药品缺陷视觉检测技术的研究  

Study on visual detection technology of drug defects

作　　者：王敏 石明航 吴佳 许永琪 朱永楠 WANG Min;SHI Ming-hang;WU Jia;XU Yong-qi;ZHU Yong-nan(School of Electronics and Information Engineering,Nanjing University of Information Science and Technology,Nanjing 210044,China;Suzhou Minjie Robot Technology Co.,Ltd.,Suzhou 215159,Jiangsu Province,China)

机构地区：[1]南京信息工程大学电子与信息工程学院,南京210044 [2]苏州敏捷机器人科技有限公司,江苏苏州215159

出　　处：《信息技术》2024年第9期8-13,共6页Information Technology

基　　金：国家自然科学基金资助项目(41775165,41775039);南京信息工程大学人才启动经费资助项目(2021r 034)。

摘　　要：针对当前人工目检在药片缺陷检测中效果不理想的情况,提出了一种在准确率、速率和成本上更优的药片缺陷检测方法。该方法基于机器视觉检测技术,使用图像处理、边缘检测、缺陷识别等算法,包含原始图像中值滤波和均值滤波两种噪声消减预处理、hough变换对药品和基于掩膜的背景消除的定位、分析比较canny, prewitt, sobel, roberts等四种不同边缘检测算子对图像的分割效应、标记缺陷所在的同时对缺陷的类型进行识别。最后实验结果表明,该方法的准确率可高达96.67%,效果较好。On the basis of the unsatisfactory effect of visual inspection in pill defect detection,a more accurate,faster and more expensive pill detection method is proposed.The method is based on machine vision detection technology,and the use of image processing,edge detection,defect recognition algorithms,including the original image median filter and mean filter which are two kinds of noise reduction preprocessing,hough transformation which locate the drugs and mask based background elimination,analysis and comparison among canny,prewitt,sobel and roberts and other different edge detection operators’identification to the types of defects while marking the segmentation effect of the image.The experiment results show that the accuracy of this method is up to 96.67%,and the effect is ideal.

关 键 词：图像处理 HOUGH变换 掩膜 边缘检测 分割效应 

分 类 号：TP274[自动化与计算机技术—检测技术与自动化装置]