论文部分内容阅读
提出一种基于多小波变换结合矢量量化的图像编码算法(MDWT+VQ)。首先对图像进行多小波分解,然后对高频系数用改进后的LBG算法形成的码书进行VQ编码。算法充分利用了多小波域不同分辨率层间各方向子图像的相似性,仅对最高分辨率层进行码书地址索引,低级分辨率层的系数按照一定的组织形式直接套用最高分辨率层的地址索引信息。对比实验的结果验证了该算法在提高图像的重建质量以及在降低位码率方面均比传统的单小波图像编码算法有一定的提高。
An image coding algorithm based on multiwavelet transform and vector quantization (MDWT + VQ) is proposed. Firstly, the image is decomposed by multi-wavelet, and then VQ codes the codebook formed by the improved LBG algorithm with high frequency coefficients. The algorithm makes full use of the similarities between sub-images of different resolution layers in multiwavelet domain, and only indexes the codebook address of the highest resolution layer. The coefficients of the lower resolution layer are directly applied to the highest resolution layer according to a certain organizational form Address Index Information. The experimental results show that the proposed algorithm can improve the quality of image reconstruction and reduce bit-rate compared with the traditional single-wavelet image coding algorithm.