Intensive population counting method and system based on hidden density distribution, and terminal

Abstract

The invention discloses a dense crowd counting method and system based on hidden density distribution and a terminal. The method comprises the steps: obtaining a self-adaptive hidden Gaussian densitygraph through a Gaussian network according to a crowd point graph; according to the counting loss item, the smooth item and the Bayesian item, guiding the hidden Gaussian density graph optimization, so as to enable the generation quality to be higher; taking the hidden Gaussian density map as a training target, combining an adversarial loss function and a Bayesian loss function, and outputting thedense crowd image as a predicted density distribution map; summing the predicted density distribution maps to obtain the predicted density people number. The density predictor, the hidden Gaussian density generator and the discriminator are alternately trained and cooperatively optimized. According to the method, the precision is improved to a large extent, the robustness is good, and the application value is high due to the fact that the parameter quantity and the operand of the inference stage are not increased.

Description

technical field [0001] The present invention relates to the technical field of computer vision, in particular to a dense crowd counting method, system and terminal based on hidden density distribution. Background technique [0002] With the rapid growth of the world's population and the acceleration of urbanization, how to accurately count the crowds under high density to give timely warning, effectively control and guide the flow of people has become a very important hot issue. Most of the existing methods are based on multi-layer convolutional neural networks to extract image features and regress the counting results. [0003] However, in existing methods, the generated density distribution maps often have problems such as low quality, inaccurate prediction of high-density parts, high parameter redundancy, large amount of calculation, and the need to manually adjust hyperparameters according to each scenario, resulting in poor generalization ability. , but in actual scene...

AI Technical Summary

Problems solved by technology

However, the counting accuracy of this patent is low, and it is difficult to solve the challenges brought about by problems such as cross-scenario, cross-scale, and cross-density levels, and it is necessary to manually adjust hyperparameters according to each scenario, which limits its application ability in actual scenarios

Images