如何在强力算法中获得匹配关键点的位置和大小

Question

我想检测两个连续视频帧（当前帧和前一帧）的特征点。为此，我使用SIFT和强力算法提取匹配的关键点。现在我只想将匹配的关键点转换为当前帧和他们的位置和大小。我怎样才能做到这一点 ？有没有具体的功能呢？非常感谢您的帮助

下面是我目前的代码：

import cv2
import numpy as np

cap = cv2.VideoCapture('video3.mov')
ret, frame = cap.read()

while(cap.isOpened()):

  prev_frame=frame[:]
  ret, frame = cap.read()

  if ret:

      graycurrent = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
      grayprev = cv2.cvtColor(prev_frame, cv2.COLOR_BGR2GRAY)

      #detect key feature points
      sift = cv2.xfeatures2d.SIFT_create()

      #kp = sift.detect(graycurrent, None)
      kp1, des1 = sift.detectAndCompute(graycurrent, None)

      #draw key points detected
      img1 = cv2.drawKeypoints(graycurrent, kp1,graycurrent, flags=cv2.DRAW_MATCHES_FLAGS_DRAW_RICH_KEYPOINTS)

      kp2, des2 = sift.detectAndCompute(grayprev, None)
      #draw key points detected
      img2 = cv2.drawKeypoints(grayprev, kp2,grayprev, flags=cv2.DRAW_MATCHES_FLAGS_DRAW_RICH_KEYPOINTS)

    # BFMatcher with default params
    bf = cv2.BFMatcher(normType=cv2.NORM_L2, crossCheck= True)
    matches = bf.match(des1,des2)


    cv2.imshow("grayframe",img1)
    cv2.imshow("prevFrame",img2)

  else:
      print("could not read the frame")

  if cv2.waitKey(100) & 0xFF == ord('q'):
        break

cap.release()
cv2.destroyAllWindows()