基于NI Vision Assistant的樱桃缺陷检测方法  

Detection of Cherry Defects Based on NI Vision Assistant

作　　者：王昭[1] 程朋乐[1] 皇甫宜龙 古宏涛 张京善 

机构地区：[1]北京林业大学,北京 [2]北京物资学院,北京

出　　处：《计算机科学与应用》2017年第12期1163-1173,共11页Computer Science and Application

基　　金：通州区西集镇沙古堆大樱桃专业合作社特色农产品物流系统研究与示范应用项目编号:KJ2017CX016。

摘　　要：为了改善我国樱桃分选仍然主要由人工作业的现状,实现樱桃缺陷果自动筛选,提高樱桃生产效率。本实验结合现有的图像处理方法,设计了一套基于视觉处理软件NI Vision Assistant的樱桃缺陷检测装置。实验中首先利用图像采集系统获取樱桃图像,然后利用算法提取樱桃缺陷部分,最后将缺陷果位置信息发送给筛选机构,从而完成樱桃的缺陷检测与筛选。最后的实验结果表明,此筛选方法速度快,实验器械和方法较为简单,综合检测正确率也比较高,基本满足实际生产中缺陷樱桃的检测要求。In order to improve the status of cherry sorting in China, the automatic screening of cherry fruit defects and the improvement of cherry production efficiency were mainly carried out. Combined with the existing image processing methods, this paper designed a set of cherry defect detection device based on visual processing software NI Vision Assistant. In this experiment, using image acquisition system to obtain the image and then the algorithm to extract the cherry fruit defect part;finally the defect location information is sent to the screening mechanism, so as to complete the detection and screening of defects of cherry. Finally, the experimental results show that the screening method is fast;the experimental instruments and methods are relatively simple;and the comprehensive detection accuracy is relatively high, which basically meets the requirements of the detection of defective cherries in actual production.

关 键 词：机器视觉 图像处理 缺陷检测 樱桃 

分 类 号：TP39[自动化与计算机技术—计算机应用技术]