摘要
针对铝方棒凹面朝向检测速度慢、准确率低的问题提出了改进的线扫描检测算法,引入图像比例法进行凹面区域提取,缩小扫描线检测范围;采用像素点均值变化替代极限亮度差统计,确定铝方棒凹面朝向。仿真试验表明:改进的线扫描检测算法准确率和平均耗时分别为99.33%、46.35 ms,比传统测算法,准确率提升了5.72%、平均耗时减少了30.92%。
To address the issues of slow detection speed and low accuracy in identifying the concave surface orientation of aluminum square bars,an improved line-scan detection algorithm was adopted.This method incorporates an image ratio technique for concave region extraction,narrows down the scanning line detection range,and replaces extreme brightness difference statistics with pixel mean value variation to determine the concave surface orientation of the aluminum square bars.Simulation results demonstrate that the improved line-scan detection algorithm achieves an accuracy of 99.33%with an average processing time of 46.35 ms.Compared with conventional detection methods,this represents a 5.72%improvement in accuracy and a 30.92%reduction in average processing time.
作者
闫磊起
樊高峰
李建华
YAN Leiqi;FAN Gaofeng;LI Jianhua(Mechanical and Electrical Engineering College,Lanzhou University of Technology,Lanzhou,Gansu 730050,China;Yunnan Wenshan Aluminum Co.,Ltd.,Wenshan,Yunnan 663001,China)
出处
《云南冶金》
2025年第3期130-136,共7页
Yunnan Metallurgy
基金
国家重点研发计划项目(2020YFB1713603)。
关键词
铝方棒
凹面朝向
图像比例法
像素点均值变化
aluminum square bars
concave surface direction
image ratio method
variation of pixel mean values
作者简介
闫磊起(1998-),男,甘肃白银人,硕士研究生,主要从事机器视觉领域相关研究;通信作者:樊高峰(1980-),男,山西运城人,工程师,主要从事电解铝智能铸造系统相关研究工作,E-mail:1968833854@qq.com。
