基于各向异性扩散算法的深度图像不匹配边缘增强方法  

Mismatched edge enhancement method for depth images based on anisotropic diffusion algorithm

作　　者：刘兴瑞 徐文鹏[2] LIU Xing-rui;XU Wen-peng(School of Computer and Information Engineering,Xinxiang University,Xinxiang 453003,China;School of Computer,Henan Polytechnic University,Jiaozuo 454003,China)

机构地区：[1]新乡学院计算机与信息工程学院,河南新乡453003 [2]河南理工大学计算机学院,河南焦作454003

出　　处：《沈阳工业大学学报》2022年第6期682-687,共6页Journal of Shenyang University of Technology

基　　金：河南省科技攻关项目(182102210310).

摘　　要：针对传统深度图像边缘增强方法存在图像边缘匹配度低、峰值信噪比小、稳定性低、增强效果不佳的问题,提出了基于各向异性扩散算法的深度图像不匹配边缘增强方法.通过被动测距方法获取深度图像,采用插值算法去除图像孔洞,并依据加权中值滤波法对图像进行滤波.对滤除噪声点的不匹配边缘进行检测,并采用各向异性扩散算法填补不匹配边缘空洞,根据像素点到空洞距离构建不匹配边缘增强模型,完成深度图像不匹配边缘的增强.结果表明,与三种传统方法对比,所提方法的图像边缘匹配度高,峰值信噪比最大,稳定性高,增强效果最佳.Aiming at the problems of lower image edge matching,lower peak signal-to-noise ratio(SNR),poor stability and poor enhancement effect of traditional depth image edge enhancement methods,a mismatched edge enhancement method for depth images based on anisotropic diffusion algorithm was proposed.Depth images were obtained by a passive ranging method,image holes were removed by an interpolation algorithm,and the images were filtered according to a weighted median filtering method.Mismatched edges without noise points were detected,and an anisotropic diffusion algorithm was used to fill holes within the mismatched edges.According to the distance from pixel points to holes,an mismatched edge enhancement model was constructed to enhance the mismatched edges of depth images.The results show that the as-proposed method has higher image edge matching,maximum peak signal-to-noise ratio(PSNR),higher stability and optimal enhancement effect,in comparison with three traditional methods.

关 键 词：各向异性扩散算法 深度图像 不匹配边缘 边缘增强 中值滤波 被动测距 插值算法 峰值信噪比 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]