Novel embedded image compression technique using second generation bandelets and zerotree coding

摘要

As a new multiscale geometric analysis tool, second generation bandelets can make full use of intrinsic geometric regularity and provide an optimal representation of the underlying high-dimensional functions adaptively. In this paper, a novel embedded image compression technique is presented, which is based on second generation bandelets and zerotree coding. The method is embedded progressive which can control the actual compression ratio and allow "best" reconstructions during decompression. Objectively, we use peak signal-to-noise ratio and edge preservation properties to measure Image quality. Numerical tests applied on natural images show that compared with other two algorithms, our method provides improvements both in peak signal-to-noise ratio and edge preservation properties. Subjectively, our method avoids visible ringing effects having better image visual fidelity.