METHOD AND SYSTEM FOR CALCULATING PASSENGER CROWDEDNESS DEGREE

• US 20170286780A1
• Filed: 06/20/2017
• Published: 10/05/2017
• Est. Priority Date: 03/18/2016
• Status: Active Grant

• Current Assignee: Shenzhen University (Guangdong Education Department)
• Original Assignee: Shenzhen University (Guangdong Education Department)
• Inventors: Zhang, Yong, Liu, Lei, Zhao, Dongning, Li, Yanshan, Chen, Jianyong
• Granted Patent: US 10,223,597 B2
• Time in Patent Office: Days
• CPC Class Codes: G06F 18/2148 characterised by the process organisation or structure, e.g. boosting cascade
• CPC Class Codes: G06T 2207/30201 Face
• CPC Class Codes: G06T 2207/30242 Counting objects in image
• CPC Class Codes: G06T 2207/30252 Vehicle exterior; Vicinity of vehicle
• CPC Class Codes: G06T 7/215 Motion-based segmentation
• CPC Class Codes: G06V 10/446 using Haar-like filters, e.g. using integral image techniques
• CPC Class Codes: G06V 10/62 relating to a temporal dimension, e.g. time-based feature extraction; Pattern tracking
• CPC Class Codes: G06V 10/7747 Organisation of the process, e.g. bagging or boosting
• CPC Class Codes: G06V 20/53 Recognition of crowd images, e.g. recognition of crowd congestion
• CPC Class Codes: G06V 20/59 inside of a vehicle, e.g. relating to seat occupancy, driver state or inner lighting conditions

Abstract

The disclosure provides a method for calculating a passenger crowdedness degree, comprising: establishing a video data collection environment and starting collecting video data of passengers getting on and off; reading the collected video data of passengers getting on and off and pre-processing a plurality of successive image frames of the video data; identifying a human head according to the pre-processing result and taking the detected human head as a target object to be tracked by mean-shift; and judging the behaviours of getting on and off of a passenger in the area where the target object is positioned and determining the crowdedness degree of passengers inside a vehicle according to the numbers of the passengers getting on and off. The disclosure also provides a system for calculating a passenger crowdedness degree. The disclosure can effectively reduce the false detection, leak detection and error detection of the head top.

Claims

1. A method for calculating a passenger crowdedness degree, characterized in that the method comprises:

establishing a video data collection environment and starting collecting video data of passengers getting on and off;

reading the collected video data of passengers getting on and off and pre-processing a plurality of successive image frames of the video data;

identifying a human head according to the pre-processing result and taking the detected human head as a target object to be tracked by mean-shift; and

judging the behaviours of getting on and off of a passenger in the area where the target object is positioned and determining the crowdedness degree of passengers inside a vehicle according to the numbers of passengers getting on and off.

2. The method for calculating a passenger crowdedness degree according to claim 1, characterized in that the step of reading the collected video data of passengers getting on and off and pre-processing a plurality of successive image frames of the video data specifically comprises:

reading the frame format of the collected video data of passengers getting on and off and determining the number of frames;

establishing a single Gaussian model for each pixel point in an initial frame;

analyzing the change of pixel points of a plurality of successive image frames and judging each pixel point in the image frame to be static background or dynamic foreground;

modifying the pixel value of the pixel point judged to be static background and not modifying the pixel value of the pixel point judged to be dynamic foreground; and

recording the number of times each pixel point in the image frame is judged to be static background continuously;

if the number of times is greater than or equal to a preset threshold, immediately updating the RGB value of this pixel point to background; and

if the pixel point is not judged to be static background continuously, recording again.

3. The method for calculating a passenger crowdedness degree according to claim 1, characterized in that the step of identifying a human head according to the pre-processing result and taking the detected human head as a target object to be tracked by mean-shift specifically comprises:

producing a cascade classifier which is configured to judge head top, by utilizing a plurality of collected positive samples containing human heads and negative samples not containing human heads;

limiting the size range of a detection window detecting head top in the cascade classifier;

identifying human heads according to the cascade classifier with the detection window size limited; and

taking the detected human head as a target object to be tracked by mean-shift.

4. The method for calculating a passenger crowdedness degree according to claim 1, characterized in that the step of judging the behaviours of getting on and off of a passenger in the area where the target object is positioned and determining the crowdedness degree of passengers inside a vehicle according to the numbers of passengers getting on and off specifically comprises:

setting two detection lines in an area seen by a camera, and, if the mass centre of the target object to be tracked by mean-shift passes through the two detection lines, judging that the passenger is getting on or off; and

determining the total number of passengers actually carried by calculating the numbers of passengers getting on and off, and measuring the crowdedness degree of passengers inside the vehicle using the ratio of the total number of passengers to the maximum passenger capacity inside the vehicle.

5. A system for calculating a passenger crowdedness degree, characterized in that the system comprises:

a data collection module, which is configured to establish a video data collection environment and start collecting video data of passengers getting on and off;

a pre-processing module, which is configured to read the collected video data of passengers getting on and off and pre-process a plurality of successive image frames of the video data;

an object determination module, which is configured to identify a human head according to the pre-processing result and take the detected human head as a target object to be tracked by mean-shift; and

a crowdedness judgment module, which is configured to judge the behaviours of getting on and off of a passenger in the area where the target object is positioned and determine the crowdedness degree of passengers inside a vehicle according to the numbers of passengers getting on and off.

6. The system for calculating a passenger crowdedness degree according to claim 5, characterized in that the pre-processing module comprises:

a frame reading sub-module, which is configured to read the frame format of the collected video data of passengers getting on and off and determine the number of frames;

a model establishment sub-module, which is configured to establish a single Gaussian model for each pixel point in an initial frame;

a state sub-module, which is configured to analyze the change of pixel points of a plurality of successive image frames and judge each pixel point in the image frame to be static background or dynamic foreground;

a modification sub-module, which is configured to modify the pixel value of the pixel point judged to be static background and not modify the pixel value of the pixel point judged to be dynamic foreground; and

an update sub-module, which is configured to;

record the number of times each pixel point in the image frame is judged to be static background continuously;

if the number of times is greater than or equal to a preset threshold, immediately update the RGB value of this pixel point to background; and

if the pixel point is not judged to be static background continuously, record again.

7. The system for calculating a passenger crowdedness degree according to claim 6, characterized in that the object determination module comprises:

a production sub-module, which is configured to produce a cascade classifier which is configured to judge head top, by utilizing a plurality of collected positive samples containing human heads and negative samples not containing human heads;

a limit sub-module, which is configured to limit the size range of a detection window detecting head top in the cascade classifier;

an identification sub-module, which is configured to identify human heads according to the cascade classifier with the detection window size limited; and

a target sub-module, which is configured to take the detected human head as a target object to be tracked by mean-shift.

8. The system for calculating a passenger crowdedness degree according to claim 7, characterized in that the crowdedness judgment module comprises:

a first judgment sub-module, which is configured to set two detection lines in an area seen by a camera, and, if the mass centre of the target object to be tracked by mean-shift passes through the two detection lines, judge that the passenger is getting on or off; and

a second judgment sub-module, which is configured to determine the total number of passengers actually carried by calculating the numbers of passengers getting on and off, and measure the crowdedness degree of passengers inside the vehicle using the ratio of the total number of passengers to the maximum passenger capacity inside the vehicle.