发明名称 GRAY IMAGE PROCESSING METHOD AND APPARATUS
摘要 The embodiments provide an image processing method and apparatus, which combine a wavelet transform method with a non-local means method, comprising: performing N-level decomposition of an original gray image by using a wavelet transform method; and performing non-local means processing and wavelet reconstruction of each level of components one by one starting from an N-th level, so as to obtain a processed gray image. The image processing method and apparatus can effectively remove various image noises, and well reserve the image details.
申请公布号 US2014294318(A1) 申请公布日期 2014.10.02
申请号 US201414228729 申请日期 2014.03.28
申请人 Fujitsu Limited 发明人 Wu Jianrong;Fu Xuan;Wang Zheng;Higashi Akihiro
分类号 G06T5/00;G06T5/20 主分类号 G06T5/00
代理机构 代理人
主权项 1. A gray image processing method, comprising: performing N-level decomposition of an original gray image by using a wavelet transform method; and processing each level of components one by one starting from an N-th level, to obtain a processed gray image; wherein the processing each level of components in N levels comprises: performing non-local means processing to all pixels of a first component, a second component, a third component and a fourth component of a current level, and performing wavelet reconstruction of the four non-local means processed components, to obtain a first component of a level preceding the current level; wherein the performing non-local means processing to one of all the pixels of a first component, a second component, a third component and a fourth component of the current level comprises: performing non-local means processing to one pixel of the first component of the current level, to obtain a pixel value of the non-local means processed pixel and a weighted value matrix of pixels within a predetermined search range around the pixel, wherein the weighted value matrix is an R×R matrix, R denoting the number of the pixels within the predetermined search range, which is an integer greater than or equal to 1; and using the weighted value matrix to respectively calculate pixel values of the non-local means processed pixels in the second component, the third component and the fourth component of the current level corresponding to the pixel in the first component; or the processing each level of components in the N levels comprises: performing non-local means processing to all the pixels of the first component of the current level, and performing wavelet reconstruction of the non-local means processed first component, second component, third component and fourth component, to obtain a first component of a level preceding the current level; wherein N is greater than or equal to 1, and when the current level is the N-th level, the first component of the current level is a first component obtained after the N-th level decomposition; and when N is greater than 1 and the current level is not the N-th level, the first component of the current level is a first component obtained after processing a component of a level next to the current level; and when N is greater than 1, manners for processing the components of different levels in the N levels are identical or different.
地址 Kawasaki JP