激光与光电子学进展, 2019, 56 (13): 131502, 网络出版: 2019-07-11   

基于支持向量机的输液袋智能检测与缺陷分类 下载: 814次

Intelligent Detection and Defect Classification of Infusion Bags Based on Support Vector Machine
作者单位
沈阳城市建设学院信息与控制工程系, 辽宁 沈阳 110167
摘要
针对目前医疗输液袋印刷过程中存在的漏印、错印等影响医疗品质的问题,提出了一种基于支持向量机的输液袋智能检测与缺陷分类方法。通过对生产过程中常见的医疗输液袋缺陷特征的分析,选取品名偏移、品名旋转和品名污迹作为缺陷分类目标,将候选区与监测区位置关系、候选区与监测区旋转角度和填充度这三种特征作为支持向量机的输入向量训练分类器。实验中使用径向基核函数结合一对一分类法,以平均运算时间和识别准确率作为评价指标进行对比实验,实验结果表明,所提方法识别准确率可达96.7%,满足企业生产的要求。
Abstract
To address the problems of missing and inaccurate prints during the medical infusion bag printing process that impact the medical quality, an intelligent detection and defect classification method based on a support vector machine is proposed for infusion bags. The selected defect classification targets, which are to be classified based on the analysis of the defect characteristics of medical infusion bags during the production process, include the product name offset, product name rotation, and product name stain. These three features, including the location relation between the region of interest and the monitoring region, rotation angle of region of interest and monitoring region, and filling degree are used as the input vectors of the support vector machine to train the classifier. Further, a radial basis function and an one-to-one classification method are used in this experiment. The average operation time and recognition accuracy are considered to be the evaluation criteria for comparing various experiments. The experimental results demonstrate that the recognition accuracy of the proposed method can become 96.7%, satisfying the requirements of commercial production.

李丹, 金媛媛, 童艳, 白国君, 杨明. 基于支持向量机的输液袋智能检测与缺陷分类[J]. 激光与光电子学进展, 2019, 56(13): 131502. Dan Li, Yuanyuan Jin, Yan Tong, Guojun Bai, Ming Yang. Intelligent Detection and Defect Classification of Infusion Bags Based on Support Vector Machine[J]. Laser & Optoelectronics Progress, 2019, 56(13): 131502.

本文已被 5 篇论文引用
被引统计数据来源于中国光学期刊网
引用该论文: TXT   |   EndNote

相关论文

加载中...

关于本站 Cookie 的使用提示

中国光学期刊网使用基于 cookie 的技术来更好地为您提供各项服务,点击此处了解我们的隐私策略。 如您需继续使用本网站,请您授权我们使用本地 cookie 来保存部分信息。
全站搜索
您最值得信赖的光电行业旗舰网络服务平台!