摘要： 提出一种基于自适应方向提升(ADL)小波变换的图像压缩算法。根据灰度共生矩阵角二阶矩的差异，将图像分割成平坦性不同的分块。对纹理信息较少的块，采用一般提升小波变换以减少变换时间。对纹理信息较多的块，采用方向提升小波以提高变换效果。结合多级树集合分裂编码和算术编码对变换系数和方向信息分别进行编码。实验结果表明，与ADL算法相比，该算法能有效减少方向小波变换时间。

关键词: 图像压缩, 自适应方向提升, 灰度共生矩阵, 角二阶矩, 多级树集合分裂

Abstract: Based on the general Adaptive Directional Lifting(ADL) promotion transformation, this paper proposes an image compression algorithm based on ADL wavelet transform. Using this algorithm, the image is divided into flat and non flat block according to different angle second moment of Gray Level Co-occurrence Matrix(GLCM). For each block, the lifting scheme is adaptively selected to reduce the computation of directional lifting wavelet. For the flat blocks, it uses ordinary horizontal and the vertical promotion directly to reduce the time. For the non flat blocks, it uses the directional lifting wavelet to enhance the result. Experimental results show that this algorithm can dramatically reduce the computational time compared with the tradditional ADL method.

Key words: image compression, Adaptive Directional Lifting(ADL), Gray Level Co-occurrence Matrix(GLCM), angle second moment, Set Partition in Hierarchical Trees(SPIHT)

中图分类号: 

• TP911.73