如何在python中访问SURF的KeyPoints的特征尺度和主导方向？

Question

I'm reading a paper titled 'Local Geometric Consistency Constraint for Image Retrieval'. The author introduces a method to make the matched SFIT points are right using local geometric consistency. This method requires points' characteristic scales and dominant orientation.

In python, I use opencv to extract SURF keypoints. How could I get points' characteristic scales and dominant orientation?

Answer 1

Creat SURF feature extractor with

surf = cv2.xfeatures2d.SURF_create()
kpu, desu = surf.detectAndCompute(image, None)

keypoints are stored in kpu, now you can do something like:

for i in range(len(kpu)):
     point = kpu[i].pt
     print("your point here", point)