Informatics Mathematical Perception Modeling of
Person Re-Identification in Multi Camera Surveillance Networks
Using Cognitive Sciences
Due to the change in illumination, expression, view angle, appearances captured in surveillance camera, the Person Re-identification Re-ID demand is increasing day by day. In order to identify and then re-identify a person again and again in different non overlapping camera views causes hindrance due to the abrupt change in appearance of any individual. For that sake, possible viewable unique descriptors i.e. gait, face, clothing color, height, carrying conditions are calculated in order to re-identify a person. Extracting a reliable descriptor is dependent upon availability of good quality observations. Most of the researchers resolve the same problem by focusing on single descriptor. The selection of any single or the combination of various biometrics drops out the accuracy of the system and hence the overall recognition rate. The accuracy of the result is affected because let if anyone use the face as a descriptor but the face is not visible then the system does not performs well. In this research, we will focus on all descriptors and present a cognitive model that automatically selects the biometric according to the current input available. Finally, we generate our own mean descriptor that works on any type of input whether any descriptor is visible or not. At the end, a comparison is made with existing methods.
BUDGET only shown to selected people
5 freelancers are bidding on average $26 for this job
We can redo, improve, and publish any engineering, scientific MATLAB, SIMULINK, PSPICE, NS2, NS3 based IEEE papers. Stay tuned, I'm still working on this proposal.