摘要
针对经典Prewitt算子在织物疵点边缘检测中存在的检测边缘较粗、定位不准确和人为选取阈值会造成边缘点误判等缺点,基于改进的Prewitt算子,提出了一种与非极大值抑制方法相结合的自适应阈值的织物疵点检测方法。该方法对织物原图像采用高斯滤波进行预处理,以消除图像上的光照不均和噪声等影响,增加45°和135°方向模板完善边缘结构,利用非极大值抑制方法细化边缘,并用自适应阈值法确定最优阈值来减少边缘点的误判。通过对不同类型织物疵点的试验结果进行分析,证明改进后的算法具有更好的自适应能力,提高了算法的有效性。
In view of the shortcomings of the classical Prewitt operator in the detection of fabric defects,such as coarse detection edges,inaccurate positioning and false edge judgment caused by artificial selection of threshold,an adaptive threshold detection method for fabric defects was proposed based on the improved Prewitt operator,which combined with non-maximum suppression method.The original image of the fabric was preprocessed by Gaussian filtering to eliminate the influence of uneven illumination and noise on the image,45°and 135°directional templates were added to improve the edge structure,non-maximum suppression method was used to refine the edge,and the adaptive threshold method was used to determine the optimal threshold value to reduce the misjudgment of edge points.Through experiments on different types of fabric defects,the results showed that the improved algorithm had better adaptive ability and improved the effectiveness of the algorithm.
作者
王春妍
沈丹峰
杨国仲
张旭祥
WANG Chun-yan;SHEN Dan-feng;YANG Guo-zhong;ZHANG Xu-xiang(School of Mechanical and Electrical Engineering,Xi′an Polytechnic University,Xi′an 710048,China)
出处
《纺织科技进展》
CAS
2020年第3期42-46,共5页
Progress in Textile Science & Technology
基金
湖北省数字化纺织装备重点实验室开放项目(DTL2018004)。
