CAJ | 학술논문

为更好地采集动车车号部分的图像,针对图像中车号部分所包含信息变化微小的特点,采用了一种基于虚拟仪器技术开发平台 LabVIEW 的去除条纹噪声的方法.首先,在频域内将含有条纹噪声的图像与原始图像的频谱图二值化,并对二值化的频谱图像进行逻辑操作,同时结合多样本进行综合处理,最后运用巴特沃斯带阻滤波法滤除条纹噪声;然后,在空域内通过IMAQ GrayMorphology.VI和IMAQ BCGLookup.VI对处理后的图像进行增强处理,锐化因频域内滤波而边缘虚化的车号图像.进行动车图像条纹去噪处理实验,运用均方误差(MSE)和峰值信噪比(PSNR)对去噪效果进行评价,并与普通带阻滤波、矩匹配法等去噪方法进行对比.结果表明,本文方法消除噪声的效果更显著.
위경호지채집동차차호부분적도상,침대도상중차호부분소포함신식변화미소적특점,채용료일충기우허의의기기술개발평태 LabVIEW 적거제조문조성적방법.수선,재빈역내장함유조문조성적도상여원시도상적빈보도이치화,병대이치화적빈보도상진행라집조작,동시결합다양본진행종합처리,최후운용파특옥사대조려파법려제조문조성;연후,재공역내통과IMAQ GrayMorphology.VI화IMAQ BCGLookup.VI대처리후적도상진행증강처리,예화인빈역내려파이변연허화적차호도상.진행동차도상조문거조처리실험,운용균방오차(MSE)화봉치신조비(PSNR)대거조효과진행평개,병여보통대조려파、구필배법등거조방법진행대비.결과표명,본문방법소제조성적효과경현저.
A new method of eliminating strip noise based on LabVIEW is proposed according to the feature that the image information of every bullet train number has imperceptible changes. Firstly,in the frequency domain,a series of logic opera-tions was used on the binaryzation image of both noise image and the original image. After multiple-sample processing,the Butterworth band rejection filter was used for eliminating the strip noise. Then,in the space domain,IMAQ GrayMorphol-ogy.Ⅵ and IMAQ BCGLookup.Ⅵ were used to enhance the processed image so as to sharpen the blurring edges of the train number image caused by filtering in the frequency domain. After evaluating the processed image via the MSE and PSNR image evaluation methods and comparing with the waves in the ordinary band stop filtering,the conclusion is that the method proposed can get rid of the strip noise very effectively.