微铣刀同轴全息图像增强方法  

Method of enhancing the quality of in-line holographic images for micro-milling tool

作　　者：程亚亚 于化东[1] 于占江[1] 许金凯[1] 张向辉[1] CHENG Ya-ya;YU Hua-dong;YU Zhan-jiang;XU Jin-kai;ZHANG Xiang-hui(Ministry of Education Key Laboratory for Cross-Scale Micro and Nano Manufacturing,ChangchunUniversity of Science and Technology,Changchun 130022,China)

机构地区：[1]长春理工大学跨尺度微纳制造教育部重点实验室,吉林长春130022

出　　处：《中国光学》2020年第4期705-712,共8页Chinese Optics

基　　金：国家重点研发计划(No.2018YFB1107403);中国“111”计划(No.D17017);吉林省科技发展计划(No.20190101005JH,No.20180201057GX)。

摘　　要：数字同轴全息对刀技术中,再现像中的零级项和离焦共轭像会形成一个强而复杂的背景噪声作用在实像上,严重降低了再现像的质量。针对全息应用中的干扰像问题,提出了一种基于改进自蛇模型滤波的全息图像增强方法,改进后的自蛇模型令每次扩散中只根据初始图像的梯度来选择扩散力度。实验结果表明,改进后的自蛇模型能够避免在扩散过程中由于受大梯度背景噪声影响而出现的“伪轮廓”和边缘锯齿化,弥补了自蛇模型在全息图像应用中的不足。此外,与相位恢复法和多重再现法去干扰像效果相比较,本文提出的改进自蛇模型滤波法不仅对干扰像有更好的抑制作用,还能够增强刀具边缘,有利于实现微铣刀的数字全息对刀。When tool setting with digital in-line holography,the zero-order image and defocused twin-image can form strong and complex background noise,which gets superimposed on the real image and seriously reduces the quality of the reconstructed image.To improve quality of interferential images in digital in-line holography,a holographic image enhancement method using an improved self-snake model is proposed.The improved self-snake model selects a diffusion intensity according to the gradient of the initial image.The experimental results show that the improved self-snake model can avoid the appearance of jagged edges and“pseudo-contours”caused by large gradient background noise during the diffusion process.This improvement outweighs the shortcomings of the self-snake model in holographic imaging.In addition,compared with the phase retrieval and multi-plane reproduction approaches,the improved self-snake model filtering method proposed in this paper not only has better suppression for interferential images but also can enhance the edge of the tool,which is conducive to the actualization of tool-setting using on digital in-line holography.

关 键 词：数字全息 共轭像 自蛇模型 图像去噪 

分 类 号：O438.1[机械工程—光学工程]