基于改进支持向量机的LCD液晶屏气泡分类识别  

Bubble Classification and Recognition of LCD Screen Based on Improved Support Vector Machine

作　　者：王曙敦 贺杰 Wang Shudun;He jie(Shenzhen Jinghua Electronics Co.,Ltd.,Shenzhen 518000,China)

机构地区：[1]深圳晶华电子股份有限公司,深圳518000

出　　处：《山西电子技术》2023年第6期1-4,20,共5页Shanxi Electronic Technology

摘　　要：为提高LCD液晶气泡分类识别正确率,采用量子粒子群算法对支持向量机的惩罚参数和核函数进行优化,建立基于QPSO-SVM的LCD液晶屏气泡分类识别模型,采用LCD液晶屏气泡数据进行算例分析,QPSO-SVM分类器的分类结果的各项指标均优于其他二分类方法,验证本文所提LCD液晶屏气泡分类识别方法的正确性和实用性。In order to improve the accuracy of LCD liquid crystal bubble classification and recognition,quantum particle swarm optimization algorithm is used to optimize the penalty parameters and kernel function of support vector machine,and a LCD liquid crystal bubble classification and recognition model based on QPSO-SVM is established.LCD liquid crystal bubble data is used for example analysis,and all indicators of the classification results of QPSO-SVM classifier are better than other two classification methods,which verifies the correctness and practicability of the bubble classification and recognition method proposed in this paper.

关 键 词：LCD液晶屏 气泡 支持向量机 量子粒子群算法 

分 类 号：TN873[电子电信—信息与通信工程]