摘要
针对含欧拉弹性能的传统图像噪声去除变分模型,提出一种全新的快速的对偶投影算法.将原问题转化为简单的迭代优化加权总变差问题求解,并通过引入1个对偶变量和1个约束条件设计相应的对偶方法,且约束可通过投影计算.实验表明,所提出的方法在取得较高的峰值信噪比的情况下具有较高的计算效率.
Aiming at the typical variational model with Euler’s elastica of image denoise,a new fast dual projection algorithm was proposed.The original problem was transformed into a simple iterative optimization weighted total variation problem,and the corresponding dual method was designed by introducing a dual variable and a constraint condition,and the constraint could be calculated by projection.Experiments showed that the proposed method had higher computational efficiency in the case of achieving a higher peak signal to noise ratio.
作者
宋金涛
于晗
潘振宽
SONG Jintao;YU Han;PAN Zhenkuan(College of Computer Science and Technology,Qingdao University,Qingdao,Shandong 266071,China;College of Computer Science and Technology,National University of Defense Technology,Changsha,Hunan 410000,China)
出处
《福州大学学报（自然科学版）》
CAS
北大核心
2020年第3期296-301,共6页
Journal of Fuzhou University(Natural Science Edition)
基金
国家自然科学基金资助项目(61772294)。
关键词
欧拉弹性能
图像恢复
变分法
对偶方法
投影法
Euler’s elastica
image restortation
variational method
dual method
projection method
