Using SIFT for object recognition

asked 2015-02-04 16:39:48 -0500

I have a group of pictures and I'm extracting the descriptor of each picture. After that I want to concatenate these descriptors to create a model that describes this category and save it as a picture. How can I do that?

What I have now is:

img = cv2.imread(unicode(file))
sift = cv2.SIFT()
keypoint, descriptor = sift.detectAndCompute(img, None)

How can I concatenate the descriptors and generate a picture of the result?


edit retag flag offensive close merge delete