一种融合加权ELM和AdaBoost的交通标志识别算法  

Traffic Sign Recognition Algorithm Combining Weighted ELM and AdaBoost

作　　者：徐岩[1] 王权威 韦镇余 

机构地区：[1]天津大学电子信息工程学院,天津300072

出　　处：《小型微型计算机系统》2017年第9期2028-2032,共5页Journal of Chinese Computer Systems

基　　金：国家自然科学基金项目(61372145)资助;天津大学自主创新基金项目(2015XZC-0005)资助

摘　　要：作为一种新型的单隐层前馈型神经网络,极限学习机(Extreme Learning Machine:ELM)相比于传统的神经网络学习算法具有参数设置少、泛化性能强、训练和识别速度快等优点.为了有效提高交通标志的识别速度和识别率,提出一种基于加权ELM和AdaBoost融合优化的交通标志识别新算法.该算法通过迭代更新原始ELM的训练权重,并利用加权后的ELM作为AdaBoost的弱分类器,最终通过加权多数表决得到最优强分类器.最终实验结果表明,该算法能够取得的交通标志总识别率为99.12%,且单张交通标志的识别时间为7.1ms,可以满足实时识别应用的需求,较好的改善了交通标志的识别性能.As a novel singlehidden layer feedforward neural network,ELM （Extreme Learning Machine） has the advantages of less parameter settings,better generation performance,faster learning and recognition speed compared with the traditional neural network algorithms.In order to improve the recognition speed and accuracy of traffic signs effectively,this paper presents a novel and efficient traffic sign recognition algorithm based on fusion optimization of weighted ELM and AdaBoost.The algorithm employs the weighted ELM as the weak classifier by updating the training weight of original ELM iteratively.Finally,an optimal strong classifier is constructed by the weighted majority vote of all the weighted ELMs.The final experimental results show that the proposed algorithm can achieve a total traffic sign recognition accuracy of 99.12%,and only need 7.1ms to recognize a single traffic sign which can meet the demand of real-time recognition applications.Therefore,the proposed algorithm efficiently improves the recognition performance of traffic signs.

关 键 词：极限学习机 权重矩阵 最优强分类器 交通标志识别 ADABOOST 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]