高级数据可视化：使用Ultralytics YOLOv8 绘制热图 🚀

热图简介

生成的热图 Ultralytics YOLOv8将复杂的数据转换成生动的彩色编码矩阵。这种可视化工具采用色谱来表示不同的数据值，暖色调表示较高的强度，冷色调表示较低的值。热图在可视化复杂的数据模式、相关性和异常情况方面表现出色，为不同领域的数据解读提供了一种易于理解且引人入胜的方法。

观看： 使用热图Ultralytics YOLOv8

为什么选择热图进行数据分析？

直观的数据分布可视化：热图可简化对数据集中和分布的理解，将复杂的数据集转换为易于理解的可视化格式。
高效模式检测：通过热图格式的可视化数据，可以更容易地发现趋势、群组和异常值，从而促进更快地分析和洞察。
增强空间分析和决策：热图有助于说明空间关系，帮助商业智能、环境研究和城市规划等领域的决策过程。

真实世界的应用

交通运输	零售

Ultralytics YOLOv8 交通热图	Ultralytics YOLOv8 零售业热图

热图配置

heatmap_alpha:确保该值在 (0.0 - 1.0) 范围内。
decay_factor:用于在帧中不再存在对象后删除热图，其值也应在 (0.0 - 1.0) 范围内。

热图使用Ultralytics YOLOv8 示例

热图数线多边形计数地区计数Im0特定等级

from ultralytics import YOLO
from ultralytics.solutions import heatmap
import cv2

model = YOLO("yolov8n.pt")
cap = cv2.VideoCapture("path/to/video/file.mp4")
assert cap.isOpened(), "Error reading video file"
w, h, fps = (int(cap.get(x)) for x in (cv2.CAP_PROP_FRAME_WIDTH, cv2.CAP_PROP_FRAME_HEIGHT, cv2.CAP_PROP_FPS))

# Video writer
video_writer = cv2.VideoWriter("heatmap_output.avi",
                               cv2.VideoWriter_fourcc(*'mp4v'),
                               fps,
                               (w, h))

# Init heatmap
heatmap_obj = heatmap.Heatmap()
heatmap_obj.set_args(colormap=cv2.COLORMAP_PARULA,
                     imw=w,
                     imh=h,
                     view_img=True,
                     shape="circle",
                     classes_names=model.names)

while cap.isOpened():
    success, im0 = cap.read()
    if not success:
        print("Video frame is empty or video processing has been successfully completed.")
        break
    tracks = model.track(im0, persist=True, show=False)

    im0 = heatmap_obj.generate_heatmap(im0, tracks)
    video_writer.write(im0)

cap.release()
video_writer.release()
cv2.destroyAllWindows()

from ultralytics import YOLO
from ultralytics.solutions import heatmap
import cv2

model = YOLO("yolov8n.pt")
cap = cv2.VideoCapture("path/to/video/file.mp4")
assert cap.isOpened(), "Error reading video file"
w, h, fps = (int(cap.get(x)) for x in (cv2.CAP_PROP_FRAME_WIDTH, cv2.CAP_PROP_FRAME_HEIGHT, cv2.CAP_PROP_FPS))

# Video writer
video_writer = cv2.VideoWriter("heatmap_output.avi",
                               cv2.VideoWriter_fourcc(*'mp4v'),
                               fps,
                               (w, h))

line_points = [(20, 400), (1080, 404)]  # line for object counting

# Init heatmap
heatmap_obj = heatmap.Heatmap()
heatmap_obj.set_args(colormap=cv2.COLORMAP_PARULA,
                     imw=w,
                     imh=h,
                     view_img=True,
                     shape="circle",
                     count_reg_pts=line_points,
                     classes_names=model.names)

while cap.isOpened():
    success, im0 = cap.read()
    if not success:
        print("Video frame is empty or video processing has been successfully completed.")
        break
    tracks = model.track(im0, persist=True, show=False)

    im0 = heatmap_obj.generate_heatmap(im0, tracks)
    video_writer.write(im0)

cap.release()
video_writer.release()
cv2.destroyAllWindows()

from ultralytics import YOLO
import heatmap
import cv2

model = YOLO("yolov8n.pt")
cap = cv2.VideoCapture("path/to/video/file.mp4")
assert cap.isOpened(), "Error reading video file"
w, h, fps = (int(cap.get(x)) for x in (cv2.CAP_PROP_FRAME_WIDTH, cv2.CAP_PROP_FRAME_HEIGHT, cv2.CAP_PROP_FPS))

# Video writer
video_writer = cv2.VideoWriter("heatmap_output.avi",
                               cv2.VideoWriter_fourcc(*'mp4v'),
                               fps,
                               (w, h))

# Define polygon points
region_points = [(20, 400), (1080, 404), (1080, 360), (20, 360), (20, 400)]

# Init heatmap
heatmap_obj = heatmap.Heatmap()
heatmap_obj.set_args(colormap=cv2.COLORMAP_PARULA,
                     imw=w,
                     imh=h,
                     view_img=True,
                     shape="circle",
                     count_reg_pts=region_points,
                     classes_names=model.names)

while cap.isOpened():
    success, im0 = cap.read()
    if not success:
        print("Video frame is empty or video processing has been successfully completed.")
        break
    tracks = model.track(im0, persist=True, show=False)

    im0 = heatmap_obj.generate_heatmap(im0, tracks)
    video_writer.write(im0)

cap.release()
video_writer.release()
cv2.destroyAllWindows()

from ultralytics import YOLO
from ultralytics.solutions import heatmap
import cv2

model = YOLO("yolov8n.pt")
cap = cv2.VideoCapture("path/to/video/file.mp4")
assert cap.isOpened(), "Error reading video file"
w, h, fps = (int(cap.get(x)) for x in (cv2.CAP_PROP_FRAME_WIDTH, cv2.CAP_PROP_FRAME_HEIGHT, cv2.CAP_PROP_FPS))

# Video writer
video_writer = cv2.VideoWriter("heatmap_output.avi",
                               cv2.VideoWriter_fourcc(*'mp4v'),
                               fps,
                               (w, h))

# Define region points
region_points = [(20, 400), (1080, 404), (1080, 360), (20, 360)]

# Init heatmap
heatmap_obj = heatmap.Heatmap()
heatmap_obj.set_args(colormap=cv2.COLORMAP_PARULA,
                     imw=w,
                     imh=h,
                     view_img=True,
                     shape="circle",
                     count_reg_pts=region_points,
                     classes_names=model.names)

while cap.isOpened():
    success, im0 = cap.read()
    if not success:
        print("Video frame is empty or video processing has been successfully completed.")
        break
    tracks = model.track(im0, persist=True, show=False)

    im0 = heatmap_obj.generate_heatmap(im0, tracks)
    video_writer.write(im0)

cap.release()
video_writer.release()
cv2.destroyAllWindows()

from ultralytics import YOLO
from ultralytics.solutions import heatmap
import cv2

model = YOLO("yolov8s.pt")   # YOLOv8 custom/pretrained model

im0 = cv2.imread("path/to/image.png")  # path to image file
h, w = im0.shape[:2]  # image height and width

# Heatmap Init
heatmap_obj = heatmap.Heatmap()
heatmap_obj.set_args(colormap=cv2.COLORMAP_PARULA,
                     imw=w,
                     imh=h,
                     view_img=True,
                     shape="circle",
                     classes_names=model.names)

results = model.track(im0, persist=True)
im0 = heatmap_obj.generate_heatmap(im0, tracks=results)
cv2.imwrite("ultralytics_output.png", im0)

from ultralytics import YOLO
from ultralytics.solutions import heatmap
import cv2

model = YOLO("yolov8n.pt")
cap = cv2.VideoCapture("path/to/video/file.mp4")
assert cap.isOpened(), "Error reading video file"
w, h, fps = (int(cap.get(x)) for x in (cv2.CAP_PROP_FRAME_WIDTH, cv2.CAP_PROP_FRAME_HEIGHT, cv2.CAP_PROP_FPS))

# Video writer
video_writer = cv2.VideoWriter("heatmap_output.avi",
                               cv2.VideoWriter_fourcc(*'mp4v'),
                               fps,
                               (w, h))

classes_for_heatmap = [0, 2]  # classes for heatmap

# Init heatmap
heatmap_obj = heatmap.Heatmap()
heatmap_obj.set_args(colormap=cv2.COLORMAP_PARULA,
                     imw=w,
                     imh=h,
                     view_img=True,
                     shape="circle",
                     classes_names=model.names)

while cap.isOpened():
    success, im0 = cap.read()
    if not success:
        print("Video frame is empty or video processing has been successfully completed.")
        break
    tracks = model.track(im0, persist=True, show=False,
                         classes=classes_for_heatmap)

    im0 = heatmap_obj.generate_heatmap(im0, tracks)
    video_writer.write(im0)

cap.release()
video_writer.release()
cv2.destroyAllWindows()

论据 `set_args`

名称	类型	默认值	说明
`view_img`	`bool`	`False`	用热图显示框架
`colormap`	`cv2.COLORMAP`	`None`	cv2.COLORMAP 用于热图
`imw`	`int`	`None`	热图宽度
`imh`	`int`	`None`	热图高度
`line_thickness`	`int`	`2`	增加边界框并计算文本厚度
`view_in_counts`	`bool`	`True`	仅在视频帧上显示内计数
`view_out_counts`	`bool`	`True`	仅在视频帧上显示外数
`classes_names`	`dict`	`model.model.names`	类名词典
`heatmap_alpha`	`float`	`0.5`	热图阿尔法值
`count_reg_pts`	`list`	`None`	物体计数区域点
`count_txt_color`	`RGB Color`	`(0, 0, 0)`	对象计数文本的前景色
`count_reg_color`	`RGB Color`	`(255, 0, 255)`	计数区域颜色
`region_thickness`	`int`	`5`	计数区厚度值
`decay_factor`	`float`	`0.99`	特定时间后删除热图区域的衰减系数
`shape`	`str`	`circle`	支持 "矩形 "或 "圆形 "热图显示形状
`line_dist_thresh`	`int`	`15`	直线计数器的欧氏距离阈值
`count_bg_color`	`RGB Color`	`(255, 255, 255)`	计数荧光笔颜色
`cls_txtdisplay_gap`	`int`	`50`	显示每个班级计数之间的差距

论据 `model.track`

名称	类型	默认值	说明
`source`	`im0`	`None`	图像或视频的源目录
`persist`	`bool`	`False`	帧与帧之间的持久轨迹
`tracker`	`str`	`botsort.yaml`	跟踪方法 "bytetrack "或 "botsort
`conf`	`float`	`0.3`	置信度阈值
`iou`	`float`	`0.5`	借据阈值
`classes`	`list`	`None`	按类别筛选结果，即 classes=0，或 classes=[0,2,3]

热图 COLORMAPs

彩色地图名称	说明
`cv::COLORMAP_AUTUMN`	秋色地图
`cv::COLORMAP_BONE`	骨骼颜色图
`cv::COLORMAP_JET`	喷气彩图
`cv::COLORMAP_WINTER`	冬季彩色地图
`cv::COLORMAP_RAINBOW`	彩虹颜色地图
`cv::COLORMAP_OCEAN`	海洋颜色地图
`cv::COLORMAP_SUMMER`	夏季彩色地图
`cv::COLORMAP_SPRING`	春色地图
`cv::COLORMAP_COOL`	酷炫的彩色地图
`cv::COLORMAP_HSV`	HSV（色调、饱和度、值）色彩图
`cv::COLORMAP_PINK`	粉色地图
`cv::COLORMAP_HOT`	热门彩色地图
`cv::COLORMAP_PARULA`	帕鲁拉彩图
`cv::COLORMAP_MAGMA`	岩浆颜色图
`cv::COLORMAP_INFERNO`	地狱彩色地图
`cv::COLORMAP_PLASMA`	等离子彩图
`cv::COLORMAP_VIRIDIS`	Viridis 彩色地图
`cv::COLORMAP_CIVIDIS`	Cividis 彩色地图
`cv::COLORMAP_TWILIGHT`	黄昏彩色地图
`cv::COLORMAP_TWILIGHT_SHIFTED`	偏移的曙光色彩图
`cv::COLORMAP_TURBO`	涡轮颜色图
`cv::COLORMAP_DEEPGREEN`	深绿色彩图

这些颜色图通常用于用不同的颜色表示可视化数据。

创建于 2023-12-07，更新于 2024-04-05
作者：RizwanMunawar(8)、glenn-jocher(7)、AyushExel(1)、1579093407@qq.com(1)