CAJ | 학술논문

图像边缘检测过程中产生的许多噪声边缘不仅影响后续算法的效率,还严重影响目标的准确识别.这里提出一种基于最大连续子序列算法的目标预定位方法.先利用LoG提取图像边缘,在标定边缘坐标的基础上,计算所有边缘的最大灰度差,经阈值初步去除大部分噪声边缘后,再用最大连续子序列算法将与缺陷边缘相接的噪声边缘滤除,最后利用区域填充得出待定缺陷区域.实验结果表明,该方法能有效地去除噪声边缘并获得包含缺陷边缘的掩膜区域,从而有效提高后续缺陷识别算法的效率.
도상변연검측과정중산생적허다조성변연불부영향후속산법적효솔,환엄중영향목표적준학식별.저리제출일충기우최대련속자서렬산법적목표예정위방법.선이용LoG제취도상변연,재표정변연좌표적기출상,계산소유변연적최대회도차,경역치초보거제대부분조성변연후,재용최대련속자서렬산법장여결함변연상접적조성변연려제,최후이용구역전충득출대정결함구역.실험결과표명,해방법능유효지거제조성변연병획득포함결함변연적엄막구역,종이유효제고후속결함식별산법적효솔.
Numerous noise edges produced in the process of image edge detection can affect the efficiency of the following algorithm and accurate recognition of the targets. A target pre-positioning method based on maximum continuous subsequence al-gorithm(MCSA)is proposed. The image edge is extracted by using LoG,and the maximum grayscale differences of all the edges are calculated on the basis of calibrating edge coordinates. Most noise edges are eliminated by threshold,then noise edges con-nected with defect edges are eliminated by using MCSA. The undetermined defect regions are obtained by region filling. The ex-perimental results show that this method can eliminate noise edges effectively and obtain mask regions containing defect edges, which can improve the efficiency of subsequent defect recognition algorithm greatly.