采用小波与神经网络的高效图像数据压缩方案

A high efficiency image data compression scheme based on wavelet and neural network

通过构造特别的映射、整函数和BP神经网络,获得一套基于神经网络的无损数据压缩方案。由于该方案能压缩已被小波编码压缩过的数据,因此将其嵌套入一好的小波编码系统就可以获得一种基于小波与神经网络的高效图像数据压缩方案。实验证明,该高效方案对于Lenna图像的压缩比为43∶1, 并且恢复的图像有较好的视觉效果。

A lossless data compression scheme based on neural network is obtained through the structure specific mapping, integral function and BP neural network. Since the scheme can compress the data that already have been compressed by wavelet coding, a high efficiency image data compression scheme based on wavelet and neural network is obtained through embedding a good wavelet coding system. Experiments have proved that the compression ratio of the scheme for Lenna image is 43 to 1 and the restored image has a better visual effect.