基于深度残差网络的道路标志识别模型构建及分析

Road sign recognition model construction and analysis based on deep residual network

道路标志识别是自动驾驶技术的重要依据,自动驾驶技术的高速发展对道路标志识别提出了更高的要求,对道路标志的识别具有重要的理论和应用价值。简单分析了道路标志识别的背景,介绍了卷积神经网络的网络结构和近年来取得较好识别效果的深度残差网络模型(ResNet),并提出了改进的ResNet18网络模型。使用德国道路标志数据集进行训练和测试,并与相关算法进行比较,证明该模型具有较高的识别精度和识别效率。

Road sign recognition is an important basis of autonomous driving technology.The rapid development of automatic driving technology has higher requirements for road sign recognition,which has important theoretical and application value.The background of road sign recognition is briefly analyzed,and the network structure of convolutional neural network is introduced,as well as the deep residual network model(ResNet)which has achieved good recognition effect in recent years.An improved ResNet18 network model is proposed,which is trained and tested by using the German road sign data set,and compared with related algorithms.It is proved that the model has higher recognition accuracy and efficiency.