Abstract—In the paper, a method was proposed to estimate population density based on the gray level co-occurrence matrix. Builded the relation between contrast and population density, and designed a three layers neural network. This paper has used particle swarm algorithm(PSO) to train and optimize the neural network’s connection weights and thresholds, which can overcome the BP algorithm’s limitations such as slow convergence and easy to fall into local minima. The experimental results indicated that this method is fast and can effectively calculate the crowd density. This paper first expounded the research method of crowd population density; then introduced the method of gray level co-occurrence matrix and established the three layer neural network model; the model was optimized by back-propagation (BP) and particle swarm algorithm (PSO); finally, the paper compared The optimization results of BP algorithm with PSO algorithm, And drew the conclusion.
Index Terms—Crowd density, neural network, particle swarm algorithm, gray level co-occurrence matrix.
Xie Lili is with Tian Jin Vocational Institute, Tianjin 300410 China (e-mail: shelly0313@126.com).
Wang Peng is with Hebei University of Technology, Tianjin, 300401, China (e-mail: Wangpeng1027@126.com).
Cite:Xie Lili and Wang Peng, "Method for Estimation of Crowd Density Using Neural Network with PSO Optimization Based on Gray Level Co-Occurrence Matrix," International Journal of Machine Learning and Computing vol.3, no. 6, pp. 520-523, 2013.