华中师范大学学报（自然科学版）2011，Vol.45Issue(4)：547-550,4.
直线运动模糊图像复原的截断奇异值分解方法
Linear motion blurred image restoration based on TSVD
摘要
Abstract
The image restoration problem with symmetry and translational invariance of the point spread function can be transformed into ill-posed deconvolution problem under the assumption of the Neumann boundary condition, and thus can be used TSVD method to solve. But the point spread function of in any direction linear motion blurred images does not have symmetry. While the linear motion blurred image caused merely by the interactions between pixels of the movement direction, in this paper a segments the blurred image with a series of straight parallel lines in the direction of motion, motion blur on each line can be considered similar to the level linear motion blur, the point spread function of the level linear motion blur is symmetry, thus it can restore the blurred image using TSVD method. Numerical experiments are given to illustrate the efficiency of our algorithm.关键词
图像复原/不适定问题/TSVDKey words
image restoration/ ill-posed problems/ TSVD分类
数理科学引用本文复制引用
周俊,吴传生,朱华平..直线运动模糊图像复原的截断奇异值分解方法[J].华中师范大学学报（自然科学版）,2011,45(4):547-550,4.基金项目
国家自然科学基金项目(60974094) （60974094）
中央高校基本科研业务费专项资金(2011-1a-06). （2011-1a-06）
