摘要
本文设计了一种神经网络矢量量化编码方法———基于自组织特征映射的神经网络的矢量量化 .利用Kohonen网络的自组织聚类功能 ,设计矢量量化器码书 ,实现矢量量化 .该神经网络速度快 ,效率高 ,适用于语音和图象数据压缩 ,并对SOFM算法应用于图象矢量量化进行系统的研究 .
In this paper a coding scheme for vector quantization is presented by using a neural network_vector quanrizer codebook designs for Image compression using neural nerwork.The clustering function of the kohonen self_organizing feature map is used to design codebook of vector quantization.Processing inout vectors tast and effectively,the vector quantization coder based upon neural networks is suitable for voice and image data compression.And then I do some researching on the SOFM using as image vector quantization.
出处
《松辽学刊（自然科学版）》
2001年第4期88-90,共3页
Songliao Journal (Natural Science Edition)
关键词
神经网络
矢量量化
图象压缩
自组织特征映射
码书
neural network
vector quantization
image compression
self_organizing feature map
codebook
