Road Target Classification Method Based on GA-BP Neural Network for Event Camera
Xuning Mao, Haiqing Liu, Miaomiao Zhang, Guang Liu · 2021 IEEE 4th International Conference on Automation, Electronics and Electrical Engineering (AUTEEE) · 2021
Vehicle classification is an important technology of intelligent transportation system, and most methods are based on video streams currently. However, video detectors have some limitations in extreme illumination or high-speed motion. In our study, the event camera is installed in the actual road scenario for vehicle and pedestrian detection. The geometrical, quantitative and Gaussian projection characteristics of the point cloud targets are extracted. The BP neural network is used for vehicle and pedestrian classification, and we optimize the weight and threshold values of the network by GA (genetic algorithm). The accuracy of classification reaches 95%.