【摘要】 基于计算机视觉的交通标志检测与识别是智能交通中重要的一部分,准确检测将有利于安全驾驶。当前复杂的道路状况使得交通标志的检测和识别很困难,而且车辆在行驶过程中不稳定以及光照的变化等,给交通标志检测带来了很大的问题,给交通标志识别的准确性和快速性也带来了极大的挑战。本文通过结合颜色和形状特征来检测交通标志,针对识别问题,构建了一个VGG-8卷积模型,而且在交通信号数据集上进行了测试,该模型有很高的准确性,对于解决实际问题有一定的可行性。更多还原

【Abstract】 The detection and identification of traffic signs based on computer vision is one of the most important parts of Intelligent transportation systerm(ITS) to promote the development of the safe driving. The difficulty to detect and identify the traffic signs for complicated road conditions as well as the instability of the vehicle during driving and the changes of illumination, brings great problems to accurate and rapid traffic sign detection and recognition. The color and shape features are combined to detect traffic signs, and a VGG-8 convolution model is established for identification problems. The testing results performed on German traffic sign recognition benchmark(GTSRB) show that high accuracy can be achieved and the feasibility can be used for solving practical problems.更多还原