基于神经元网络和分形的JPEG改进算法

Modified JPEG Based on Artificial Neural Network and Fractal Theory

文结合分形理论和自组织神经元网络来进行静止图像的压缩编码 ,自适应地调整量化矩阵 ,最大限度地压缩图象的冗余信息 ,并结合快速算法 ,使得压缩过程最优。实验结果表明 ,在视觉效果相同的情况下 ,比

In this paper,one approach on still image compression is presented.Modified JPEG based on ANN and fractal theory could adjust Q matrix automatically and compress useless information in still image as more as possible.It has the more compression ratio than standard JPEG on the base of the same video quality.