基于改进Canny算法和Hu矩的物表区域识别  

Object Surface Region Recognition Based on Improved Canny Algorithm and Hu Moment

作　　者：何坚强 翁嘉鑫 陆群 骆杨 蒋成晨 HE Jianqiang;WENG Jiaxin;LU Qun;LUO Yang;JIANG Chengchen(School of Electrical Engineering,Yancheng Institute of Technology,Yancheng 224000;School of Electrical and Information Engineering,Jiangsu University,Zhenjiang 212000)

机构地区：[1]盐城工学院电气工程学院,盐城224000 [2]江苏大学电气信息工程学院,镇江212000

出　　处：《计算机与数字工程》2024年第8期2393-2399,共7页Computer & Digital Engineering

基　　金：国家自然科学基金青年科学基金项目(编号:62003292)资助。

摘　　要：使用轮廓Hu矩匹配法进行物表检测的过程中,为解决轮廓匹配受噪声、比例变换等客观因素影响导致区域匹配识别率降低的问题,提出一种基于改进Canny算法和改进Hu矩的轮廓匹配方法进行物表区域识别。首先采用改进Canny算法降低噪声等对轮廓提取的影响,获取更优的轮廓信息,然后利用尺度归一法改进Hu矩进行物表区域的轮廓匹配,实现区域的匹配与识别。实验结果表明,该方法在噪声等因素干扰下可提取高精度的区域轮廓,且识别率至少达到98.3%,具有强鲁棒性和高识别率,满足物表检测对区域识别的高精度、高识别率要求。In the process of object surface detection using contour Hu moment matching method,in order to solve the problem that contour extraction and geometric moment are affected by objective factors such as noise and scale transformation,a contour matching method based on improved Canny algorithm and improved Hu moment is proposed for object surface area recognition.Firstly,the improved Canny algorithm is used to reduce the influence of noise on contour extraction and obtain better contour information.Then,the scale normalization method is used to improve Hu moment for contour matching of object surface region to realize region matching and recognition.The experimental results show that this method can extract high-precision regional contour under the interference of noise and other factors,and the recognition rate is at least 98.3%.It has strong robustness and a high recognition rate and meets the requirements of high precision and high recognition rate for regional recognition in object surface detection.

关 键 词：物表检测 区域识别 轮廓提取 特征匹配 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]