2

I am trying to extract keypoints from an image. The image is a frame from a video I captured myself.

enter image description here I already extracted the keypoints using SURF.

SURF

I found cv2.FileStorage and cv2.FileNodein the help but I don't know how to use them. 

FileNode(...)
        FileNode() -> <FileNode object>

    FileStorage(...)
        FileStorage([source, flags[, encoding]]) -> <FileStorage object>

Every time I try to dump the keypoints to a file using cPickle, json or numpy

I get an error from these libraries for not being able to save a cv2.FileStorage file type.

I found the official documentation of opencv for C++ on how to save these type of files, but I haven't found one for Python yet.

What I want to do is to dump these keypoints into an XML file for later use for object detection and recognition.

Tes3awy
  • 2,166
  • 5
  • 29
  • 51
  • Btw, SURF are more useful if computed on a grayscale image ;D – Miki Jul 30 '16 at 18:07
  • @Miki I need to extract the keypoints for BOVW so I don't need the extra keypoints observed in the background. That's why I am applying it on B&W image. – Tes3awy Jul 30 '16 at 18:09

0 Answers0