Subway people flow counting method based on scene flow

Abstract

The invention provides a subway people flow counting method based on scene flow, and belongs to the field of computer vision. The method includes: using RGB and TOF sensors to obtain RGB and depth image sequences when passengers get on and get off the bus; establishing a data item and introducing a smooth item, and constructing a scene flow energy functional; establishing an Euler-Lagrange equation to solve a scene flow; taking the calculated scene flow as a feature, performing moving target detection by adopting an ISODATA clustering analysis algorithm, and extracting a moving region; carrying out passenger head extraction in the motion area according to the passenger head features; achieving passenger target tracking according to the passenger head positions and the scene flow information detected by two adjacent frames; setting two detection lines in the image, judging get-on and get-off behaviors of the passenger if the passenger target passes through the detection lines in the tracking process, and updating a counting result; counting the number of passengers getting on and off the train at the stop according to the counting result of each train door, and obtaining the numberof passengers currently carried by the train.

Description

technical field [0001] The invention relates to a method for counting subway people flow based on scene flow, which belongs to the field of computer vision. Background technique [0002] With the continuous acceleration of urban development in my country and the gradual increase of urban population density, the subway, as a fast and convenient mode of transportation, is favored by the travel crowd, but it also creates certain safety hazards, such as crowding and stampede incidents during peak passenger flow . In recent years, as people pay more and more attention to the safety of rail transit, how to make statistics on passenger flow during train running has become particularly important. On the one hand, this data can provide an important basis for the construction of urban rail transit, and help subway operators formulate scientific operation management, passenger flow planning, and vehicle scheduling; Respond to incidents in a timely and effective manner. [0003] At pr...

AI Technical Summary

Problems solved by technology

[0007] The RGB binocular camera needs to calculate the depth information through stereo matching, which takes a long time to calculate; in addition, under the conditions of dark light or overexposure, and the lack of texture of the measured scene, the stereo matching effect is poor, resulting in inaccurate depth information , the depth map is incomplete

Images