查閱資料得在經過sift匹配之后，特征點類如下所示

            
              class KeyPoint

{ ? ? ? ? ? ?Point2f ?pt; ?//坐標

? ? ? ? ? ?? float ?size; //特征點鄰域直徑

? ? ? ? ? ?? float ?angle; //特征點的方向，值為[零,三百六十)，負值表示不使用

? ? ? ? ? ?? float ?response;

? ? ? ? ? ?? int ?octave; //特征點所在的圖像金字塔的組

? ? ? ? ? ?? int ?class_id; //用于聚類的id

}

            
          

opencv-python中也一樣

            
              #coding=utf-8
'''
@project : binocular_vision
@author  : Hoodie_Willi
#@description: $輸出sift匹配后，匹配點的坐標
#@time   : 2019-05-28 10:25:36
'''
import numpy as np
import cv2
sift = cv2.xfeatures2d.SIFT_create()
img1 = cv2.imread("./img/l/left.jpg", cv2.COLOR_BGR2GRAY)
img2 = cv2.imread("./img/r/right.jpg", cv2.COLOR_BGR2GRAY)
kp1, des1 = sift.detectAndCompute(img1, None)
kp2, des2 = sift.detectAndCompute(img2, None)
bf =cv2.BFMatcher()
matches = bf.knnMatch(des1, des2, k=2)

# ## Create flann matcher
# FLANN_INDEX_KDTREE = 1 # bug: flann enums are missing
# flann_params = dict(algorithm = FLANN_INDEX_KDTREE, trees = 5)
# #matcher = cv2.FlannBasedMatcher_create()
# matcher = cv2.FlannBasedMatcher(flann_params, {})

## Ratio test
print(len(matches))
matchesMask = [[0, 0] for i in range(len(matches))]
for i, (m1, m2) in enumerate(matches):
    if m1.distance < 0.7 * m2.distance:# 兩個特征向量之間的歐氏距離，越小表明匹配度越高。
        matchesMask[i] = [1, 0]
        pt1 = kp1[m1.queryIdx].pt  # trainIdx    是匹配之后所對應關鍵點的序號，第一個載入圖片的匹配關鍵點序號
        pt2 = kp2[m1.trainIdx].pt  # queryIdx  是匹配之后所對應關鍵點的序號，第二個載入圖片的匹配關鍵點序號
        # print(kpts1)
        print(i, pt1, pt2)
        if i % 5 ==0:
            cv2.circle(img1, (int(pt1[0]),int(pt1[1])), 5, (255,0,255), -1)
            cv2.circle(img2, (int(pt2[0]),int(pt2[1])), 5, (255,0,255), -1)

# 匹配點為藍點, 壞點為紅點
draw_params = dict(matchColor = (255, 0,0),
        singlePointColor = (0,0,255),
        matchesMask = matchesMask,
        flags = 0)

res = cv2.drawMatchesKnn(img1, kp1, img2, kp2, matches, None, **draw_params)
cv2.imshow("Result", res)
cv2.waitKey(0)
cv2.destroyAllWindows()
            
          

匹配結果如圖：

坐標如圖所示