矢量量化是圖像壓縮的重要方法。論文提出了基于Hopfield 神經(jīng)網(wǎng)絡(luò)的圖像矢量量化方法，該方法首先構(gòu)造聚類表格；然后聚類表格按離散Hopfield 神經(jīng)網(wǎng)絡(luò)串行方式運(yùn)行；
最后根據(jù)得到的最終碼字集，對圖像進(jìn)行矢量量化。論文最后給出模擬實(shí)驗(yàn)和結(jié)果比較，結(jié)果表明該方法是有效的，生成的碼本質(zhì)量優(yōu)于傳統(tǒng)的LBG 算法。
關(guān)鍵詞：矢量量化；碼本；LBG 算法；Hopfield 神經(jīng)網(wǎng)絡(luò)
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