针对非负和支持域受限递归逆滤波算法(NAS-RIF)的缺点, 提出了一种改进方案。首先, 引入高阶统计量去噪, 解决了NAS-RIF算法敏感于噪声的特性; 其次, 利用图像分割技术进行图像支持域的自估计, 改变了NAS-RIF算法中支持域必须是方形的不利限制; 最后, 运用该算法对不同峰值信噪比(PSNR)和背景的含噪图像进行了算法仿真, 并与原算法进行了比较。仿真结果表明, 改进的NAS-RIF算法有效解决了原始算法中存在的问题, 显著提高了恢复图像的质量。

An improved method is presented to overcome the drawbacks of the original NAS-RIF algorithm.To deal with the noise sensitivity of NAS-RIF, higher order statistics denoising was introduced.To solve the problem that the NAS-RIF requires square support region, the image segmentation based self-estimation of image support region was used.At last, simulation was made to various PSNR images and backgrounds.The improved algorithm was compared with the original one.The simulation results show that the improved algorithm solves the problems of the original algorithm, and can improve the quality of recovery image observably.