implementation of emotional behavior of autistic children using SIFT target FPGA

Abstract

Children suffering from autism have difficulties in recognizing the emotions of others which hampers their interpersonal communication. A stable detection system can help such children to easily interact with the external world. In this thesis, an efficient way is proposed which can successfully detect the signs (facial or hand gestures) of such children using Scale Invariant Feature Transform (SIFT) algorithm. Ways of implementing the algorithm on Field Programmable Gate Arrays(FPGA) is also mentioned briefly in this thesis. The algorithm helps in extracting distinctive features from an image using scale invariant features and obtains feature descriptors from the keypoints