基于遺傳算法的矢量量化
VQ BASED ON GENETIC ALGORITHM
-
摘要: 本文提出了一種基于遺傳算法的矢量化方法。矢量量化碼書設(shè)計(jì)本質(zhì)是搜索訓(xùn)練矢量的最佳分類。遺傳算法有卓越的全局優(yōu)化搜索能力,易搜索到全局最優(yōu)的矢量分類,形成高度優(yōu)化的碼書,可克服傳統(tǒng)方法局部?jī)?yōu)化的缺陷。該算法不依賴初始條件、魯棒性好、結(jié)構(gòu)規(guī)則、并行性高。
-
關(guān)鍵詞:
- 矢量量化; 遺傳算法; 適應(yīng)度函數(shù); 分類
Abstract: A new vector quantization(VQ) approach based on genetic algorithm(GA) is presented in this paper. VQ codebook design is essentially a classification of training vectors. Because of GA s global optimum ability, using GA for codebook training can obstain a global optimum codebook, and overcome local optimum limitation of traditional algorithms. This method is also independent on initial conditions, more robust, highly regular and parallel in architecture. -
Linde Y,et al. An algorithm for vector quantizer design. IEEE Trans. on comm., 1980, COM 28(l): 84-95.[2]Laregetto F, et al. Unbalanced tree structure frame adaptive vector quantization of image se[3]quences.SPIE,1990.1224: 281-304.[4][3][5]Holland J H. Adaptation in Nature and Artificial Systems.The University of Michigan Press, 1975. -
計(jì)量
- 文章訪問數(shù): 2086
- HTML全文瀏覽量: 101
- PDF下載量: 574
- 被引次數(shù): 0
下載:
