一种级联式神经网络框架的路标识别方法研究  被引量:2

Research on road sign recognition method based on a cascaded neural network framework

在线阅读下载全文

作  者:许亮[1] 王亚龙 刘颖 陈曦[1] 翟翔 XU Liang;WANG Ya-long;LIU Ying;CHEN Xi;ZHAI Xiang(Tianjin Key Laboratory for Control Theory Application in Complicated Systems,Tianjin University of Technology,Tianjin 300384,China;National Key Laboratory for Complex Systems Simulation,Beijing 100020,China)

机构地区:[1]天津理工大学天津市复杂系统控制理论与应用重点实验室,电气电子工程学院,天津300384 [2]武器装备体系国防科技重点实验室,北京100020

出  处:《光电子.激光》2020年第3期310-317,共8页Journal of Optoelectronics·Laser

基  金:国家自然科学基金(61975151,61308120);天津市“科技重大专项与工程”(16ZXHLGX00040)资助项目。

摘  要:为了提升无人驾驶汽车对于外界环境感知的能力,本文提出了一种级联式神经网络框架对虚拟环境中的路标进行检测与分类。该框架将添加了辅助结构的全卷积神经网络与改进后的经典LeNet-5网络进行组合,在处理所提取出的路标区域边缘不平整以及产生杂项问题上使用传统的腐蚀膨胀开运算图像处理算子进行优化和解决,实现虚拟道路图像中雨雪等多种情况下的多类路标进行定位与识别。通过与经典的不变矩特征、ORB全局特征提取方法,以及YOLO,SSD人工智能方法对比试验表明,本文所提出方法具备检测准确度高,运算速度快的优势。Automobile assisted driving system can not only reduce the driver′s mental stress during drivingdriving,but also greatly reduce the probability of traffic accidents.In order to improve the ability of driverless vehicles to perceive the external environment,this paper proposes a cascaded neural network framework to detect and classify road signs in virtual environments.The framework combines a full convolutional neural network with an auxiliary structure and a modified classic LeNet-5 network to perform traditional image processing using corrosion processing on the edges of the extracted road sign areas and the generation of miscellaneous problems.The operator is optimized and solved to realize the positioning and recognition of multiple types of road signs in various situations such as rain and snow in virtual road images.Compared with the classical invariant moment feature,ORB global feature extraction method,and YOLO,SSD artificial intelligence method,the proposed method has the advantages of high detection accuracy and fast calculation speed.

关 键 词:目标检测 路标识别 卷积神经网络 深度学习 无人驾驶车 

分 类 号:TP751.1[自动化与计算机技术—检测技术与自动化装置]

 

参考文献:

正在载入数据...

 

二级参考文献:

正在载入数据...

 

耦合文献:

正在载入数据...

 

引证文献:

正在载入数据...

 

二级引证文献:

正在载入数据...

 

同被引文献:

正在载入数据...

 

相关期刊文献:

正在载入数据...

相关的主题
相关的作者对象
相关的机构对象