Crowd number determination method and device, equipment and storage medium

Abstract

The invention discloses a crowd number determination method and device, equipment and a storage medium, and relates to the technical field of image processing. The method comprises the following steps: acquiring a first image; performing feature extraction based on the first image to obtain a crowd density feature map corresponding to the first image; based on the at least two people number classification intervals, performing classification processing on the crowd density feature map; based on the sub-intervals corresponding to the sub-regions in the first image in the at least two people number classification intervals, obtaining people number information corresponding to the sub-regions in the first image; and obtaining the number of people corresponding to the first image based on thenumber information of people corresponding to each subarea in the first image. The scheme can be applied to the field of intelligent traffic, the number of people in each area of the first image is estimated according to the classification condition of the first image in the two people classification intervals with different sub-intervals, the accuracy of estimating the number of people in the image is improved, and the accuracy of intelligent traffic scheduling is improved.

Description

technical field [0001] The present application relates to the field of computer vision, and in particular to a method, device, equipment and storage medium for determining the number of people. Background technique [0002] Crowd density estimation, an application that automatically infers the total number of people in an image, plays an important role in video surveillance, public safety, and more. [0003] In related technologies, the traditional methods based on detection and direct regression, most of the crowd density estimation algorithms at this stage combine heat map regression, and use deep learning technology for end-to-end training and reasoning, which can better solve the crowd density distribution range. The problem of large scale and large variation range of head scale improves the counting accuracy to a certain extent. And in the process of identifying the number of people in the image, the image blocks can be divided into different categories according to th...

AI Technical Summary

Problems solved by technology

[0004] In the above scheme, when the image blocks are divided into different categories, the predicted number of people will be uniformly set as the proxy count value of the corresponding interval, and the discretization error is large, which affects the accuracy of predicting the number of people in the image

Images