摘要
针对应用于图像无损压缩的传统SPIHT算法没有充分利用小波系数低频子带带内的相关性且存在编码冗余的不足之处,提出了基于改进SPIHT的图像无损压缩算法。首先对原始图像进行整数小波变换,然后对小波变换后的低频子带和高频子带分开编码,即对低频子带进行预测编码;对高频子带,当阈值小于等于2时,改变了传统SPIHT算法的编码方式,减少了比特输出。实验结果表明,与传统SPIHT算法相比,比特率平均降低了0.0653bpp。
In view of the problems of traditional SPIHT algorithm applied to lossless image compression, not make full use of the cor- relation in low frequency subband and code redundancy, a new image lossless compression algorithm based on improved SPIHT is proposed. Firstly, the original image is transformed by the integer wavelet transform. Then code the low frequency subband and the high frequency subbands separately that the low frequency subband is coded by predictive coding; for the high frequency subbands, change the coding style of traditional SPIRT when the threshold is less than or equal to 2, reduce the producing bits. The experi- mental results show that bit rate is reduced by 0.0653bpp On average compared with the traditional SPIHT.
出处
《微计算机信息》
2012年第3期142-144,共3页
Control & Automation
关键词
图像无损压缩
整数小波变换
多级树集合分裂算法
预测编码
image lossless compression
integer wavelet transform
Set Partitioning in Hierarchical Trees
predictive coding
