Method, system, device and medium for determination of car congestion degree based on machine vision

Abstract

The invention discloses a method, system, device and medium for determining the congestion degree of a carriage based on machine vision. The method includes: obtaining first image information in the carriage; obtaining background image information in the carriage, and according to the first image information and the background image information to obtain the foreground image information in the compartment, and then perform image binarization processing, first open and then close operation and filter processing on the foreground image information to obtain the second image information; perform edge detection on the second image information to obtain multiple continuous contours, Then perform random Hough transform on the continuous contours, and select the continuous contours that meet the preset threshold conditions as the head contours; determine the number of passengers in the compartment according to the number of head contours, and then determine the degree of congestion in the compartment according to the number of passengers and the area of ​​the compartment . The invention improves the accuracy and efficiency of compartment congestion detection, improves the space resource utilization rate of subway compartments and the ride comfort of passengers, and can be widely used in the technical field of image processing.

Description

technical field [0001] The present invention relates to the technical field of image processing, in particular to a machine vision-based method, system, device and medium for determining the congestion degree of a carriage. Background technique [0002] As a means of urban rail transportation, the subway can achieve the effect of reducing urban traffic congestion and traffic pollution. It has the advantages of fast running speed, accurate running time, short train intervals, and large passenger capacity. It has become the first choice of most citizens. travel tools. However, since the subway lines in most cities do not completely cover every corner of the city, the passenger flow on the subway lines is relatively large during the morning and evening peak hours, and there will be congestion inside the subway cars and platforms, so there are certain hidden dangers to public safety. In the long-term operation of the subway, it can be found that due to the large number of subwa...

AI Technical Summary

Problems solved by technology

However, since the subway lines in most cities do not completely cover every corner of the city, the passenger flow on the subway lines is relatively large during the morning and evening peak hours, and there will be congestion inside the subway cars and platforms, so there are certain public safety hazards

In the long-term operation of the subway, it can be found that due to the large number of subway cars and the long platform, the door positions of passengers getting on and off the cars are random, and passengers cannot predict the specific number of people in each car in advance, so some The carriages are relatively crowded, and some carriages have vacancies, resulting in uneven distribution of the number of carriages, and the inability to rationally utilize the space resources of the subway carriages, which affects the comfort of passengers

[0003] A related technology discloses a method and device for counting people in real time in a rail transit car, which obtains image frame data through cameras installed in the car, and then inputs the image frame data into the number detection model to obtain the car area captured by each camera However, this method needs to pre-train the number detection model on the one hand, and requires more training samples and longer training time to obtain the number detection model that meets the requirements, which is not efficient. For a long time, there will be occlusion, so the number detection model cannot accurately detect the number of people in the compartment

Images