期刊名称：International Journal of Advanced Research in Computer Engineering & Technology (IJARCET)
印刷版ISSN：2278-1323
出版年度：2015
卷号：4
期号：6
页码：2625-2628
出版社：Shri Pannalal Research Institute of Technolgy
摘要：we present a robust Face Recognition (FR) algorithm based on scale invariant feature transform. As it is based on SIFT, this algorithm is also invariant to translation, rotation, scale and illumination parameters. It is also robust to noise in the image for FR. We find local descriptors from image by using SIFT, but as there are number of local descriptors in every image, we select only highly stable local descriptors. In order to reduce the computational complexity and dimension of the feature vector which contain information about the local descriptors, a detailed study of face recognition using SIFT has been done. As we are selecting only highly stable local descriptors, every local descriptor is matched accurately with high probability against large database of local descriptors from many images. The algorithm based on SIFT show the results which are very accurate and very good at discrimination of the features of different images and provide efficient recognition of the face. It is robust against the challenges like illumination, pose variation and expression change, etc. It gives recognition rate more than 92%.
