一种新的非因果递归模型在纹理合成及压缩中的应用

A NEW APPLICATION OF RECURSIVE NONCAUSAL MODEL TO TEXTURE SYNTHESIS AND DATA COMPRESSION

非因果模型在纹理合成与压缩中得到了广泛应用.但由于非因果模型不能递归实现且建模方法非常复杂,所以限制了它的实用化.本文针对这一课题提出一种可递归实现的非因果模型,并将它应用于纹理合成及数据压缩.实验结果表明这种模型可代替一般的非因果SAR模型进行纹理图象合成.

Using noncausal model to texture synthesis and data compression has been discussed. In practice, its application has been limited as the model can not be implemented recursively. In this paper, we propose a recursive noncausal model and use it for synthesis and data compression of natural textures. Experimental results have shown that the proposed model can be used in such applications in lieu of the conventional noncausal SAR models.