Designing a new storage technique for face images transmission using JPEG2000

Pictures and video sequences showing human faces are of high importance in content-based retrieval systems. During flourishing Internet time, more and more human face image databases are stored on the Internet. They provide a great number of human face images to access. Therefore, considering the storage space and the network bandwidth, how to access human face image more efficiently has become an important topic. The primary goal of our method is to minimize the number of bits required to preserve the crucial facial features without any loss of information. In this paper, we propose a new technique based on JPEG2000 for the storage and retrieval of the human face images. We combine lossy and lossless image compression of JPEG2000 to reach our goal. Aiming at the feature of human face image data, we extract the facial features in the image as the most important information. The lossless image compression is used to retain the facial features of human face image. The background and the rest of the image are processed by lossy compression to further save the required storage space. The preliminary result has shown that our method can have higher compression ratio without loss of the crucial facial feature for low bandwidth communication applications, such as videophone and teleconferencing.