Multi-level supervision crowd counting method and device, medium and electronic equipment

Abstract

The invention provides a multi-level supervision crowd counting method and device, a medium and electronic equipment, and the method comprises the following steps: enabling an image to be sequentiallyinputted into all subordinate feature extraction modules of a feature extraction module 1 through the feature extraction module 1, so as to obtain an output feature map corresponding to each featureextraction module i; fusing the output feature map corresponding to each feature extraction module i and the output feature map corresponding to each subordinate feature extraction module of the feature extraction module i; respectively carrying out density map estimation on each fusion result so as to respectively obtain a sub-density map corresponding to each fusion result; performing joint calculation on the sub-density maps to obtain a final density map; and obtaining the total number of people according to the final density map. The multi-level supervised network adopted by the inventioncan fuse a plurality of density maps to generate a more accurate final density map, so that some errors are avoided, the error influence is reduced, and the total number of people can be better counted.

Description

technical field [0001] The invention relates to the field of image processing, in particular to a multi-level supervised crowd counting method, device, medium and electronic equipment. Background technique [0002] With the increasing urban population, in subway stations, shopping malls, and various festival gatherings, excessive crowds bring potential dangers and easily cause tragedies such as stampedes and riots. In recent years, crowd counting and population density distribution estimation have become hot topics in the security field. At the same time, the algorithm performance of crowd counting has been greatly improved under the combination of deep learning. Therefore, crowd counting and crowd density estimation have important research value. [0003] The purpose of crowd counting is to find out the number of people in a specific scene, while crowd density distribution estimation needs to obtain spatial density information and number of people (summation of density map...

AI Technical Summary

Problems solved by technology

[0009] In view of the above-mentioned shortcomings of the prior art, the object of the present invention is to provide a multi-level supervised crowd counting method, device, medium and electronic equipment, which are used to solve the ineffective and accurate crowd counting in the prior art. estimated problem

Images