Hi,
A computer system engineer by profession, i have 10+ yr experience working in image processing field. The problem u specify, is first the face recognition and then the generation of UID. The UID generation is a seperate process, the 3D face recognition without surface reconstruction contain novel methods, which can be optimized for real time implementation. The UID generation is a mathematical process and would largely depends on the population of the database.
Suppose our database is empty and a new person comes in front of shop. the program will catch its image,compute its features, fill in database and generate a UID for that person. The next time same person comes, it will again compute the features and will compare with database for best feature match, this is a very robust process and implemented in real time and same UID will be generated for that person.
I will implemenet it in C++, using visual studio 2010, with OpenCV and some other dependant libraries. the time line i specify is very much needed to build this full fledge application.
Looking forward to your response.
Thanks n regards
DP