Computer Vision: SURF (Speeded Up Robust Features) with Color consideration

Question

Is it possible to enhance the Speeded Up Robust Features (SURF) Computer Vision algorithm so that it can differentiate equally distributed and shaped objects with different colors?

I am looking for research papers or any other sources that try to enhance SURF so that it can differentiate the same objects with different colors.

Answer 1

Color-SURF: A surf descriptor with local kernel color histograms

Answer 2

Rather than add color to the algorithm it's often easier to convert the image and then run the algorithm on each single color scaler image.

It's worth looking at doing this in another color space rather than RGB - depending on what you are trying to find, doing it in HSV or YUV (at least the UV) parts might make sense.

Answer 3

I think the addition of color information to the SURF may enhance the performance a little, but not significantly