使用Ultralytics YOLOv8 计数物体 🚀

什么是对象计数？

使用 Ultralytics YOLOv8YOLOv8 在实时应用中表现出色，凭借其先进的算法和深度学习能力，可为人群分析和监控等各种场景提供高效、精确的物体计数。

观看： 使用对象计数Ultralytics YOLOv8

物体计数的优势？

资源优化：对象计数通过提供准确的计数和优化库存管理等应用中的资源分配，促进了高效的资源管理。
增强安全性：物体计数可准确跟踪和计数实体，有助于主动探测威胁，从而增强安全性和监控能力。
知情决策：物体计数为决策、优化零售、交通管理和其他各种领域的流程提供了宝贵的见解。

真实世界的应用

物流	水产养殖

使用传送带包装计数Ultralytics YOLOv8	利用海洋数鱼Ultralytics YOLOv8

使用YOLOv8 进行对象计数示例

地区计数多边形计数排队计数特定等级

from ultralytics import YOLO
from ultralytics.solutions import object_counter
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))

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

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

# Init Object Counter
counter = object_counter.ObjectCounter()
counter.set_args(view_img=True,
                 reg_pts=region_points,
                 classes_names=model.names,
                 draw_tracks=True,
                 line_thickness=2)

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 = counter.start_counting(im0, tracks)
    video_writer.write(im0)

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

from ultralytics import YOLO
from ultralytics.solutions import object_counter
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))

# Define region points as a polygon with 5 points
region_points = [(20, 400), (1080, 404), (1080, 360), (20, 360), (20, 400)]

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

# Init Object Counter
counter = object_counter.ObjectCounter()
counter.set_args(view_img=True,
                 reg_pts=region_points,
                 classes_names=model.names,
                 draw_tracks=True,
                 line_thickness=2)

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 = counter.start_counting(im0, tracks)
    video_writer.write(im0)

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

from ultralytics import YOLO
from ultralytics.solutions import object_counter
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))

# Define line points
line_points = [(20, 400), (1080, 400)]

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

# Init Object Counter
counter = object_counter.ObjectCounter()
counter.set_args(view_img=True,
                 reg_pts=line_points,
                 classes_names=model.names,
                 draw_tracks=True,
                 line_thickness=2)

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 = counter.start_counting(im0, tracks)
    video_writer.write(im0)

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

from ultralytics import YOLO
from ultralytics.solutions import object_counter
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))

line_points = [(20, 400), (1080, 400)]  # line or region points
classes_to_count = [0, 2]  # person and car classes for count

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

# Init Object Counter
counter = object_counter.ObjectCounter()
counter.set_args(view_img=True,
                 reg_pts=line_points,
                 classes_names=model.names,
                 draw_tracks=True,
                 line_thickness=2)

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_to_count)

    im0 = counter.start_counting(im0, tracks)
    video_writer.write(im0)

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

区域可移动

点击区域边缘，可将其移动到框架内的任何位置

可选参数 `set_args`

名称	类型	默认值	说明
`view_img`	`bool`	`False`	显示带有计数的帧
`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
`line_thickness`	`int`	`2`	增加边界框并计算文本厚度
`reg_pts`	`list`	`[(20, 400), (1260, 400)]`	定义区域面积的点
`classes_names`	`dict`	`model.model.names`	类名词典
`count_reg_color`	`RGB Color`	`(255, 0, 255)`	计数区域或线条对象的颜色
`track_thickness`	`int`	`2`	跟踪线的厚度
`draw_tracks`	`bool`	`False`	启用绘制轨迹线
`track_color`	`RGB Color`	`(0, 255, 0)`	每条轨道线的颜色
`line_dist_thresh`	`int`	`15`	直线计数器的欧氏距离阈值
`count_txt_color`	`RGB Color`	`(255, 255, 255)`	对象计数文本的前景色
`region_thickness`	`int`	`5`	对象计数器区域或线条的厚度
`count_bg_color`	`RGB Color`	`(255, 255, 255)`	计数荧光笔颜色

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]
`verbose`	`bool`	`True`	显示物体跟踪结果

创建于 2023-12-02，更新于 2024-04-18
作者：glenn-jocher(9)、RizwanMunawar(1)、AyushExel(1)