Decision fusion of multiple features from small data set

This is not a much specific question, I just want to gather idea from colleagues in this forum.

My question is;

I estimated human height, size and cloth color from video data collected from one camera, let's call these features and I have similar set of people as another data set but collected from another camera. So I want to identify the people in the second dataset if they match the features in the first dataset (something like re-identification) but, I am confused on how to fuse the features together and how to match or rank them(which is like classification).

Please I need suggestions and advise.