Python 2.7 OpenCV：从图像坐标中提取特征/描述符

Question

I extracted several points (x,y-coordinates) of interest from my image using OpenCV in Python (ie Thresholding, contourfinding, moments). Now, I want to match those points with a reference pattern to exclude falsy detected points. My approach is to match the features/descriptors from the image points with the features/descriptors of the reference pattern, but I don't know how to extract features/descriptors from the image points.

How do I extract features/descriptors from the given coordinates using OpenCV in Python?

Answer 1

Maybe you should use KeyPoint

import cv2
surf = cv2.SURF() #Should works with another descriptors
customKeyPoints = [cv2.KeyPoint(<your coords, size, angle, etc>), 
                   cv2.KeyPoint(<your coords, size, angle, etc>), ...]
surf.compute(img, customKeyPoints)
#Should return you descriptors from specified points

Similiar question with answer

PS Sorry for my English