I really don't know what tag this question should be under, however I'm developing with C#, so I chose C# as its tag, could any mods here please edit my question to make a suitable tag. (for some reason, I can't see all supported tags), thank you very much. Now is my question's main idea:
I have tried running some projects involving edge detection (one of image feature detections), the input is some image and the output is another image of edge features. All are clear, but I wonder how I can save these feature as some kind of mathematic stuff like vectors, matrices... to end the process called 'feature extraction'? You know that these stored features will be used then for next processes such as recognition, matching....
I have searched all the web pages but they don't explain clearly about storing features, all the algorithms they have seem to DETECT the features not EXTRACT the features (of course, detection would be the necessary stage before jumping to extraction) and see detection as extraction. My capitalized words don't mean my loud shout, I only want to emphasize them, so please pardon me if you mind. Could you please give me some example on some image features, I want to know exactly what stuff (figures, vectors...) of the features should be stored, my interested features are color features, edge features, corner features, ...
Please help, if you don't know some of the answer but you know someone who is experienced in this field, please notify him/her about my question or give me his/her e-mail so that I can ask for help privately. I really need a sufficient answer to this question of mine.
Thank you very much!