论文部分内容阅读
详细研究了小波变换传递误差、尺度、快速 Mallat算法加权量化对分层树剖分 (SPIHT)编码图象压缩质量的影响 ,并给出了基于 VC++的仿真结果 ,结果表明 ,小波变换尺度对图象压缩编码效果有明显的影响 ,通常要在大于 4的尺度上作图象压缩编码才可以得到较好的压缩效果 ,不同的边界延拓 ,不同尺度的取整重构、加权量化对图象压缩效果的影响是不同的 ,恰当选取加权量化步长能达到更好的图象压缩效果 ,一般而言 ,同一尺度下随加权量化步长的增加 ,压缩效果变差 .
The influence of wavelet transformation transfer error, scale and fast Mallat algorithm weighting on the compression quality of SPIHT coding image is studied in detail. The simulation results based on VC ++ are given. The results show that the scale of wavelet transform As the compression coding effect has a significant impact, usually more than 4 in the scale of the image compression coding can be better compression effect, different boundary extension, different scales rounding reconstruction, weighted quantization image The effect of compression is different. Selecting proper weighted quantization step can achieve better image compression effect. Generally speaking, with the increase of weighted quantization step on the same scale, the compression effect becomes worse.