Accurate Statistical Method of Bus Passenger Flow

Abstract

A method for accurate statistics of bus passenger flow. A monocular camera is set above the bus door, and head portraits of passengers getting on and off the bus are collected from the video captured by the camera. The collected positive, Negative samples are used for training, and the images trained by the cascade classifier are stored in the form of xml; the video captured by the monocular camera is framed, and combined with LBP texture features for detection and counting; the cascade classifier detection program is directly called The detection function of opencv obtains the coordinate value of the center point of the passenger's head; at the same time, draws a trapezoidal detection line at the position of the door, and judges the passenger's getting on and off the car based on the position of the starting point of the running track of the center point of the passenger's head within the range of the camera The number of passengers. The starting point of the running trajectory formed by the coordinate points formed by the center point of the passenger's head passes through the trapezoidal detection line and finally runs out of the camera range to easily record the number of passengers getting on and off, simplifying the complexity.

Description

technical field [0001] The invention relates to the technical field of pattern recognition, in particular to a method for accurate statistics of bus passenger flow. Background technique [0002] With the continuous progress of society, there is an urgent need for effective identity detection in the fields of human-computer interaction technology, access control, criminal investigation of public security police departments, personal identification, etc., and the existing detection methods have their own irreparable shortcomings. Retinal detection cannot ensure the safety of the human body. Fingerprint and palmprint detection are neither concealed nor require the active cooperation of the detected person, which cannot meet the needs of practical applications. A new detection method is urgently needed. For public transport, the general bus passenger flow statistics use infrared detection, credit card counting, and coin counting to achieve the technical purpose of passenger flow...

AI Technical Summary

Problems solved by technology

[0002] With the continuous progress of society, there is an urgent need for effective identity detection in the fields of human-computer interaction technology, access control, criminal investigation of public security police departments, personal identification, etc., and the existing detection methods have their own irreparable shortcomings. Retinal detection cannot ensure the safety of the human body, fingerprint and palmprint detection are neither concealed nor need the active cooperation of the detected person, and cannot meet the needs of practical applications. A new detection method is urgently needed

For public transport, the general bus passenger flow statistics use infrared detection, credit card counting, and coin counting to achieve the technical purpose of passenger flow, but relatively speaking, there are still shortcomings such as inaccurate statistics and complicated statistics. For bus passenger flow, There is also an urgent need for a faster way to achieve more accurate passenger flow statistics

Images