I would like to implement the same kind of feature extraction process in this paper. That is, I'd like the extracted features to be related to color and texture attributes of the image (and shape perhaps).
Region of interest of the image has already been segmented. However, I cannot decide which feature descriptor (extractor) in OpenCV is appropriate. (i.e., which descriptor gives color and texture features).
Questions:
- Are there feature descriptors (SIFT, SURF, ORB, FREAK, BRIEF etc.) that represent/detect for the color and texture features?
Or should it be done/coded/computed manually? - As for texture features, it is said to be achieved using gray level co-occurrence matrix (GLCM) but OpenCV no longer supports this class (deprecated). Are there other alternatives OpenCV offers regarding this?
I am stuck in looking for feature descriptor to use for classification. Examples and code implementation, pseudocode or good references (for beginners) would be helpful.
