在Opencv中使用multiTracker对象实现多目标跟踪初始化

Question

我在cv2中使用multiTracker来跟踪多个对象。我的代码是基于this link和this one构建的。我想在视频中的任何时候一次初始化所有的边界框。然而，我很难做到这一点。代码如下：

import imutils
import cv2
from random import randint

trackerName = 'csrt'
videoPath = "C:/Users/Art/testVideo.mp4"

OPENCV_OBJECT_TRACKERS = {
    "csrt": cv2.TrackerCSRT_create,
    "kcf": cv2.TrackerKCF_create,
    "boosting": cv2.TrackerBoosting_create,
    "mil": cv2.TrackerMIL_create,
    "tld": cv2.TrackerTLD_create,
    "medianflow": cv2.TrackerMedianFlow_create,
    "mosse": cv2.TrackerMOSSE_create
}


trackers = cv2.MultiTracker_create()
cap = cv2.VideoCapture(videoPath)

while cap.isOpened():

    ret, frame = cap.read()

    if frame is None:
        break

    # for fast processing resize the frame 
    frame = imutils.resize(frame, width=600)

    (success, boxes) = trackers.update(frame)
    for box in boxes:
        (x, y, w, h) = [int(v) for v in box]
        cv2.rectangle(frame, (x, y), (x + w, y + h), (0, 255, 0), 2)

    cv2.imshow("Frame", frame)
    key = cv2.waitKey(1) & 0xFF

    # if the 'p' key is selected, we select a bounding box to track
    if key == ord("p"):
        boxes = []
        colors = []
        boxFlag = True
        while boxFlag:
            box = cv2.selectROI('MultiTracker', frame, fromCenter=False, 
                                showCrosshair=True)
            boxes.append(box)
            colors.append((randint(64, 255), randint(64, 255), randint(64, 
                          255)))
            print("Press q to quit selecting boxes and start tracking")
            print("Press any other key to select next object")
            if key == ord("q"):  # q is pressed
                boxFlag = False
    # Initialize MultiTracker
    for box in boxes:
        tracker = OPENCV_OBJECT_TRACKERS[trackerName]()
        trackers.add(tracker, frame, box)

cap.release()
cv2.destroyAllWindows()

然而，也存在一些问题。首先，当我点击“p”键选择边界框时，视频暂停，另一个窗口打开，显示视频暂停的帧，我只能在新窗口上选择边界框。此外，当我按下'q‘键时，什么也不会发生，基本上它永远都在那里。我的问题是如何解决这个问题，并能够初始化跟踪后，我选择了所有的边界框。

Answer 1

我发现了这一点，并认为这可能对其他人有用。我对前面的代码做了一些修改。我还意识到cv2有selectROIs方法，如果有人想要一次选择多个边界框，这个方法会很有用。以下是更新后的代码：

import imutils
import cv2
from random import randint

trackerName = 'csrt'
videoPath = "C:/Users/Art/testVideo.mp4"

OPENCV_OBJECT_TRACKERS = {
    "csrt": cv2.TrackerCSRT_create,
    "kcf": cv2.TrackerKCF_create,
    "boosting": cv2.TrackerBoosting_create,
    "mil": cv2.TrackerMIL_create,
    "tld": cv2.TrackerTLD_create,
    "medianflow": cv2.TrackerMedianFlow_create,
    "mosse": cv2.TrackerMOSSE_create
}

# initialize OpenCV's special multi-object tracker
trackers = cv2.MultiTracker_create()
cap = cv2.VideoCapture(videoPath)

while cap.isOpened():

    ret, frame = cap.read()

    if frame is None:
        break

    frame = imutils.resize(frame, width=600)
    (success, boxes) = trackers.update(frame)

    # loop over the bounding boxes and draw them on the frame
    for box in boxes:
        (x, y, w, h) = [int(v) for v in box]
        cv2.rectangle(frame, (x, y), (x + w, y + h), (0, 255, 0), 2)

    cv2.imshow("Frame", frame)
    key = cv2.waitKey(1) & 0xFF

    # if the 's' key is selected, we are going to "select" a bounding
    # box to track
    if key == ord("s"):
        colors = []
        # select the bounding box of the object we want to track (make
        # sure you press ENTER or SPACE after selecting the ROI)
        box = cv2.selectROIs("Frame", frame, fromCenter=False,
                             showCrosshair=True)
        box = tuple(map(tuple, box)) 
        for bb in box:
            tracker = OPENCV_OBJECT_TRACKERS[trackerName]()
            trackers.add(tracker, frame, bb)

    # if you want to reset bounding box, select the 'r' key 
    elif key == ord("r"):
        trackers.clear()
        trackers = cv2.MultiTracker_create()

        box = cv2.selectROIs("Frame", frame, fromCenter=False,
                            showCrosshair=True)
        box = tuple(map(tuple, box))
        for bb in box:
            tracker = OPENCV_OBJECT_TRACKERS[trackerName]()
            trackers.add(tracker, frame, bb)

    elif key == ord("q"):
        break
cap.release()
cv2.destroyAllWindows()

在选择了每个边界框之后，您需要单击"Enter“或"Space”按钮来完成它，并开始选择一个新的边界框。一旦你完成了边界框选择，按"Esc“键结束ROI选择并开始追踪。此外，如果出于任何原因需要重置边界框，请按"r“按钮。

Answer 2

也许您应该在没有循环的"while cap.isOpened():“循环之前尝试"cv2.selectROIs”函数，因为它看起来像是"cv2.selectROIs“函数在选择ROI时作为循环本身工作。