In collaboration with my college(Keshav Memorial Institute of Technology, Hyderabad, India), I have started a small project(in C++) that analyses wardrobe of each and every student in the campus. Well, the idea is pretty simple, a camera is put at a gateway where all the students are made to stand in front of it for a few seconds, before they enter their classes for about 30 days. I am given the faces of all the students as a reference data set. The first functionality I thought of finishing at the earliest is counting the number of outfits a particular student has. And then a lot can be done with the obtained data, as in, determining a student's favorite color and types of outfits one prefers to wear the most etc. So far, I have implemented recognition of faces(LBPH Recognizer) only and right now I am planning to create a cascade classifier for an outfit. Although I was suggested to just come up with a static algorithm of detecting an outfit in an input image, given the present situation, as the camera position is static and the distance of the person from the camera is almost constant. So please feel free to suggest me any more ideas over this. Thank you in advance!