Codec designs for image browsing

We propose codec designs for image browsing. Image/video browsing is an important requirement in many multimedia applications. In image browsing, a lower resolution version of the image is first viewed. Upon request, the resolution is further enhanced until an acceptable fidelity is obtained. Recently, many new codecs have been designed for image/video compression. However these codecs are not directly applicable for image browsing as they cannot interact with the user. In this paper, we propose a codec design based on Gabor decomposition which is well adapted for image browsing applications. However we note that the generic design can be implemented with any compression algorithm. Gabor decomposition is a promising approach for image browsing. The Gabor basis functions match the human visual profiles and hence result in better subjective quality compared to JPEG. For example, Gabor decomposition does not suffer from the blocking artifacts prevalent in JPEG-DCT. The proposed codec can achieve real-time performance, is simple and modular and hence can be implemented in VLSI