Eureka AIR delivers breakthrough ideas for toughest innovation challenges, trusted by R&D personnel around the world.

Abnormal face detection method in front of ATM (automatic teller machine)

A technology of face detection and cash machine, which is applied in the field of abnormal face detection to achieve the effect of improving speed and accuracy and improving practicability

Active Publication Date: 2013-11-20
SHANGHAI JIAO TONG UNIV
View PDF2 Cites 36 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

But their disadvantages: when the contour shape of the head changes greatly, the head cannot be positioned well

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
  • Abnormal face detection method in front of ATM (automatic teller machine)
  • Abnormal face detection method in front of ATM (automatic teller machine)
  • Abnormal face detection method in front of ATM (automatic teller machine)

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0050]This embodiment provides a method for detecting abnormal faces in front of an ATM cash machine. This method is mainly developed under the platform of matlab2010a. The core program is written in matlab. Hybrid programming (malab calls c++) is used in target tracking, based on Gaussian energy The growth algorithm program of the function is written in Matlab. The test used is a video (432×240 pixels, 30fps) in front of an ATM cash machine in a simulated real scene. There are a lot of occluded faces in the scene.

[0051] The specific implementation steps of this embodiment are as follows:

[0052] The first step is to extract accurate foreground contours based on the Gaussian mixture model and the frame difference method. The process is as follows:

[0053] (1) The RGB color model is used for modeling, and each pixel is modeled separately. Each pixel can use up to N Gaussian models to describe their colors in different "states" (according to experience, N is generally sel...

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 provides an abnormal face detection method in front of an ATM (automatic teller machine). The method comprises three steps that the background of a scene is subjected to modeling to obtain the sketchy moving foreground, a money withdrawal person in front of the ATM is positioned by using the rate of the foreground block occupying the whole image and the target position, then, the precise positioning of the foreground block is realized through being combined with a communication domain for analysis, and next, the preprocessing is carried out; the position of the shoulders of the money withdrawal person in the horizontal direction is fast estimated through the statistics of the change of the foreground body width from bottom to top; the growth algorithm based on the Gaussian energy function is used for finding the box tightly covering the head region, the face outline at the edge of the box is utilized, the position of the jaw is determined through random Hough conversion and the least square method head fitting ellipse, and the final accurate position of the head target is obtained through the amendment; after the initial head target position is obtained, the real-time tracking is carried out on the initial head target position on the basis of the L1 tracking operator; and the abnormal face identification is carried out. The abnormal face detection method can be applied to real-time ATM video monitoring and video retrieval.

Description

technical field [0001] The invention relates to a method in the technical field of video image processing and pattern recognition, in particular to an abnormal face detection method applied to a video surveillance system. Background technique [0002] The main task of abnormal face detection in front of the ATM cash machine is to determine whether the behavior of the moving target is abnormal, so as to achieve the purpose of abnormal detection. Abnormal face detection is to further identify specific abnormal faces on the basis of pedestrian target detection, such as detecting the behavior of wearing masks, sunglasses, and peaked caps in a given video. Abnormal face detection technology can be applied to video surveillance, video retrieval and other fields, thereby reducing the burden on operators brought by massive surveillance video data. [0003] A typical pedestrian target detection method is based on background modeling, compares the image information of the current seq...

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
Inventor 张涛刘宁杨杰
Owner SHANGHAI JIAO TONG UNIV
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
Eureka Blog
Learn More
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com