为了提升药品缺陷检测效率,降低制药企业的生产成本,设计了一款基于计算机视觉的药片实时检测系统.介绍了计算机视觉检测系统的工作原理和硬件结构组成,并在此硬件结构基础上,提出了一种新型计算机图像分割和特征提取方法.实验结果表明,所述算法和系统能够显著提高药片缺陷检测准确度,可以达到99%,计算机视觉检测系统能够完全满足实时性、高识别率等要求.%In order to improve the drug defect detection efficiency and cut down production cost of the pharmaceutical enterprise,this paper designed a real-time tablet detection system based on computer vi-sion.The working principle and hardware structure of the proposed system was introduced,based on which,a novel method for segmenting images and extracting image features was proposed.The experi-mental results show that the proposed algorithm and system can significantly improve the accuracy of tab-let defect detection,reaching 9 9%,and it can fully meet the requirements of real-time and high recogni-tion rate.
展开▼