报告题目:Non-local means and learning based algorithm
报 告 人:金其余
工作单位:内蒙古大学
报告时间:2021-06-23 16:00-17:00
腾讯会议ID: 857 784 218
报告摘要:
We introduce a local patch matching based denoising algorithm as preprocessing algorithm and use preprocessed image to get high accuracy patch matching group and then replace all preprocessed patches by noisy patches. After patch matching, all noisy patches yi ,1≤i≤q in a similar patch group can be regarded as independent and identically distributed (i.i.d.) multivariate random variable following a multivariate Gaussian distribution. For this reason, all sparse covariance matrix estimation methods can be used to remove Gaussian noise.
报告人简介:
金其余,内蒙古大学教授。法国南布列塔尼大学应用数学博士,巴黎六大、上海交通大学博士后,巴黎-萨克雷高等师范学校访问学者,2015年由内蒙古大学数学科学学院高层次引进。长期与国内外多所大学保持合作,包括法国巴黎-萨克雷高等师范学校、巴黎六大、Centre Inria Rennes等。研究领域包括:图像处理、计算机视觉与最优化。相应成果发表于SIAM Journal on Imaging Sciences、Cell子刊Structure、Journal of scientific computing、Journal of Mathematical Imaging and Vision等期刊。主持国家自然科学基金、内蒙古自然科学基金等项目多项。