基于Hopfield神经网络的图像矢量量化

矢量量化是图像压缩的重要方法。论文提出了基于Hopfield 神经网络的图像矢量量化方法，该方法首先构造聚类表格；然后聚类表格按离散Hopfield 神经网络串行方式运行；
最后根据得到的最终码字集，对图像进行矢量量化。论文最后给出模拟实验和结果比较，结果表明该方法是有效的，生成的码本质量优于传统的LBG 算法。
关键词：矢量量化；码本；LBG 算法；Hopfield 神经网络
Abstract：Image vector quantization is an important method in the image compression field. This paper proposed a method of image vector quantization based on hopfield Neural Network.This method first built a clustering form; then the clustering form worked according to the asynchronous mode of discrete hopfield neural network; Finally carried on the image vector quantization according to the final codebook set. The paper finally produced the
experiments and the result comparison, the result indicated this method is effective and the codebook quality surpassed traditional the LBG algorithm.
Key Words：Vector Quantization; Codebook ;LBG Algorithm; Hopfield Neural Nework1