Research on Visual Recognition Method for Residual Materials in Wooden Spoon Molding Production Line
张庆功 1赵辉 1翟庆磊 1曹文豪 1王锦添1
扫码查看
点击上方二维码区域，可以放大扫码查看
作者信息
1. 东北林业大学机电工程学院,黑龙江 哈尔滨 150040
折叠
摘要
针对木勺模压生产线,生产过程中一次性木勺在模具中产生遗料现象,利用相关图像识别技术,设计了一种在木勺生产过程中使用的遗料识别系统.以Raspberry Pi 4B作为控制板,利用Python设计编写相关遗料识别程序.利用该程序对相关图像进行霍夫圆检测以及透视变换的处理,借以帧差法对比图像与图像之间的差异从而判断模具中有无遗料.实验证明,编写的遗料识别程序可以满足预期生产要求,程序的准确率接近100%,稳定性好.
Abstract
In response to the phenomenon of leftover materials generated by disposable wooden spoons in the mold during the production process of the wooden spoon molding production line,a leftover material recognition system used in the production process of wooden spoons was designed using relevant image recognition technology.Using Raspberry Pi 4B as the control board,use Python to design and write relevant legacy identification programs.Use this program to perform HoughCircles detection and perspective transformation on relevant images,and use frame difference method to compare the differences between images to determine whether there are any leftover materials in the mold.The experiment has proven that the written residual material identification program can meet the expected production requirements,with an accuracy rate of nearly 100%and good stability.
关键词
图像识别/木勺/Raspberry/Pi/4B/霍夫圆检测
Key words
image recognition/wooden spoon/Raspberry Pi 4B/HoughCircles detection
维普
万方数据