摘要
育秧盘因长期使用或运输不当等原因易出现破损,影响了工厂化自动育秧生产线的作业性能。针对水稻育秧盘的裂缝缺陷,采用机器视觉技术,利用图像灰度化、自适应阈值处理和形态学运算等方法对育秧盘的裂缝缺陷进行检测;利用平均值法对RGB图像进行灰度化,采用自适应阈值处理对灰度化后的图像进行二值化,然后通过形态学膨胀对断裂处的裂缝进行连通,以求得最大连通区域,再运用最小外接矩形把最大连通区域标记出来,实现对裂缝缺陷的识别。试验结果表明:对带有裂缝缺陷的育秧盘的正确识别率可达到94.38%。本文的研究为育秧盘裂缝缺陷的检测和判定奠定了基础。
The seedling tray was easily damaged due to long-term use or improper transportation,which affects the operation performance of the factory automatic seedling raising planter.For the crack of rice seedling-raising tray,machine vision technology was used to detect the crack of rice seedling-raising tray by means of image gray-scale,adaptive threshold and morphological operation.The average method was used to gray the RGB image,and the gray-scale image was binarized by means of adaptive threshold,and then the crack at the fracture are connected by morphological expansion to obtain the maximum connected area.At last,the minimum enclosing rectangle was used to mark the maximum connected area to realize the recognition of crack defects.The experimental results show that the correct recognition rate of seedling trays with crack can reach 94.38%.This study laid a foundation for the detection and determination of seedling tray crack detection.
作者
马旭
袁志成
王宇唯
季传栋
温志成
邓向武
Ma Xu;Yuan Zhicheng;Wang Yuwei;Ji Chuandong;Wen Zhicheng;Deng Xiangwu(College of Engineering,South China Agricultural University,Guangzhou 510642,China)
出处
《现代农业装备》
2019年第1期25-31,共7页
Modern Agricultural Equipment
基金
国家重点研发计划课题(2018YFD0700703)
现代农业产业技术体系建设专项资金项目(CARS-01-43)
关键词
育秧盘
阈值处理
形态学运算
裂缝检测
机器视觉
nursing seedling tray
threshold processing
morphological operation
crack detection
machine vision
