基于深度学习的计算机视觉技术在交通场景中的应用

Application of computer vision technology based on deep learning in traffic scene

基于深度学习的计算机视觉技术实现了采用YOLO V3神经网络模型对视频中交通目标物体进行检测、识别与分析。实现车牌识别、路口饱和度计算、机动车违章判断等功能。结果表明,检测结果准确度高、检测速度较快,达到了交通自动化监管的目的。

Based on the computer vision technology of deep learning,this paper realizes the system of detecting,recognizing and analyzing the target object in video by using Yolo V3 neural network model.This page realizes the functions of license plate recognition,intersection saturation calculation,motor vehicle violation judgment and so on.It is proved by the experiment that the design is practicable,accurate and fast,and that it can contribute to automizing the management of traffic.