Passenger flow counting method based on deep learning in vertical visual angle

Abstract

The invention provides a passenger flow counting method based on deep learning in a vertical visual angle. According to the method, passenger flow statistics is performed in the vertical visual angle. Compared with an oblique photographing visual angle, the visual angle realizes easier coping with conditions such as market, supermarket and subway with high passenger density. The method provides head and shoulder detection by means of deep learning. Strong learning capability of deep learning is utilized. Background modeling and foreground extraction on a video are not required, and passenger cutting on the foreground is not required. Head information and shoulder information can be more accurately detected with higher robustness. A deep convolutional characteristic is utilized for performing matched tracking. Compared with manually designed characteristics such as HOG and LBP, the deep convolutional characteristic has better expression capability and can better cope with various scenes. The passenger flow counting method directly use the deep convolutional characteristic of a certain layer for matching, thereby preventing repeated characteristic calculation and realizing high time saving effect.

Description

technical field [0001] The present invention relates to the field of digital image processing, and more specifically, to a passenger flow counting method based on deep learning in a vertical perspective. Background technique [0002] In recent years, video passenger flow counting technology has been a research hotspot that has attracted much attention in the industry. It has also been gradually applied to major shopping malls, chain stores, supermarkets, hotels, airports, subways, scenic spots, etc. The people flow data generated in these scenarios can be used in many fields. Provides valuable information. For major shopping mall chain stores and supermarkets, facing the current hot online e-commerce systems, such as Jingdong, Taobao, Tmall, Amazon, etc., the offline sales market has been crowded, and scientific management is obviously to improve their own effective means of competitiveness. The flow of people data in different time periods and regions in the shopping mall...

AI Technical Summary

Problems solved by technology

But the disadvantage of this feature is that it is easily affected by other objects with similar outlines, such as balloons, etc.

In the prior art, an effective pedestrian counting method is also proposed. For the first time, Hough circle Transform is used for head detection, followed by optical flow (optical flow) method for tracking. However, this method The disadvantage is that optical flow requires a lot of computing power and is difficult to deploy on embedded devices.

In the prior art, it is also proposed to first use Gaussian mixture background modeling to extract the foreground area, and then use the adaptive boosting (Adaptive Boosting) method in the foreground area combined with the local binary pattern (Local Binary Pattern) feature to perform head detection, and finally Connected with the mean shift (meanshift) tracking algorithm for in-and-out statistics, such a system relies on the integrity of the foreground extraction, and it is often difficult to extract a complete foreground, especially when there are dense pedestrians in the picture

Images