复杂环境下螺栓防松线快速分割算法  

Fast Segmentation Algorithm for Bolt Anti-loosening Line in Complex Environment

作　　者：康晶杰 张立君[1] 孙远东[1] 杨晓禹 王若兰 赵天豪 Tang Jingjie;Zhang Lijun;Sun Yuandong

机构地区：[1]中国兵器科学研究院宁波分院,浙江宁波315000

出　　处：《装备机械》2024年第2期5-11,共7页The Magazine on Equipment Machinery

基　　金：宁波市自然科学基金资助项目(编号:202003N4342)。

摘　　要：针对现有螺栓防松线分割方法无法适应复杂环境的问题,提出了一种复杂环境下螺栓防松线快速分割算法。这一算法可以对Lab和RGB颜色空间进行转换,通过选取适当参数在Lab和RGB颜色空间实现a分量非线性拉伸和R分量最优阈值分割,得到螺栓防松线图像。选取50张光照和背景复杂的图像作为测试集,与传统阈值算法、基于Lab颜色空间的K均值聚类算法在准确率、精确度、查全率、运行时间方面进行对比。试验结果表明,所提出的快速分割算法在复杂环境下具有更强的环境适应能力和更短的耗时,具备较高的工程应用价值。In order to solve the problem that the existing bolt anti-loosening line segmentation method cannot adapt to the complex environment,a fast segmentation algorithm for bolt anti-loosening line in complex environment was proposed.This algorithm can convert the Lab and RGB color spaces,and realize the nonlinear stretching of the a component and the optimal threshold segmentation of the R component in the Lab and RGB color spaces by selecting the appropriate parameter,and obtain the bolt anti-loosening line image.50 images with lighting and complex background were selected as the test set,and compared with the traditional threshold algorithm and the K-means clustering algorithm based on Lab color space,in terms of accuracy,precision,recall and running time.The experimental result shows that the proposed fast segmentation algorithm has stronger environmental adaptability and shorter time consumption in complex environment,and has high engineering application value.

关 键 词：螺栓 防松线 快速分割 算法 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]