Method and system for predicting people crowdedness

A prediction method and technology for human flow, applied in the computer field, can solve problems such as error-prone, large workload of manual analysis of video data, etc., and achieve the effects of enhancing adaptability, improving system computing efficiency, and improving the accuracy of prediction results.

Inactive Publication Date: 2017-07-11
CHONGQING UNIV OF ARTS & SCI
View PDF4 Cites 9 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

After obtaining a large amount of video data through the camera, the traffic statistics data in the video data can be accurately extracted; however, the workl

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Method and system for predicting people crowdedness
  • Method and system for predicting people crowdedness
  • Method and system for predicting people crowdedness

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0066] As shown in the figure, the crowd flow congestion prediction method provided by this embodiment includes the following steps:

[0067] Obtain the image frame sequence of the area to be tested in real time through the camera module;

[0068] Input the picture frame sequence to the pedestrian detection module;

[0069] Through the pedestrian detection module, the pedestrian detector based on convolutional neural network and support vector machine trained offline is used to detect the face of the image frame sequence input by the camera module to obtain the face image feature vector;

[0070] Through the face image feature vectors of two adjacent frames, the full connection layer of the convolutional neural network is used to process and output the number of pedestrians;

[0071] According to the number of pedestrians, the pedestrian flow density is calculated according to the following formula:

[0072]

[0073] where x i is the number of pedestrians appearing in th...

Embodiment 2

[0137] The pedestrian detection process provided in this embodiment: the model used in the pedestrian detection module is a pedestrian detector based on convolutional neural network + support vector machine that is trained offline, and the convolutional neural network uses different rectangular frame sizes for the image frames collected by the camera , different vertical and horizontal displacements are scanned to extract feature vectors, and different rectangular frame feature vectors are sent to the support vector machine to make a binary classification judgment of whether they are pedestrians, regression merges face frames at the same position, and counts the number of faces appearing in a frame of pictures ;Continuously count the number of faces in the picture frame per unit time, and calculate the crowd density value per unit time according to the formula of crowd density. The specific convolutional neural network can be selected from the Google network, and the BN layer i...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention discloses a method for predicting people crowdedness and the method comprises: first, obtaining the image frame sequence of a to-be-measured area; inputting the image frame sequence into a pedestrian detection module; through the pedestrian detection module, obtaining the characteristic vectors of the human face images; through the characteristic vectors of the human face images of two adjacent frames, obtaining and outputting the number of the traveling people; based on the number of the traveling people, calculating the people density; and inputting the people density into a people crowdedness predicting module to calculate and output the people crowdedness predicting vector signal. According to the method and the system of the invention, a support vector machine is utilized to train the predicting model, which increases the computation efficiency of the system and the precision of the predicting result so that the method and the system become more adaptable to the data and the environment. In addition, as the method and the system are equipped with a people crowdedness training module and a historical database, the system is empowered with the ability for self-adaptive learning, and therefore, the manual involvement is reduced, the human resource and use difficulty can be saved, making the method and system adaptable to complex and diverse environments. Along with the increase in the learning time and the enlargement of the historical data, the people crowdedness predicting model can achieve more precise predicting results, and the cost in doing so is also reduced.

Description

technical field [0001] The invention relates to the field of computer technology, in particular to a method and system for predicting crowd flow congestion. Background technique [0002] In public scenes and traffic control, it is often necessary to determine the pedestrian flow in the space area to be detected, which is generally convenient for real-time control of crosswalk signal lights, or to ensure the safety of public places. In recent years, a large number of surveillance cameras have been added in cities across the country. After obtaining a large amount of video data through the camera, the traffic statistics data in the video data can be accurately extracted; however, the workload of manual analysis of video data is very large and error-prone, so intelligent video surveillance has become the focus of domestic and foreign researchers. Among them, pedestrian flow counting has become a very important research field in intelligent video surveillance because of its wide...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
IPC IPC(8): G06K9/00G06K9/62
CPCG06V40/161G06V40/172G06V40/168G06V40/10G06F18/214G06F18/2411
Inventor 齐逸
Owner CHONGQING UNIV OF ARTS & SCI
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap