基于机器视觉的坚果尺寸检测系统设计

Design of nut size detection system based on machine vision

针对坚果分类效率低、精确度低、成本高等问题,该文研究并设计了基于机器视觉的坚果尺寸检测系统,以核桃坚果为研究对象,对核桃的尺寸进行检测,以获取品质高、质量好的核桃。首先,通过工业相机对核桃图像进行采集;其次,利用视觉软件Halcon进行Blob分析、分水岭阈值分割、矩形检测等方法进行图像处理,获取核桃尺寸的相关数据;最后,将被检测物尺寸的实时数据传输到工控机上,利用Visual Studio 2019软件实现数据通信。实验结果表明,该坚果尺寸检测系统稳定,能够保证较高的准确率。

Aiming at the problems of low efficiency,low accuracy and high cost of nut classification,a nut size detection system based on machine vision was researched and designed.In this paper,walnut nuts are used as the research object,and the size of walnuts is tested to obtain high quality and good quality walnuts.First,the walnut image is collected by an industrial camera.Secondly,the visual software Halcon is used for image processing by Blob analysis,watershed threshold segmentation,and rectangular detection to obtain data about the size of the walnut.Finally,the real-time data of the size of the detected object transfer to the industrial computer,use Visual Studio 2019 software to realize data communication.The experimental results show that the nut size detection system is stable and can guarantee a high accuracy.