一种有效的基于中值滤波的BTC图像编码算法

Efficient BTC image coding algorithms with median filtering

块截短编码 ( BTC)是一种基块的快速有损图象压缩技术。标准 BTC算法的主要缺点是其比特率 ( 2 .0比特 /像素 )相对其他基块图象编码的算法较高 ,如变换编码和矢量量化。为了降低比特率 ,本文提出四种有效的 BTC算法以降低对每个图象块的两个量化数据和位平面编码所需的比特数。其中有两种算法用简单的查表算法对块的高平均值和低平均值编码。在此基础上 ,另外两种算法利用一维中值滤波根信号序列进一步降低位平面的比特率。为了得到更好的性能 ,在每种算法中 ,本文用最优阈值而不用平均值作为量化阈值。实验结果证明所提出算法的有效性。

Block truncation coding (BTC) is a fast lossy image compression technique. The main shortcoming of the standard BTC algorithm is a relatively high bit rate of 2.0 bits per pixel compared to other block\|based image coding schemes such as transform coding and vector quantization. In order to reduce the bit rate, four efficient BTC algorithms are presented to reduce the number of bits required to code two quantization data and the bit plane, which are created by BTC for each image block. Two of these algorithms use simple look\|up\|table methods to code the higher mean and the lower mean of a block. Besides these, the other two algorithms also use the root signal set of 1\|D median filter to further reduce the number of bits required to code the bit plane. In order to obtain better performance, we use the optimal threshold instead of the mean value as the quantization threshold in each proposed algorithm. Experimental results prove the effectiveness of proposed algorithms.