BP神经网络的内场驾考电子教练系统设计

Driving Test Electronic Training System Based on BP Neural Network

针对驾考人员在校考方面训练时间、场地大小的不足以及自考方面缺乏合理指导的问题,设计了基于BP神经网络的内场驾考电子教练系统,该系统以S3C2440微处理器为核心硬件平台,介绍了采集(视觉模块和速度模块)、处理(Canny算子、单目摄像头测距算法和Harris算子)、分析判定(BP神经网络算法)、结果输出(语音模块)的流程化方法。最后,结合各模块算法和C-Free5软件对核心控制算法进行了仿真,得到了满足驾驶考试中触边、启停和夹角分析等相关设定的实验结果。

Aiming at the problems such as the lack of training time and space in school examination and the limitation of reasonable guid- ance in college examination for the self-taught,the driving test electronic training system based on BP neural network is designed, which adopts the S3C2440 microprocessor as the core hardware platform.The procedural approaches are introduced such as the collection（visual module and speed module）, the processing（Canny operator, monocular camera ranging algorithm and Harris operator）, the analysis and judgement （BP neural network algorithm） and the result output（voice module）.Finally,the algorithm for each module and C-Free5 soft- ware are combined to simulate core control algorithm, the results of rim, the starting and stopping as well as the angle analysis in driving test are got.