基于整形小波变换的零树编码算法的改进

Improvement of zero-tree coding algorithm based on integer wavelet transform

针对整形小波变换能量集中性较差的问题,提出了基于整数平方阈值的小波图像压缩算法。该算法用整数平方阈值代替传统零树编码中2的整数幂量化阈值;通过调整码字的表示方式,充分利用游程编码的优点,在提高压缩效率的同时,简化了编码与解码的过程;使零树编码在整形小波变换中得到高效应用。

A novel wavelet image compression algorithm based on integer square threshold is presented for integer wavelet transforms having worse energy concentration. The integer power of two of general zerotree coding is replaced by integer square threshold. The algorithm makes full use of RLE coding by adjusting form of code. It not only improves compression efficiency, but also simplifies coding and encoding processing. Zerotree coding is applied in integer wavelet transforms efficiently.