基于机器视觉的纪念币表面划痕提取算法研究

Machine Vision Based on Scratch Extraction Algorithm for Commemorative Coin

针对纪念币表面划痕处理问题,本文设计一个基于机器视觉的纪念币表面划痕的提取算法,将该算法应用在纪念币的表面划痕检测中。将图像的预处理做为特征提取的前提,采用粒子群算法对图像的阈值进行确定,将局部与全局的阈值相结合,使纪念币的阈值分割更加明显,对阈值分割后的图像逐行扫描,提取划痕信息,最终实现纪念表面划痕的定位和标记。

Scratch treatment on commemorative coins,Designed an automatic recognition system for scratches on commemorative coin surfaces based on machine vision,Applying its scratch extraction algorithm to the surface scratch detection of commemorative coins. This article research content,Image preprocessing prepares for feature extraction,Using PSO to determine the threshold of the image,And improved its,When combining local and global,More suitable for threshold segmentation of commemorative coins,By scanning the threshold divided image line by line,Scratch information is extracted and marked,Finally,the characteristic mark on the surface of the commemorative coin is realized.