I have to create color histograms of the GTSRB dataset. Since these images were created under different lighting and weather conditions, I consider whether or how I should pre-process the images before I create the color histograms.
With the help of these color histograms, a classifier should then be trained, which can recognize new traffic sign images.
Which preprocessing would you recommend? Histogram equalization of the RGB images, histogram equalization in the HSV color space etc …?
I look forward to your opinions and thank you in advance.