Histogram$-$based Skin Color Detection

Exercise for a clean implementation of color based segmentation with application to flesh tone $($skin color$)$ detection. Due to its efficiency in computation and implementation, such a technique is very likely to be used in our final project. What you need to do in this MP are the following steps:

$*$ Collecting flesh tone training data. A couple of images are provided. It is very easy to find on the Web some color images contain skin color $($people$).$ You can crawl such images from the Web. Then, you should design a tool to collect skin pixels. Of course, a simple interface will do$.$ E$.$g$.,$ you can use imcrop or ginput in Matlab. If you are using C$++,$ you have to select pixels or small regions using a mouse.

$*$ Selecting a good color space. You$$d better try a couple of color spaces, e$.$g$.,$ RGB$,$ N$-$RGB$,$ HSI, etc, before you make your decision.

$*$ Training a color histogram$-$based flesh tone detector. Basically, you just construct a $2$D color histogram based on the color pixels you have collected, i$.$e$.,$ R$-$G$,$ NR$-$NG$,$ H$-$S$,$ etc. Note: pay more attention to normalization.

$*$ Finding skin regions in test images. You apply your color detector to segment skin color regions in test images.

Some training and testing images are shown here$1.$ You can also create your own test images. Please pay attention that this image is an $24-$bit bitmap images. You should test your color detector on all these three images.