改进向量量化算法的图像压缩研究

针对于传统的LBG向量量化压缩图像效果不是很好,而提高其效果的方法之一是增加向量的维数,但这也会增加运算的复杂性的情况下,提出一种改进的LBG算法来实现对图像分解效果的提高.该方法通过LBG算法先对图像进行量化编码,再用原图像减去该编码恢复的图像而得到量化误差;对这个量化误差还用LBG算法进行编码量化.还原时先恢复原图像的量化编码,再加上恢复的误差量化编码.仿真结果表明,改进算法运算复杂度不会增加,图像压缩后还原效果较好,同等条件下能得到更高的信噪比和峰值信噪比.

The Study of Image Compression Method with Improved Vector Quantization Algorithm

LBG vector quantization algorithm is used to compress image traditionally, but the performance is not always perfect, one method to improve it is just to add the dimension at the cost of increasing the computing complexity. An improved LBG algorithm is proposed to improve the effect of image decomposition in this paper. This method uses LBG algorithm for quantization coding, then get quantization error via the original image subtracts the restoration image of coding, also use the LBG algorithm to quantization-code the quantization error. Restore the quantization coding firstly and add it to restoration error quantization coding. Simulation result shows that the new algorithm is simple, it has nice performance, and the SNR and PSNR are improved in the same precondition.