Precise Cable Contour Extraction Method Based On Candidate Point Merge Search

Wang Hao-tian,Li Xiao-run,Feng Yue-ming

PROCEEDINGS OF 2020 5TH INTERNATIONAL CONFERENCE ON MULTIMEDIA AND IMAGE PROCESSING (ICMIP 2020)(2020)

引用 0|浏览0
暂无评分
摘要
Aiming at the problem that the existing image contour extraction algorithm can't be accurate to the pixel level, this paper proposes a cable contour extraction method based on candidate point merge search. The algorithm is divided into three steps: Firstly, the rough position of the target area is extracted by color, and the contour candidate points with larger gradients near the boundary are obtained; Then the adjacent candidate contour points are uniformly connected into several line segments and the angle of each line segment and the gradient sum through the pixel points are calculated; Finally define the target of the line segment merge and the principle of distance attenuation and angle attenuation and perform a merge search, by iterating until it merges into a complete contour. For contour sections where the correct candidate point cannot be found, it can be skipped by adding flag bit. The experimental results show that this algorithm can extract the cable contour at pixel level without manual parameter setting, which is of great significance for cable quality detection.
更多
查看译文
关键词
cable contour, pixel level, candidate point, merge search, sum of gradients
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要