KNN分类算法在停车场车牌识别系统中的应用  

Application of KNN Classification Algorithm in Car Park Automatic License Plate Recognition System

作　　者：屠菁 刘登胜 钟雪景 TU Jing;LIU Dengsheng;ZHONG Xuejing

机构地区：[1]合肥学院人工智能与大数据学院,安徽合肥230601 [2]东南大学软件学院,江苏苏州215123

出　　处：《淮南师范学院学报》2021年第2期143-148,共6页Journal of Huainan Normal University

基　　金：2019年安徽省高等学校自然科学研究项目“轨道交通深基坑监测大数据的状态建模与灾害预警关键技术研究”(KJ2019A0833)。

摘　　要：车牌识别(ALPR)的准确度易受对比度、光照、运动图像等多种因素影响。车牌识别包括预处理、车牌定位、字符分割和识别多个步骤,文章针对分割字符的识别等理论研究成果在实际应用中存在的问题,提出将KNN(K-nearest neighbor)分类算法应用在字符识别中,通过将向量化字符提取特征向量的方法,降低了识别特征的维度和复杂度,在保证检索效率和准确性的同时易于代码实现。将此方法应用在实际停车场管理系统开发中,实验结果表明在车牌定位和车牌识别方面具有较高的可靠性和准确率,提高了车牌识别系统的稳定性和实用性。The accuracy of ALPR is affected by many factors,such as contrast,illumination,moving image,etc.License plate recognition includes several steps,such as preprocessing,license plate location,character segmentation and recognition.In this paper,in view of the difficulty in combining the theoretical research with practical application of the recognition algorithm of segmented characters,the KNN(K-Nearest Neighbor)classification algorithm was applied to character recognition.The method of extracting feature vector from vector characters could reduce the dimension and complexity of the recognition feature,and it was easy to implement in code while ensuring the efficiency and accuracy of retrieval,which could be applied to the development of a practical parking lot management system.The experimental results showed that the method has high reliability and accuracy in license plate location and recognition,and improves the stability and practicability of the license plate recognition system.

关 键 词：车牌识别 车牌定位 KNN分类算法 字符分割 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]