The application of video image recognition technology to identify and count the moving targets in traffic scenes has become a research hotspot in the field of intelligent transportation. In order to identify multiple moving targets in traffic scenes accurately, this paper proposes a pedestrian target intelligent recognition method that combines neural networks and genetic algorithms. It uses BP feedforward neural networks to achieve target classification and recognition, and uses a hierarchical genetic algorithm (HGA) with global search capabilities to optimize the structure of neural network. It solves the shortcomings of BP algorithm, such as easy to fall into local minimum, slow convergence, weak global search ability, difficult to determine network structure, etc., thereby improving the effectiveness and accuracy of pedestrian recognition. The results show that the method can significantly distinguish pedestrians from other negative moving targets, accurately count the number of pedestrian targets in the entire traffic scene, and achieve better results in pedestrian recognition in dynamic scenes.

CICTP