基于机器视觉的穴盘苗检测试验研究

Experimental Study on Detection of Plug Tray Seedlings Based on Machine Vision

针对穴盘苗出现空穴和不合格坏苗,影响蔬菜种苗销售价格,不利于机械化移栽及后续栽培,而人工剔苗补苗费时费力的问题,提出了利用机器视觉技术检测穴盘苗空穴及不合格苗、传输检测结果的方法,为穴盘苗自动化剔除空穴与不合格苗及补苗作业提供技术基础。穴盘苗空穴及不合格苗检测硬件系统由工业相机、PC机及PLC构成,通过CKVisionBuilder软件对图像进行处理,获取每个穴孔区域的像素数量,判断幼苗状态,并将获取的判断结果传输至PLC中。试验结果表明:苗龄13天的72穴意大利生菜、白玫瑰白菜及广府1号油菜心穴盘苗坏穴的检测正确率达到95.8%以上。

In order to solve the problem of vacancy and unqualified bad seedlings in potted tray seedlings,which affect the selling price of vegetable seedlings,mechanized transplanting and follow-up cultivation,and time-consuming and laborious manual culling and filling seedlings,this paper puts forward a method of detecting vacancy and unqualified seedlings in potted Tray Seedlings by machine vision technology and transferring test results.It provides technical basis for automatic removal of bad holes and unqualified seedlings,and automatic filling.The detection hardware system of cavitation and unqualified seedlings of plug seedlings is made up of industrial cameras,PC and PLC.The image is processed by CKVisionBuilder software.It obtains the number of pixels in each hole area,is judges the seedling state,transmits the judgement result to PLC.The results showed that the correct rate of detecting bad holes in 72 seedlings of Italian lettuce,White rose cabbage and Guangfu No.1 Chinese flowering cabbage was over 95.8%.