Train Number Recognition System Based on Image Processing
Railway freight car number is used in railway information management to quickly and accurately judge the im-portant information such as vehicle location and usage.The traditional identification of vehicle number mainly relies on man-ual copying and recording and the use of radio frequency technology,which is easy to make mistakes in the identification of vehicle number,and also consumes a lot of manpower and material resources.Aiming at the above problems,a vehicle number recognition system based on image processing is designed to collect vehicle number information data and con-struct experimental data set.Firstly,MobileNet is used to compress the training model and extract the train number fea-tures.Secondly,YOLOv2 algorithm is used to detect objects in images.Finally,PID algorithm is used to control the two-degree-of-freedom head equipped with a visual recognition system to track the location of the train number,and the iden-tified train number is uploaded to the PC.The design of the train number recognition system is expected to lay a founda-tion for the subsequent upgrading.
image processingvehicle number recognitionautomatic trackingrailway train number
NETL
NSTL
万方数据
