Considering pixels of different levels have different significance, we utilize improved embedded zerotree wavelet (EZW) to get multilevel thresholds quantization.
根据不同等级像素重要性不同的特性,采用改进的嵌入零树小波编码方法对预测结果进行多阈值量化。
ZBP exploits the correlation among the Zerotree symbols and the bit data of wavelet coefficients, so the efficiency of arithmetic coding is improved.
ZBP不仅充分利用了零树符号之间的相关性,而且从位数据的层面上挖掘出了小波系数值之间的相关性,从而提高了算术编码的性能。
The experiment shows that the embedded zerotree coding method can be used in the case of nonseparable wavelet transform and has good PSNR at low compression ratio.
实验表明嵌入式零材编码方法可以用在非分离型二维子波变换情况,在低比特率时有很好的巅值信噪比PSNR。
The experiment shows that the embedded zerotree coding method can be used in the case of nonseparable wavelet transform and has good PSNR at low compression ratio.
实验表明嵌入式零材编码方法可以用在非分离型二维子波变换情况,在低比特率时有很好的巅值信噪比PSNR。
应用推荐