基于神经网络的汽车动态称重CCD抓拍双核系统设计

Dual-system design of vehicles' dynamic weighing and CCD capture based on BP neural network

介绍了基于TMS320C5410DSP和MSP430F149单片机相结合的汽车动态称重双核系统。该系统采用了BP神经网络改进算法Levenberg-Marquardt法实现精确称重,具有对超重汽车进行CCD抓拍,图像车牌识别等功能。仿真结果表明动态称重误差小于0.1%。

This paper introduced a vehicles dynamic weighing dual-core system based on the combination of MSP430F149 MCU and TMS320C5410 DSP.BP neural network was adopted to complete the precise weighing.The system had the function of realizing the capture of overweighing vehicles and recognition of their images and license plates.The emulation presents that the error of dynamic vehicles＇ weighing is under 0.1%.