可变深度嵌入小波图像压缩算法

Depth Variable Compression Algorithm Using Embedded Image Wavelet

为了提高小波嵌入压缩算法的速度,提出了一种可变深度嵌入小波压缩算法。由图像小波变换能量聚集和子带间的幅度衰减性等特性,在小波分层嵌入编码过程中,对不同的重要图,部分小波子带的系数幅度是不会超过当前的阈值,从而可以忽略不计,即小波树的深度随着重要图的不同而改变,也就是变深度嵌入编码。实验结果显示,对于自然图像不需要考虑的层中的小波系数幅值几乎全部小于当前的阈值,从而和EZW算法相比,编码不会影响压缩率(或图像质量),而编解码过程中的运算量明显下降。

The embedded image compression algorithm is discussed. To speed the embedded image compression algorithm with wavelet,a depth variable embedded coding scheme is proposed. With the concentration of the energy and the decay of the magnitude of wavelet coefficients across sub-bands, the embedded algorithm only scans parts of the wavelet sub-bands in original steps. The depth variable embedding is thus realized. The experimental results show that the speed is higher and the recovery quality is the same as the previous schemes.