People counting method, device, equipment and storage medium based on video images

Abstract

The present invention proposes a method and device for people counting based on video images, wherein the method includes: obtaining the setting parameters of the camera device in the shooting scene; according to the setting parameters, obtaining the imaging weight matrix when the human body is imaged in the shooting scene; The pixel matrix of the image mask of the scene; according to the pixel matrix of the image mask and the imaging weight matrix, the current weighted pixel value of the shooting scene is obtained; according to the mapping relationship between the current weighted pixel value and the preset weighted pixel value and the number of people, Obtain the statistics of the number of people shooting the scene. It can effectively improve the applicability of the method. In addition, the computational complexity caused by the feature detection can be avoided, the system resource requirement can be reduced, and the real-time performance of the system can be improved.

Description

Technical field [0001] The present invention relates to the field of video image processing, and more particularly to a video image based on video image. Background technique [0002] With the continuous development of video image processing technology, you can have real-time statistics on supermarkets, banks, subways such as supermarkets, banks, subways, etc. through video surveillance, and obtain statistics such as passenger flow distribution, crowds, and provide effective public regional management, resource scheduling. References. [0003] In the prior art, by extracting the characteristics of the pedestrians in the image captured in the image device, for example, the head shoulder shape characteristics, human face characteristics, pedestrian direction gradient histogram (HOG) feature, etc. feature. Specifically, the extraction of the characteristics can be achieved by a machine learning method, such as a venoming neural network training line portrait detector, and then input...

AI Technical Summary

Problems solved by technology

[0004] However, in different shooting scenarios, the features selected by machine learning methods are quite different, which may lead to the inability to recognize portraits. When there are key features, missed detection may occur

Therefore, when the shooting scene changes, it is necessary to re-extract samples and retrain the pedestrian image detector, so that the system can work normally in the new shooting scene, resulting in a large workload

In addition, the computational complexity of feature detection is large, and it is necessary to extract the region of interest (Region of Interest, referred to as ROI), run the feature extraction algorithm and the classification algorithm, and the demand for computing resources is large, so it is difficult to meet the real-time requirements

Images