People counting method based on video analysis

Abstract

The invention discloses a people counting method based on video analysis. The method comprises the following steps of inputting a video image, acquiring a foreground picture by using a background subtraction method, and clustering the foreground picture into a plurality of blocks; extracting features, and performing perspective correction on the features; estimating the number of people by using a two-layer regression model, wherein the first-layer regression divides the different blocks in each frame into different disperse density layers, and the second regression combines virtual standardization and regression counting as a joint learning process; training a counting model synchronously covering virtual standardization and regression problems for the different disperse density layers; and at last counting each block by using the different regression models according to the different crowd density layers, and accumulating the numbers of all blocks to obtain the number of people. According to the people counting method, based on the two-layer regression model, and through combining the vision normalization with the number regression, the defect of a single regression model is overcome, and better robustness and adaptability are provided for shielding multi-density crowd scenes and crowds and segmenting incomplete images.

Description

technical field [0001] The present invention relates to the field of computer vision, and more specifically, to a crowd counting method based on video analysis. Background technique [0002] With the development of the economy, large-scale crowd activities are becoming more and more frequent, and the high crowding of the crowd is likely to cause various emergencies. Therefore, it is very necessary to estimate the number of people in public places. However, the traditional manual counting method is not only time-consuming, laborious and expensive, but also cannot guarantee the counting accuracy. Therefore, it is of great practical significance to develop an intelligent real-time monitoring crowd counting system. [0003] At present, the research work on crowd counting can be mainly divided into two categories: methods based on individual tracking and methods based on regression. The basic idea of ​​the individual-based tracking method is to detect and track people as individ...

AI Technical Summary

Problems solved by technology

The current regression method generally separates the visual standardization and regression estimation. When performing visual standardization, the traditional method is usually used to set the perspective standardization weight (such as assuming that the edge features are linearly related to the size of the crowd), but this setting is not necessarily all feasible

Secondly, this method of separately performing perspective standardization and regression ignores the local nonlinear relationship between the visual normalization weight and the crowd size, and the effect is not very ideal due to the influence of segmentation and occlusion.

Images