An Improved Human Body Contour Extraction Method Based on Marr-Hildreth Operator
Traditional human contour extraction is mostly based on RGB images and edge detection operators,but RGB imag-es are greatly affected by the environment.Operators such as Marr-Hildreth and Canny will also weaken the edge information of the human body and cannot effectively distinguish the human body from the background.In view of the above deficiencies,an improved human contour extraction method is proposed based on the depth map obtained by Kinect.First,the human depth map is obtained from Kinect,median filter and edge-preserving Gaussian filter is used to form a joint filter to replace the Gaussian filter in the tradi-tional Marr-Hildreth operator,and finally Laplacian and zero-cross point detection processing are performed on the filtered image,the outline of the human body edge is gotten.The experimental results show that the algorithm in this paper can effectively extract the outline of the human body,and the protection effect of edge points is better.
depth mapMarr-Hildreth algorithmcombined filtersimage edge information
国家科技期刊平台
NSTL
万方数据
