大尺度三维视觉测量中的离焦模糊图像恢复 .pdfVIP

学兔兔 第3l卷 第l2期 仪 器 仪 表 学 报 V01．3l No．12 2010年12月 Chinese Journa1 of Scientific Instrument Dec．2010 大尺度三维视觉测量中的离焦模糊图像恢复术 张益昕，王 顺，张旭苹 (南京大学光通信工程研究中心 南京 210008) 摘 要：为了实现大型设备的几何量测量，需要对多幅立体罔像进行拼接。南于镜头的焦距同定，当观测不同距离的场景时往 往使图像产生离焦模糊，从而丧失图像细节，降低了拼接精度。提出了一种基于维纳滤波的离焦模糊罔像恢复算法。利用模糊 图像频谱上的暗环半径1卉计退化函数的参数，并通过对原始冈像的边界进行处理消除了图像边界条件的不连续性。实验结果 显示该算法可以有效地降低罔像恢复过程中产生的振铃现象，对离焦模糊图像的恢复质量良好。 关键词：模糊识别；离焦模糊；维纳滤波：振铃现象 中图分类号：TI~91．7 文献标识码：A 国家标准学科分类代码：510．40 Defocused blur image restoration in large scale 3D vision measurement Zhang Yixin，Wang Shun，Zhang Xuping (Institute of Optical Communication Engineering，Nanfing University，Nanjing 210008，China) Abstract：In order to realize geometric quantity measurement of large scale equipment a stereo image mosaic ， method is needed．Unfortunately，the lens usually has fixed focal length，which causes defocused blue image while observing scenes in different distances，which makes the captured image lose its high frequency feature and limits the resolution of image mosaic．In this paper，a Wiener filter based algorithm for defocused blur image restoration is presented．Blur identification with dark loop searching in frequency dimension is proposed．The boundary of image is symmetrized to reduce the ringing effects during the process of image restoration．Experiment results indicate that the new method effectively reduces the ringing effects and restores the defoeused blur image effectively． Key words：blur identification；defocused blur；Wiener filter；ringing effect