
    """
    if mask.size == 0:  # no masks to plot
        return

    cv2.polylines(self.im, [np.int32([mask])], isClosed=True, color=mask_color, thickness=2)
    text_size, _ = cv2.getTextSize(label, 0, self.sf, self.tf)

    if label:
        cv2.rectangle(
            self.im,
            (int(mask[0][0]) - text_size[0] // 2 - 10, int(mask[0][1]) - text_size[1] - 10),
            (int(mask[0][0]) + text_size[0] // 2 + 10, int(mask[0][1] + 10)),
            mask_color,
            -1,
        )
        cv2.putText(
            self.im, label, (int(mask[0][0]) - text_size[0] // 2, int(mask[0][1])), 0, self.sf, txt_color, self.tf
        )

show

show(title=None)

Show the annotated image.

Source code in ultralytics/utils/plotting.py
def show(self, title=None):
    """Show the annotated image."""
    im = Image.fromarray(np.asarray(self.im)[..., ::-1])  # Convert numpy array to PIL Image with RGB to BGR
    if IS_COLAB or IS_KAGGLE:  # can not use IS_JUPYTER as will run for all ipython environments
        try:
            display(im)  # noqa - display() function only available in ipython environments
        except ImportError as e:
            LOGGER.warning(f"Unable to display image in Jupyter notebooks: {e}")
    else:
        im.show(title=title)

sweep_annotator

sweep_annotator(
    line_x=0,
    line_y=0,
    label=None,
    color=(221, 0, 186),
    txt_color=(255, 255, 255),
)

Function for drawing a sweep annotation line and an optional label.

Parameters:

Name Type Description Default
line_x int

The x-coordinate of the sweep line.

0
line_y int

The y-coordinate limit of the sweep line.

0
label str

Text label to be drawn in center of sweep line. If None, no label is drawn.

None
color tuple

RGB color for the line and label background.

(221, 0, 186)
txt_color tuple

RGB color for the label text.

(255, 255, 255)
Source code in ultralytics/utils/plotting.py
def sweep_annotator(self, line_x=0, line_y=0, label=None, color=(221, 0, 186), txt_color=(255, 255, 255)):
    """
    Function for drawing a sweep annotation line and an optional label.

    Args:
        line_x (int): The x-coordinate of the sweep line.
        line_y (int): The y-coordinate limit of the sweep line.
        label (str, optional): Text label to be drawn in center of sweep line. If None, no label is drawn.
        color (tuple): RGB color for the line and label background.
        txt_color (tuple): RGB color for the label text.
    """
    # Draw the sweep line
    cv2.line(self.im, (line_x, 0), (line_x, line_y), color, self.tf * 2)

    # Draw label, if provided
    if label:
        (text_width, text_height), _ = cv2.getTextSize(label, cv2.FONT_HERSHEY_SIMPLEX, self.sf, self.tf)
        cv2.rectangle(
            self.im,
            (line_x - text_width // 2 - 10, line_y // 2 - text_height // 2 - 10),
            (line_x + text_width // 2 + 10, line_y // 2 + text_height // 2 + 10),
            color,
            -1,
        )
        cv2.putText(
            self.im,
            label,
            (line_x - text_width // 2, line_y // 2 + text_height // 2),
            cv2.FONT_HERSHEY_SIMPLEX,
            self.sf,
            txt_color,
            self.tf,
        )

text

text(xy, text, txt_color=(255, 255, 255), anchor='top', box_style=False)

Adds text to an image using PIL or cv2.

Source code in ultralytics/utils/plotting.py
def text(self, xy, text, txt_color=(255, 255, 255), anchor="top", box_style=False):
    """Adds text to an image using PIL or cv2."""
    if anchor == "bottom":  # start y from font bottom
        w, h = self.font.getsize(text)  # text width, height
        xy[1] += 1 - h
    if self.pil:
        if box_style:
            w, h = self.font.getsize(text)
            self.draw.rectangle((xy[0], xy[1], xy[0] + w + 1, xy[1] + h + 1), fill=txt_color)
            # Using `txt_color` for background and draw fg with white color
            txt_color = (255, 255, 255)
        if "\n" in text:
            lines = text.split("\n")
            _, h = self.font.getsize(text)
            for line in lines:
                self.draw.text(xy, line, fill=txt_color, font=self.font)
                xy[1] += h
        else:
            self.draw.text(xy, text, fill=txt_color, font=self.font)
    else:
        if box_style:
            w, h = cv2.getTextSize(text, 0, fontScale=self.sf, thickness=self.tf)[0]  # text width, height
            h += 3  # add pixels to pad text
            outside = xy[1] >= h  # label fits outside box
            p2 = xy[0] + w, xy[1] - h if outside else xy[1] + h
            cv2.rectangle(self.im, xy, p2, txt_color, -1, cv2.LINE_AA)  # filled
            # Using `txt_color` for background and draw fg with white color
            txt_color = (255, 255, 255)
        cv2.putText(self.im, text, xy, 0, self.sf, txt_color, thickness=self.tf, lineType=cv2.LINE_AA)

text_label

text_label(
    box, label="", color=(128, 128, 128), txt_color=(255, 255, 255), margin=5
)

Draws a label with a background rectangle centered within a given bounding box.

Parameters:

Name Type Description Default
box tuple

The bounding box coordinates (x1, y1, x2, y2).

required
label str

The text label to be displayed.

''
color tuple

The background color of the rectangle (B, G, R).

(128, 128, 128)
txt_color tuple

The color of the text (R, G, B).

(255, 255, 255)
margin int

The margin between the text and the rectangle border.

5
Source code in ultralytics/utils/plotting.py
def text_label(self, box, label="", color=(128, 128, 128), txt_color=(255, 255, 255), margin=5):
    """
    Draws a label with a background rectangle centered within a given bounding box.

    Args:
        box (tuple): The bounding box coordinates (x1, y1, x2, y2).
        label (str): The text label to be displayed.
        color (tuple, optional): The background color of the rectangle (B, G, R).
        txt_color (tuple, optional): The color of the text (R, G, B).
        margin (int, optional): The margin between the text and the rectangle border.
    """
    # Calculate the center of the bounding box
    x_center, y_center = int((box[0] + box[2]) / 2), int((box[1] + box[3]) / 2)
    # Get the size of the text
    text_size = cv2.getTextSize(label, cv2.FONT_HERSHEY_SIMPLEX, self.sf - 0.1, self.tf)[0]
    # Calculate the top-left corner of the text (to center it)
    text_x = x_center - text_size[0] // 2
    text_y = y_center + text_size[1] // 2
    # Calculate the coordinates of the background rectangle
    rect_x1 = text_x - margin
    rect_y1 = text_y - text_size[1] - margin
    rect_x2 = text_x + text_size[0] + margin
    rect_y2 = text_y + margin
    # Draw the background rectangle
    cv2.rectangle(self.im, (rect_x1, rect_y1), (rect_x2, rect_y2), color, -1)
    # Draw the text on top of the rectangle
    cv2.putText(
        self.im,
        label,
        (text_x, text_y),
        cv2.FONT_HERSHEY_SIMPLEX,
        self.sf - 0.1,
        self.get_txt_color(color, txt_color),
        self.tf,
        lineType=cv2.LINE_AA,
    )

visioneye

visioneye(box, center_point, color=(235, 219, 11), pin_color=(255, 0, 255))

Function for pinpoint human-vision eye mapping and plotting.

Parameters:

Name Type Description Default
box list

Bounding box coordinates

required
center_point tuple

center point for vision eye view

required
color tuple

object centroid and line color value

(235, 219, 11)
pin_color tuple

visioneye point color value

(255, 0, 255)
Source code in ultralytics/utils/plotting.py
def visioneye(self, box, center_point, color=(235, 219, 11), pin_color=(255, 0, 255)):
    """
    Function for pinpoint human-vision eye mapping and plotting.

    Args:
        box (list): Bounding box coordinates
        center_point (tuple): center point for vision eye view
        color (tuple): object centroid and line color value
        pin_color (tuple): visioneye point color value
    """
    center_bbox = int((box[0] + box[2]) / 2), int((box[1] + box[3]) / 2)
    cv2.circle(self.im, center_point, self.tf * 2, pin_color, -1)
    cv2.circle(self.im, center_bbox, self.tf * 2, color, -1)
    cv2.line(self.im, center_point, center_bbox, color, self.tf)





ultralytics.utils.plotting.plot_labels

plot_labels(boxes, cls, names=(), save_dir=Path(''), on_plot=None)

Plot training labels including class histograms and box statistics.

Source code in ultralytics/utils/plotting.py
@TryExcept()  # known issue https://github.com/ultralytics/yolov5/issues/5395
@plt_settings()
def plot_labels(boxes, cls, names=(), save_dir=Path(""), on_plot=None):
    """Plot training labels including class histograms and box statistics."""
    import pandas  # scope for faster 'import ultralytics'
    import seaborn  # scope for faster 'import ultralytics'

    # Filter matplotlib>=3.7.2 warning and Seaborn use_inf and is_categorical FutureWarnings
    warnings.filterwarnings("ignore", category=UserWarning, message="The figure layout has changed to tight")
    warnings.filterwarnings("ignore", category=FutureWarning)

    # Plot dataset labels
    LOGGER.info(f"Plotting labels to {save_dir / 'labels.jpg'}... ")
    nc = int(cls.max() + 1)  # number of classes
    boxes = boxes[:1000000]  # limit to 1M boxes
    x = pandas.DataFrame(boxes, columns=["x", "y", "width", "height"])

    # Seaborn correlogram
    seaborn.pairplot(x, corner=True, diag_kind="auto", kind="hist", diag_kws=dict(bins=50), plot_kws=dict(pmax=0.9))
    plt.savefig(save_dir / "labels_correlogram.jpg", dpi=200)
    plt.close()

    # Matplotlib labels
    ax = plt.subplots(2, 2, figsize=(8, 8), tight_layout=True)[1].ravel()
    y = ax[0].hist(cls, bins=np.linspace(0, nc, nc + 1) - 0.5, rwidth=0.8)
    for i in range(nc):
        y[2].patches[i].set_color([x / 255 for x in colors(i)])
    ax[0].set_ylabel("instances")
    if 0 < len(names) < 30:
        ax[0].set_xticks(range(len(names)))
        ax[0].set_xticklabels(list(names.values()), rotation=90, fontsize=10)
    else:
        ax[0].set_xlabel("classes")
    seaborn.histplot(x, x="x", y="y", ax=ax[2], bins=50, pmax=0.9)
    seaborn.histplot(x, x="width", y="height", ax=ax[3], bins=50, pmax=0.9)

    # Rectangles
    boxes[:, 0:2] = 0.5  # center
    boxes = ops.xywh2xyxy(boxes) * 1000
    img = Image.fromarray(np.ones((1000, 1000, 3), dtype=np.uint8) * 255)
    for cls, box in zip(cls[:500], boxes[:500]):
        ImageDraw.Draw(img).rectangle(box, width=1, outline=colors(cls))  # plot
    ax[1].imshow(img)
    ax[1].axis("off")

    for a in [0, 1, 2, 3]:
        for s in ["top", "right", "left", "bottom"]:
            ax[a].spines[s].set_visible(False)

    fname = save_dir / "labels.jpg"
    plt.savefig(fname, dpi=200)
    plt.close()
    if on_plot:
        on_plot(fname)





ultralytics.utils.plotting.save_one_box

save_one_box(
    xyxy,
    im,
    file=Path("im.jpg"),
    gain=1.02,
    pad=10,
    square=False,
    BGR=False,
    save=True,
)

Save image crop as {file} with crop size multiple {gain} and {pad} pixels. Save and/or return crop.

This function takes a bounding box and an image, and then saves a cropped portion of the image according to the bounding box. Optionally, the crop can be squared, and the function allows for gain and padding adjustments to the bounding box.

Parameters:

Name Type Description Default
xyxy Tensor or list

A tensor or list representing the bounding box in xyxy format.

required
im ndarray

The input image.

required
file Path

The path where the cropped image will be saved. Defaults to 'im.jpg'.

Path('im.jpg')
gain float

A multiplicative factor to increase the size of the bounding box. Defaults to 1.02.

1.02
pad int

The number of pixels to add to the width and height of the bounding box. Defaults to 10.

10
square bool

If True, the bounding box will be transformed into a square. Defaults to False.

False
BGR bool

If True, the image will be saved in BGR format, otherwise in RGB. Defaults to False.

False
save bool

If True, the cropped image will be saved to disk. Defaults to True.

True

Returns:

Type Description
ndarray

The cropped image.

Example
from ultralytics.utils.plotting import save_one_box

xyxy = [50, 50, 150, 150]
im = cv2.imread("image.jpg")
cropped_im = save_one_box(xyxy, im, file="cropped.jpg", square=True)
Source code in ultralytics/utils/plotting.py
def save_one_box(xyxy, im, file=Path("im.jpg"), gain=1.02, pad=10, square=False, BGR=False, save=True):
    """
    Save image crop as {file} with crop size multiple {gain} and {pad} pixels. Save and/or return crop.

    This function takes a bounding box and an image, and then saves a cropped portion of the image according
    to the bounding box. Optionally, the crop can be squared, and the function allows for gain and padding
    adjustments to the bounding box.

    Args:
        xyxy (torch.Tensor or list): A tensor or list representing the bounding box in xyxy format.
        im (numpy.ndarray): The input image.
        file (Path, optional): The path where the cropped image will be saved. Defaults to 'im.jpg'.
        gain (float, optional): A multiplicative factor to increase the size of the bounding box. Defaults to 1.02.
        pad (int, optional): The number of pixels to add to the width and height of the bounding box. Defaults to 10.
        square (bool, optional): If True, the bounding box will be transformed into a square. Defaults to False.
        BGR (bool, optional): If True, the image will be saved in BGR format, otherwise in RGB. Defaults to False.
        save (bool, optional): If True, the cropped image will be saved to disk. Defaults to True.

    Returns:
        (numpy.ndarray): The cropped image.

    Example:
        ```python
        from ultralytics.utils.plotting import save_one_box

        xyxy = [50, 50, 150, 150]
        im = cv2.imread("image.jpg")
        cropped_im = save_one_box(xyxy, im, file="cropped.jpg", square=True)
        ```
    """
    if not isinstance(xyxy, torch.Tensor):  # may be list
        xyxy = torch.stack(xyxy)
    b = ops.xyxy2xywh(xyxy.view(-1, 4))  # boxes
    if square:
        b[:, 2:] = b[:, 2:].max(1)[0].unsqueeze(1)  # attempt rectangle to square
    b[:, 2:] = b[:, 2:] * gain + pad  # box wh * gain + pad
    xyxy = ops.xywh2xyxy(b).long()
    xyxy = ops.clip_boxes(xyxy, im.shape)
    crop = im[int(xyxy[0, 1]) : int(xyxy[0, 3]), int(xyxy[0, 0]) : int(xyxy[0, 2]), :: (1 if BGR else -1)]
    if save:
        file.parent.mkdir(parents=True, exist_ok=True)  # make directory
        f = str(increment_path(file).with_suffix(".jpg"))
        # cv2.imwrite(f, crop)  # save BGR, https://github.com/ultralytics/yolov5/issues/7007 chroma subsampling issue
        Image.fromarray(crop[..., ::-1]).save(f, quality=95, subsampling=0)  # save RGB
    return crop





ultralytics.utils.plotting.plot_images

plot_images(
    images: Union[torch.Tensor, np.ndarray],
    batch_idx: Union[torch.Tensor, np.ndarray],
    cls: Union[torch.Tensor, np.ndarray],
    bboxes: Union[torch.Tensor, np.ndarray] = np.zeros(0, dtype=np.float32),
    confs: Optional[Union[torch.Tensor, np.ndarray]] = None,
    masks: Union[torch.Tensor, np.ndarray] = np.zeros(0, dtype=np.uint8),
    kpts: Union[torch.Tensor, np.ndarray] = np.zeros((0, 51), dtype=np.float32),
    paths: Optional[List[str]] = None,
    fname: str = "images.jpg",
    names: Optional[Dict[int, str]] = None,
    on_plot: Optional[Callable] = None,
    max_size: int = 1920,
    max_subplots: int = 16,
    save: bool = True,
    conf_thres: float = 0.25,
) -> Optional[np.ndarray]

Plot image grid with labels, bounding boxes, masks, and keypoints.

Parameters:

Name Type Description Default
images Union[Tensor, ndarray]

Batch of images to plot. Shape: (batch_size, channels, height, width).

required
batch_idx Union[Tensor, ndarray]

Batch indices for each detection. Shape: (num_detections,).

required
cls Union[Tensor, ndarray]

Class labels for each detection. Shape: (num_detections,).

required
bboxes Union[Tensor, ndarray]

Bounding boxes for each detection. Shape: (num_detections, 4) or (num_detections, 5) for rotated boxes.

zeros(0, dtype=float32)
confs Optional[Union[Tensor, ndarray]]

Confidence scores for each detection. Shape: (num_detections,).

None
masks Union[Tensor, ndarray]

Instance segmentation masks. Shape: (num_detections, height, width) or (1, height, width).

zeros(0, dtype=uint8)
kpts Union[Tensor, ndarray]

Keypoints for each detection. Shape: (num_detections, 51).

zeros((0, 51), dtype=float32)
paths Optional[List[str]]

List of file paths for each image in the batch.

None
fname str

Output filename for the plotted image grid.

'images.jpg'
names Optional[Dict[int, str]]

Dictionary mapping class indices to class names.

None
on_plot Optional[Callable]

Optional callback function to be called after saving the plot.

None
max_size int

Maximum size of the output image grid.

1920
max_subplots int

Maximum number of subplots in the image grid.

16
save bool

Whether to save the plotted image grid to a file.

True
conf_thres float

Confidence threshold for displaying detections.

0.25

Returns:

Type Description
Optional[ndarray]

np.ndarray: Plotted image grid as a numpy array if save is False, None otherwise.

Note

This function supports both tensor and numpy array inputs. It will automatically convert tensor inputs to numpy arrays for processing.

Source code in ultralytics/utils/plotting.py
@threaded
def plot_images(
    images: Union[torch.Tensor, np.ndarray],
    batch_idx: Union[torch.Tensor, np.ndarray],
    cls: Union[torch.Tensor, np.ndarray],
    bboxes: Union[torch.Tensor, np.ndarray] = np.zeros(0, dtype=np.float32),
    confs: Optional[Union[torch.Tensor, np.ndarray]] = None,
    masks: Union[torch.Tensor, np.ndarray] = np.zeros(0, dtype=np.uint8),
    kpts: Union[torch.Tensor, np.ndarray] = np.zeros((0, 51), dtype=np.float32),
    paths: Optional[List[str]] = None,
    fname: str = "images.jpg",
    names: Optional[Dict[int, str]] = None,
    on_plot: Optional[Callable] = None,
    max_size: int = 1920,
    max_subplots: int = 16,
    save: bool = True,
    conf_thres: float = 0.25,
) -> Optional[np.ndarray]:
    """
    Plot image grid with labels, bounding boxes, masks, and keypoints.

    Args:
        images: Batch of images to plot. Shape: (batch_size, channels, height, width).
        batch_idx: Batch indices for each detection. Shape: (num_detections,).
        cls: Class labels for each detection. Shape: (num_detections,).
        bboxes: Bounding boxes for each detection. Shape: (num_detections, 4) or (num_detections, 5) for rotated boxes.
        confs: Confidence scores for each detection. Shape: (num_detections,).
        masks: Instance segmentation masks. Shape: (num_detections, height, width) or (1, height, width).
        kpts: Keypoints for each detection. Shape: (num_detections, 51).
        paths: List of file paths for each image in the batch.
        fname: Output filename for the plotted image grid.
        names: Dictionary mapping class indices to class names.
        on_plot: Optional callback function to be called after saving the plot.
        max_size: Maximum size of the output image grid.
        max_subplots: Maximum number of subplots in the image grid.
        save: Whether to save the plotted image grid to a file.
        conf_thres: Confidence threshold for displaying detections.

    Returns:
        np.ndarray: Plotted image grid as a numpy array if save is False, None otherwise.

    Note:
        This function supports both tensor and numpy array inputs. It will automatically
        convert tensor inputs to numpy arrays for processing.
    """
    if isinstance(images, torch.Tensor):
        images = images.cpu().float().numpy()
    if isinstance(cls, torch.Tensor):
        cls = cls.cpu().numpy()
    if isinstance(bboxes, torch.Tensor):
        bboxes = bboxes.cpu().numpy()
    if isinstance(masks, torch.Tensor):
        masks = masks.cpu().numpy().astype(int)
    if isinstance(kpts, torch.Tensor):
        kpts = kpts.cpu().numpy()
    if isinstance(batch_idx, torch.Tensor):
        batch_idx = batch_idx.cpu().numpy()

    bs, _, h, w = images.shape  # batch size, _, height, width
    bs = min(bs, max_subplots)  # limit plot images
    ns = np.ceil(bs**0.5)  # number of subplots (square)
    if np.max(images[0]) <= 1:
        images *= 255  # de-normalise (optional)

    # Build Image
    mosaic = np.full((int(ns * h), int(ns * w), 3), 255, dtype=np.uint8)  # init
    for i in range(bs):
        x, y = int(w * (i // ns)), int(h * (i % ns))  # block origin
        mosaic[y : y + h, x : x + w, :] = images[i].transpose(1, 2, 0)

    # Resize (optional)
    scale = max_size / ns / max(h, w)
    if scale < 1:
        h = math.ceil(scale * h)
        w = math.ceil(scale * w)
        mosaic = cv2.resize(mosaic, tuple(int(x * ns) for x in (w, h)))

    # Annotate
    fs = int((h + w) * ns * 0.01)  # font size
    annotator = Annotator(mosaic, line_width=round(fs / 10), font_size=fs, pil=True, example=names)
    for i in range(bs):
        x, y = int(w * (i // ns)), int(h * (i % ns))  # block origin
        annotator.rectangle([x, y, x + w, y + h], None, (255, 255, 255), width=2)  # borders
        if paths:
            annotator.text((x + 5, y + 5), text=Path(paths[i]).name[:40], txt_color=(220, 220, 220))  # filenames
        if len(cls) > 0:
            idx = batch_idx == i
            classes = cls[idx].astype("int")
            labels = confs is None

            if len(bboxes):
                boxes = bboxes[idx]
                conf = confs[idx] if confs is not None else None  # check for confidence presence (label vs pred)
                if len(boxes):
                    if boxes[:, :4].max() <= 1.1:  # if normalized with tolerance 0.1
                        boxes[..., [0, 2]] *= w  # scale to pixels
                        boxes[..., [1, 3]] *= h
                    elif scale < 1:  # absolute coords need scale if image scales
                        boxes[..., :4] *= scale
                boxes[..., 0] += x
                boxes[..., 1] += y
                is_obb = boxes.shape[-1] == 5  # xywhr
                boxes = ops.xywhr2xyxyxyxy(boxes) if is_obb else ops.xywh2xyxy(boxes)
                for j, box in enumerate(boxes.astype(np.int64).tolist()):
                    c = classes[j]
                    color = colors(c)
                    c = names.get(c, c) if names else c
                    if labels or conf[j] > conf_thres:
                        label = f"{c}" if labels else f"{c} {conf[j]:.1f}"
                        annotator.box_label(box, label, color=color, rotated=is_obb)

            elif len(classes):
                for c in classes:
                    color = colors(c)
                    c = names.get(c, c) if names else c
                    annotator.text((x, y), f"{c}", txt_color=color, box_style=True)

            # Plot keypoints
            if len(kpts):
                kpts_ = kpts[idx].copy()
                if len(kpts_):
                    if kpts_[..., 0].max() <= 1.01 or kpts_[..., 1].max() <= 1.01:  # if normalized with tolerance .01
                        kpts_[..., 0] *= w  # scale to pixels
                        kpts_[..., 1] *= h
                    elif scale < 1:  # absolute coords need scale if image scales
                        kpts_ *= scale
                kpts_[..., 0] += x
                kpts_[..., 1] += y
                for j in range(len(kpts_)):
                    if labels or conf[j] > conf_thres:
                        annotator.kpts(kpts_[j], conf_thres=conf_thres)

            # Plot masks
            if len(masks):
                if idx.shape[0] == masks.shape[0]:  # overlap_masks=False
                    image_masks = masks[idx]
                else:  # overlap_masks=True
                    image_masks = masks[[i]]  # (1, 640, 640)
                    nl = idx.sum()
                    index = np.arange(nl).reshape((nl, 1, 1)) + 1
                    image_masks = np.repeat(image_masks, nl, axis=0)
                    image_masks = np.where(image_masks == index, 1.0, 0.0)

                im = np.asarray(annotator.im).copy()
                for j in range(len(image_masks)):
                    if labels or conf[j] > conf_thres:
                        color = colors(classes[j])
                        mh, mw = image_masks[j].shape
                        if mh != h or mw != w:
                            mask = image_masks[j].astype(np.uint8)
                            mask = cv2.resize(mask, (w, h))
                            mask = mask.astype(bool)
                        else:
                            mask = image_masks[j].astype(bool)
                        try:
                            im[y : y + h, x : x + w, :][mask] = (
                                im[y : y + h, x : x + w, :][mask] * 0.4 + np.array(color) * 0.6
                            )
                        except Exception:
                            pass
                annotator.fromarray(im)
    if not save:
        return np.asarray(annotator.im)
    annotator.im.save(fname)  # save
    if on_plot:
        on_plot(fname)





ultralytics.utils.plotting.plot_results

plot_results(
    file="path/to/results.csv",
    dir="",
    segment=False,
    pose=False,
    classify=False,
    on_plot=None,
)

Plot training results from a results CSV file. The function supports various types of data including segmentation, pose estimation, and classification. Plots are saved as 'results.png' in the directory where the CSV is located.

Parameters:

Name Type Description Default
file str

Path to the CSV file containing the training results. Defaults to 'path/to/results.csv'.

'path/to/results.csv'
dir str

Directory where the CSV file is located if 'file' is not provided. Defaults to ''.

''
segment bool

Flag to indicate if the data is for segmentation. Defaults to False.

False
pose bool

Flag to indicate if the data is for pose estimation. Defaults to False.

False
classify bool

Flag to indicate if the data is for classification. Defaults to False.

False
on_plot callable

Callback function to be executed after plotting. Takes filename as an argument. Defaults to None.

None
Example
from ultralytics.utils.plotting import plot_results

plot_results("path/to/results.csv", segment=True)
Source code in ultralytics/utils/plotting.py
@plt_settings()
def plot_results(file="path/to/results.csv", dir="", segment=False, pose=False, classify=False, on_plot=None):
    """
    Plot training results from a results CSV file. The function supports various types of data including segmentation,
    pose estimation, and classification. Plots are saved as 'results.png' in the directory where the CSV is located.

    Args:
        file (str, optional): Path to the CSV file containing the training results. Defaults to 'path/to/results.csv'.
        dir (str, optional): Directory where the CSV file is located if 'file' is not provided. Defaults to ''.
        segment (bool, optional): Flag to indicate if the data is for segmentation. Defaults to False.
        pose (bool, optional): Flag to indicate if the data is for pose estimation. Defaults to False.
        classify (bool, optional): Flag to indicate if the data is for classification. Defaults to False.
        on_plot (callable, optional): Callback function to be executed after plotting. Takes filename as an argument.
            Defaults to None.

    Example:
        ```python
        from ultralytics.utils.plotting import plot_results

        plot_results("path/to/results.csv", segment=True)
        ```
    """
    import pandas as pd  # scope for faster 'import ultralytics'
    from scipy.ndimage import gaussian_filter1d

    save_dir = Path(file).parent if file else Path(dir)
    if classify:
        fig, ax = plt.subplots(2, 2, figsize=(6, 6), tight_layout=True)
        index = [2, 5, 3, 4]
    elif segment:
        fig, ax = plt.subplots(2, 8, figsize=(18, 6), tight_layout=True)
        index = [2, 3, 4, 5, 6, 7, 10, 11, 14, 15, 16, 17, 8, 9, 12, 13]
    elif pose:
        fig, ax = plt.subplots(2, 9, figsize=(21, 6), tight_layout=True)
        index = [2, 3, 4, 5, 6, 7, 8, 11, 12, 15, 16, 17, 18, 19, 9, 10, 13, 14]
    else:
        fig, ax = plt.subplots(2, 5, figsize=(12, 6), tight_layout=True)
        index = [2, 3, 4, 5, 6, 9, 10, 11, 7, 8]
    ax = ax.ravel()
    files = list(save_dir.glob("results*.csv"))
    assert len(files), f"No results.csv files found in {save_dir.resolve()}, nothing to plot."
    for f in files:
        try:
            data = pd.read_csv(f)
            s = [x.strip() for x in data.columns]
            x = data.values[:, 0]
            for i, j in enumerate(index):
                y = data.values[:, j].astype("float")
                # y[y == 0] = np.nan  # don't show zero values
                ax[i].plot(x, y, marker=".", label=f.stem, linewidth=2, markersize=8)  # actual results
                ax[i].plot(x, gaussian_filter1d(y, sigma=3), ":", label="smooth", linewidth=2)  # smoothing line
                ax[i].set_title(s[j], fontsize=12)
                # if j in {8, 9, 10}:  # share train and val loss y axes
                #     ax[i].get_shared_y_axes().join(ax[i], ax[i - 5])
        except Exception as e:
            LOGGER.warning(f"WARNING: Plotting error for {f}: {e}")
    ax[1].legend()
    fname = save_dir / "results.png"
    fig.savefig(fname, dpi=200)
    plt.close()
    if on_plot:
        on_plot(fname)





ultralytics.utils.plotting.plt_color_scatter

plt_color_scatter(v, f, bins=20, cmap='viridis', alpha=0.8, edgecolors='none')

Plots a scatter plot with points colored based on a 2D histogram.

Parameters:

Name Type Description Default
v array - like

Values for the x-axis.

required
f array - like

Values for the y-axis.

required
bins int

Number of bins for the histogram. Defaults to 20.

20
cmap str

Colormap for the scatter plot. Defaults to 'viridis'.

'viridis'
alpha float

Alpha for the scatter plot. Defaults to 0.8.

0.8
edgecolors str

Edge colors for the scatter plot. Defaults to 'none'.

'none'

Examples:

>>> v = np.random.rand(100)
>>> f = np.random.rand(100)
>>> plt_color_scatter(v, f)
Source code in ultralytics/utils/plotting.py
def plt_color_scatter(v, f, bins=20, cmap="viridis", alpha=0.8, edgecolors="none"):
    """
    Plots a scatter plot with points colored based on a 2D histogram.

    Args:
        v (array-like): Values for the x-axis.
        f (array-like): Values for the y-axis.
        bins (int, optional): Number of bins for the histogram. Defaults to 20.
        cmap (str, optional): Colormap for the scatter plot. Defaults to 'viridis'.
        alpha (float, optional): Alpha for the scatter plot. Defaults to 0.8.
        edgecolors (str, optional): Edge colors for the scatter plot. Defaults to 'none'.

    Examples:
        >>> v = np.random.rand(100)
        >>> f = np.random.rand(100)
        >>> plt_color_scatter(v, f)
    """
    # Calculate 2D histogram and corresponding colors
    hist, xedges, yedges = np.histogram2d(v, f, bins=bins)
    colors = [
        hist[
            min(np.digitize(v[i], xedges, right=True) - 1, hist.shape[0] - 1),
            min(np.digitize(f[i], yedges, right=True) - 1, hist.shape[1] - 1),
        ]
        for i in range(len(v))
    ]

    # Scatter plot
    plt.scatter(v, f, c=colors, cmap=cmap, alpha=alpha, edgecolors=edgecolors)





ultralytics.utils.plotting.plot_tune_results

plot_tune_results(csv_file='tune_results.csv')

Plot the evolution results stored in an 'tune_results.csv' file. The function generates a scatter plot for each key in the CSV, color-coded based on fitness scores. The best-performing configurations are highlighted on the plots.

Parameters:

Name Type Description Default
csv_file str

Path to the CSV file containing the tuning results. Defaults to 'tune_results.csv'.

'tune_results.csv'

Examples:

>>> plot_tune_results("path/to/tune_results.csv")
Source code in ultralytics/utils/plotting.py
def plot_tune_results(csv_file="tune_results.csv"):
    """
    Plot the evolution results stored in an 'tune_results.csv' file. The function generates a scatter plot for each key
    in the CSV, color-coded based on fitness scores. The best-performing configurations are highlighted on the plots.

    Args:
        csv_file (str, optional): Path to the CSV file containing the tuning results. Defaults to 'tune_results.csv'.

    Examples:
        >>> plot_tune_results("path/to/tune_results.csv")
    """
    import pandas as pd  # scope for faster 'import ultralytics'
    from scipy.ndimage import gaussian_filter1d

    def _save_one_file(file):
        """Save one matplotlib plot to 'file'."""
        plt.savefig(file, dpi=200)
        plt.close()
        LOGGER.info(f"Saved {file}")

    # Scatter plots for each hyperparameter
    csv_file = Path(csv_file)
    data = pd.read_csv(csv_file)
    num_metrics_columns = 1
    keys = [x.strip() for x in data.columns][num_metrics_columns:]
    x = data.values
    fitness = x[:, 0]  # fitness
    j = np.argmax(fitness)  # max fitness index
    n = math.ceil(len(keys) ** 0.5)  # columns and rows in plot
    plt.figure(figsize=(10, 10), tight_layout=True)
    for i, k in enumerate(keys):
        v = x[:, i + num_metrics_columns]
        mu = v[j]  # best single result
        plt.subplot(n, n, i + 1)
        plt_color_scatter(v, fitness, cmap="viridis", alpha=0.8, edgecolors="none")
        plt.plot(mu, fitness.max(), "k+", markersize=15)
        plt.title(f"{k} = {mu:.3g}", fontdict={"size": 9})  # limit to 40 characters
        plt.tick_params(axis="both", labelsize=8)  # Set axis label size to 8
        if i % n != 0:
            plt.yticks([])
    _save_one_file(csv_file.with_name("tune_scatter_plots.png"))

    # Fitness vs iteration
    x = range(1, len(fitness) + 1)
    plt.figure(figsize=(10, 6), tight_layout=True)
    plt.plot(x, fitness, marker="o", linestyle="none", label="fitness")
    plt.plot(x, gaussian_filter1d(fitness, sigma=3), ":", label="smoothed", linewidth=2)  # smoothing line
    plt.title("Fitness vs Iteration")
    plt.xlabel("Iteration")
    plt.ylabel("Fitness")
    plt.grid(True)
    plt.legend()
    _save_one_file(csv_file.with_name("tune_fitness.png"))





ultralytics.utils.plotting.output_to_target

output_to_target(output, max_det=300)

Convert model output to target format [batch_id, class_id, x, y, w, h, conf] for plotting.

Source code in ultralytics/utils/plotting.py
def output_to_target(output, max_det=300):
    """Convert model output to target format [batch_id, class_id, x, y, w, h, conf] for plotting."""
    targets = []
    for i, o in enumerate(output):
        box, conf, cls = o[:max_det, :6].cpu().split((4, 1, 1), 1)
        j = torch.full((conf.shape[0], 1), i)
        targets.append(torch.cat((j, cls, ops.xyxy2xywh(box), conf), 1))
    targets = torch.cat(targets, 0).numpy()
    return targets[:, 0], targets[:, 1], targets[:, 2:-1], targets[:, -1]





ultralytics.utils.plotting.output_to_rotated_target

output_to_rotated_target(output, max_det=300)

Convert model output to target format [batch_id, class_id, x, y, w, h, conf] for plotting.

Source code in ultralytics/utils/plotting.py
def output_to_rotated_target(output, max_det=300):
    """Convert model output to target format [batch_id, class_id, x, y, w, h, conf] for plotting."""
    targets = []
    for i, o in enumerate(output):
        box, conf, cls, angle = o[:max_det].cpu().split((4, 1, 1, 1), 1)
        j = torch.full((conf.shape[0], 1), i)
        targets.append(torch.cat((j, cls, box, angle, conf), 1))
    targets = torch.cat(targets, 0).numpy()
    return targets[:, 0], targets[:, 1], targets[:, 2:-1], targets[:, -1]





ultralytics.utils.plotting.feature_visualization

feature_visualization(
    x, module_type, stage, n=32, save_dir=Path("runs/detect/exp")
)

Visualize feature maps of a given model module during inference.

Parameters:

Name Type Description Default
x Tensor

Features to be visualized.

required
module_type str

Module type.

required
stage int

Module stage within the model.

required
n int

Maximum number of feature maps to plot. Defaults to 32.

32
save_dir Path

Directory to save results. Defaults to Path('runs/detect/exp').

Path('runs/detect/exp')
Source code in ultralytics/utils/plotting.py
def feature_visualization(x, module_type, stage, n=32, save_dir=Path("runs/detect/exp")):
    """
    Visualize feature maps of a given model module during inference.

    Args:
        x (torch.Tensor): Features to be visualized.
        module_type (str): Module type.
        stage (int): Module stage within the model.
        n (int, optional): Maximum number of feature maps to plot. Defaults to 32.
        save_dir (Path, optional): Directory to save results. Defaults to Path('runs/detect/exp').
    """
    for m in {"Detect", "Segment", "Pose", "Classify", "OBB", "RTDETRDecoder"}:  # all model heads
        if m in module_type:
            return
    if isinstance(x, torch.Tensor):
        _, channels, height, width = x.shape  # batch, channels, height, width
        if height > 1 and width > 1:
            f = save_dir / f"stage{stage}_{module_type.split('.')[-1]}_features.png"  # filename

            blocks = torch.chunk(x[0].cpu(), channels, dim=0)  # select batch index 0, block by channels
            n = min(n, channels)  # number of plots
            _, ax = plt.subplots(math.ceil(n / 8), 8, tight_layout=True)  # 8 rows x n/8 cols
            ax = ax.ravel()
            plt.subplots_adjust(wspace=0.05, hspace=0.05)
            for i in range(n):
                ax[i].imshow(blocks[i].squeeze())  # cmap='gray'
                ax[i].axis("off")

            LOGGER.info(f"Saving {f}... ({n}/{channels})")
            plt.savefig(f, dpi=300, bbox_inches="tight")
            plt.close()
            np.save(str(f.with_suffix(".npy")), x[0].cpu().numpy())  # npy save



📅 Created 1 year ago ✏️ Updated 2 months ago