当前位置: 首页 > 文章 > 基于支持向量机的交通标志人工智能检测与识别 山东农业大学学报(自然科学版) 2017 (3) 400-404
Position: Home > Articles > The Intelligent Detection and Recognition for Traffic Signs Based on Support Vector Machine Journal of Shandong Agricultural University(Natural Science Edition) 2017 (3) 400-404

基于支持向量机的交通标志人工智能检测与识别

作  者:
何诚刚
单  位:
西安交通大学城市学院
关键词:
支持向量机;交通标志;智能检测;识别
摘  要:
针对人工智能检测与识别交通标志准确率不高的问题,本文提出了一种以支持向量机(SVM)为基础的多方法相融合的交通标志检测与识别方法。该方法首先采用方向梯度直方图进行交通标志的特征数据提取,然后利用网格搜索法和交叉验证方法对支持向量机模型最优化参数组合(惩罚因子C和核参数r)进行搜索,最后利用优化的支持向量机模型现实交通标志识别。实验仿真结果表明:基于支持向量机的最优化交通标志识别的准确率可达98%。
译  名:
The Intelligent Detection and Recognition for Traffic Signs Based on Support Vector Machine
作  者:
HE Cheng-gang;Xi'an Jiaotong University City College;
关键词:
Support vector machine;;traffic signs;;intelligent detection;;recognition
摘  要:
To solve the problem of low accuracy of intelligent detection and recognition for traffic signs, a fused multi-phase method based on support vector machine theory was put forward for traffic signs detection and identification. Firstly, the histogram of oriented gradient(HOG) was used to extract the feature data of traffic signs. Then the grid search method and cross validation method were applied to search the optimal parameters combination(penalty factor C and kernel parameter r)of the support vector machine model. Finally, the optimal method of support vector machine was applied to identify the traffic signs. Experimental results indicated that the proposed method went up to accuracy 98% of the recognition for traffic signs.

相似文章

计量
文章访问数: 15
HTML全文浏览量: 0
PDF下载量: 0

所属期刊

推荐期刊