Self-Adjusted Tracker Based on Genetic Neural-Networks for Tracking Multi-Target
Xiaowei Fu, Kangling Fang, Xi Li · 2005
Neural-networks technique is used to establish a self-adjusted compensator for tracing moving multi-object based on the sampled images. A novel genetic algorithm (NGA) is applied to optimize the weights of neural network rapidly. The algorithm is used for tracking the moving peoples. The results of simulation and experiment are given in the end. The validity of the algorithm is demonstrated.