摘要
以桑塔纳14寸轮毂为研究对象,Halcon软件为研究工具,提出一种基于canny算子的轮毂区域识别方法。具体的,采集均匀光源下的轮毂图像,通过灰度化、二值化、高斯滤波、膨胀处理从原始图像中提取“战斧”形区域与模板匹配,实现轮毂区域识别。研究结果表明该方法对轮毂“战斧”形区域匹配良好,同时适用于其它类型轮毂区域识别,为轮毂识别技术提供理论基础。
With the improvement of the automation degree of wheel hub production line,the production requirements of wheel hub increase accordingly.The existing wheel hub recognition methods extract more regions of the wheel hub,and the recognition repetition rate is high.In order to solve this problem,the method of hub recognition is studied by extracting the target area of hub to meet the requirements of hub production.Taking Santana 14 inch wheel hub as recognition object and Halcon software as research tool,a canny operator method for identifying“Tomahawk”shaped area of wheel hub is proposed.Specifically,the hub image under uniform light source is collected,and the“Tomahawk”shape area is extracted from the original image by graying,binarization,Gaussian filtering and dilation processing to realize the hub recognition.The results show that the method has high accuracy and robustness to the intensity of light source.This method improves the efficiency of wheel hub recognition,and also provides reference for solving similar image processing problems.
作者
狄玉鹏
孙见君
马晨波
郑伟
马琳博
Di Yupeng;Sun Jianjun;Ma Chenbo;Zheng Wei;Ma Linbo(School of mechanical and electronic engineering,Nanjing Forestry University,Nanjing Jiangsu,210000)
出处
《电子测试》
2021年第3期30-33,84,共5页
Electronic Test
基金
江苏省科技项目,产业前瞻与共性关键技术——竞争项目(BE2017026)。
