摘要
BestNeighborhoodMatching (BNM)是目前对损坏图像进行修复并能产生高质量可接受图像的图像恢复算法 ,然而BNM却具有计算复杂度高的局限性 ,因而限制了其实际应用 .文中从BNM中影响算法计算复杂度的关键环节———为坏块寻找最优匹配块的搜索路径出发 ,参照人在寻找相似块时的快速智能方法 ,提出跳跃环顾BNM图像恢复算法 (JLBNM ) .同时 ,对于进行匹配的核心步骤 ,提出具有自适应阀值匹配标准的优化方法 .有关的计算复杂度度量分析及模拟实验均证明JLBNM具有计算复杂度低 ,对损坏图像恢复质量高的特点 .
Best Neighborhood Matching (BNM) is a very good error concealment algorithm to achieve high quality image restoration. However, BNM needs intensive computation, which restricts its real application. A fast image restoration algorithm: Jump and Look all around BNM (JLBNM) is proposed, where an intelligent search method is developed for the most time consuming stage. Meanwhile, an adaptive method is employed for optimization. Theoretical analysis has been carried out for the proposed algorithm. Experiment results show that our JLBNM can efficiently reduce computation time for image restoration, while the restored image with high quality can be obtained.
出处
《计算机辅助设计与图形学学报》
EI
CSCD
北大核心
2002年第4期351-355,368,共6页
Journal of Computer-Aided Design & Computer Graphics
基金
国家自然科学基金 ( 6 99330 2 0 )
国家"八六三"高技术研究发展计划( 86 3 30 6 Z0 0 1 0 1)资
