NFGS Enabled Face Re-identification for Efficient Surveillance in Low Quality Video

People tracking, and re-identification of human faces have been widely applied in large areas namely, surveillance and human-robot interaction. The accuracy of the face reidentification largely depends on the quality of the image likewise the face recognition systems. In this piece of work, we have designed an algorithm to extract the important facial features from a low- face image by utilizing the method of figure-ground segregation inspired from neuro-visual structure in human visual cortex (NFGS). The aforementioned method describes the role of glial cells and formation of Extended classical receptive field (ECRF) further. This work illustrates improved face reidentification, tracking and recognition in low quality images including low resolution or small images, low contrast images, non-uniformly illuminated images and blurred images affected through motion. This in turn helps to perform re-identification and person tracking in offline videos. Face re-identification and face matching has been addressed in low quality image space through the proposed framework effectively as supported by the experimental evidences.