摘要
利用三维自组织映射网络的自组织功能,设计矢量量化码书,实现图像压缩矢量量化。实验结果表明,码书尺寸一定时,三维SOM网络所恢复的图像质量高于二维SOM网络。
The clustering function of the three-dimensional self-organization mapping network is used to design codebook Of vector quantization. Experiment results shows that the peak signal-to-noise ratio （PSNR） of the three- dimensional self-organization mapping neural network is high than the PSNR of the two-dimensional self-organization mapping neural network.
出处
《科技通报》
北大核心
2012年第2期67-68,73,共3页
Bulletin of Science and Technology
关键词
自组织映射
矢量量化
图像压缩
Self-organization mapping
Vector Quantization
Image compressing
