Python OpenCV：如何在不间断视频流中选取ROI进行目标跟踪

我正在学习OpenCV目标跟踪教程，该教程通过视频流的静态截图创建ROI来实现基础目标跟踪（非检测）。我需要在摄像头/视频流播放过程中（不暂停、不打开新窗口）选取Bounding Box，但标准的cv2.selectROI()会打开新窗口并暂停视频流，不符合需求。

参考相关内容后，我实现了鼠标事件来在实时视频上绘制矩形，但将自定义BBox坐标传入OpenCV跟踪器后出现异常。尝试过的旧示例到2022年已过时，现在能触发跟踪器，但初始BBox并未出现在鼠标选取的位置。

我的设置规则：两次左键点击创建BBox，一次右键点击移除BBox并停止跟踪，相关代码如下：

import cv2, sys

(major_ver, minor_ver, subminor_ver) = (cv2.__version__).split('.')
print(cv2.__version__)
video = cv2.VideoCapture(r'C:\Users\longc\Downloads\race_test.mp4')
cv2.namedWindow('Frame', cv2.WINDOW_NORMAL)
import selectinwindow

# Our ROI, defined by two points
p1, p2 = None, None
state = 0
bbox = (0,50,100,150)
# Set up tracker.
# Instead of MIL, you can also use
cv2.namedWindow("Frame", cv2.WINDOW_NORMAL)
tracker_types = ['BOOSTING', 'MIL','KCF', 'TLD', 'MEDIANFLOW', 'CSRT', 'MOSSE']
tracker_type = tracker_types[6]

if int(major_ver) < 4 and int(minor_ver) < 3:
    tracker = cv2.cv2.Tracker_create(tracker_type)
else:
    if tracker_type == 'BOOSTING':
        tracker = cv2.TrackerBoosting_create()
    if tracker_type == 'MIL':
        tracker = cv2.TrackerMIL_create()
    if tracker_type == 'KCF':
        tracker = cv2.TrackerKCF_create()
    if tracker_type == 'TLD':
        tracker = cv2.TrackerTLD_create()
    if tracker_type == 'MEDIANFLOW':
        tracker = cv2.TrackerMedianFlow_create()
    if tracker_type == 'CSRT':
        tracker = cv2.TrackerCSRT_create()
    if tracker_type == 'MOSSE':
        tracker = cv2.legacy.TrackerMOSSE_create()


# Called every time a mouse event happen
def on_mouse(event, x, y, flags, userdata):
    global state, p1, p2, bbox, val, tracker, frame

    # Left click
    if event == cv2.EVENT_LBUTTONUP:
        # Select first point
        if state == 0:
            p1 = (x, y)
            state += 1
        # Select second point
        elif state == 1:
            p2 = (x, y)
            state += 1
            bbox = p1 + p2
            print(p1,p2)
            print(bbox)
            # Initialize tracker with first frame and bounding box
            if int(major_ver) < 4 and int(minor_ver) < 3:
                tracker = cv2.cv2.Tracker_create(tracker_type)
            else:
                if tracker_type == 'BOOSTING':
                    tracker = cv2.TrackerBoosting_create()
                if tracker_type == 'MIL':
                    tracker = cv2.TrackerMIL_create()
                if tracker_type == 'KCF':
                    tracker = cv2.TrackerKCF_create()
                if tracker_type == 'TLD':
                    tracker = cv2.TrackerTLD_create()
                if tracker_type == 'MEDIANFLOW':
                    tracker = cv2.TrackerMedianFlow_create()
                if tracker_type == 'CSRT':
                    tracker = cv2.TrackerCSRT_create()
                if tracker_type == 'MOSSE':
                    tracker = cv2.legacy.TrackerMOSSE_create()
            val = tracker.init(frame, bbox)
    # Right click (erase current ROI)
    if event == cv2.EVENT_RBUTTONUP:
        p1, p2 = None, None
        state = 0
        val = tracker.init(frame, bbox)

# Register the mouse callback
cv2.setMouseCallback('Frame', on_mouse)


test_counter = 0
while video.isOpened():
    val, frame = video.read()
    if not val:
        break
    # If a ROI is selected, draw it
    if state > 1:
        #cv2.rectangle(frame, p1, p2, (255, 0, 0), 1)
        # Read a new frame
        val, frame = video.read()
        if not val:
            break

        # Start timer
        timer = cv2.getTickCount()

        # Update tracker
        val, bbox = tracker.update(frame)

        # Calculate Frames per second (FPS)
        fps = cv2.getTickFrequency() / (cv2.getTickCount() - timer);

        # Draw bounding box
        if val:
            # Tracking success
            p1 = (int(bbox[0]), int(bbox[1]))
            p2 = (int(bbox[0] + bbox[2]), int(bbox[1] + bbox[3]))
            cv2.rectangle(frame, p1, p2, (255, 0, 0), 2, 1)
            cv2.putText(frame, "Tracking Successful", (100, 80), cv2.FONT_HERSHEY_SIMPLEX, 0.75, (50, 170, 50), 2)
        else:
            # Tracking failure
            cv2.putText(frame, "Tracking failure detected", (100, 80), cv2.FONT_HERSHEY_SIMPLEX, 0.75, (0, 0, 255), 2)

        # Display tracker type on frame
        cv2.putText(frame, tracker_type + " Tracker", (100, 20), cv2.FONT_HERSHEY_SIMPLEX, 0.75, (50, 170, 50), 2);

        # Display FPS on frame
        cv2.putText(frame, "FPS : " + str(int(fps)), (100, 50), cv2.FONT_HERSHEY_SIMPLEX, 0.75, (255, 127, 0), 2);

    # Show image
    cv2.imshow('Frame', frame)

    # Let OpenCV manage window events
    key = cv2.waitKey(50)
    # If ESCAPE key pressed, stop
    if key == 27:
        video.release()

    # Delayed test console
    test_counter += 1
    if test_counter % 30 == 0:
        print("Current BBox: {}".format(str(bbox)))
        print(val)
        #print(frame)
        print(tracker)

内容的提问来源于stack exchange，提问作者ansib1e