摘要
首先介绍了图像处理技术的特点和优势,然后从图像预处理、特征值提取和磨损面积计算等方面分析了收割机刀具磨损检测原理,并从软硬件体系结构入手,实现了一套基于图像处理的收割机刀具磨损状态监测系统。试验结果表明:在336次实际监测中,系统成功识别诊断刀具状态312次,成功率高达92.86%,平均监测一次的时间为6.67s,准确性高,识别速度快。
It introduces the features and advantages of image processing technology. And then from the image preprocessing, feature extraction and wear area calculation and other aspects of the detection principle of tool wear harvester, and from the hardware and software architecture design, it realizes a set based on the state monitoring system of harvester tool wear image processing. The experimental results show that in the 336 actual monitoring system, the system successfully identified 312 times of tool condition, and the success rate was 92.86%. The average monitoring time was 6.67s, with high accuracy and fast recognition speed.
作者
杨德义
王平岗
吴东林
Yang Deyi;Wang Pinggang;Wu Donglin(Luohe Vocational Technology College, Luohe 462300,China;Luohe Vocational College of Food , Luohe 462300,China)
出处
《农机化研究》
北大核心
2019年第4期228-232,共5页
Journal of Agricultural Mechanization Research
基金
河南省高等学校重点科研计划项目(17B520022)
关键词
收割机刀具
磨损检测
图像处理
预处理
特征值提取
harvester cutters
wear detection
image processing
preprocessing
eigenvalue extraction
