Behavior identification method through combination of confidence and contribution degree on the basis of space-time context

A technology of spatiotemporal context, recognition methods, applied in character and pattern recognition, instrumentation, computing, etc.

Inactive Publication Date: 2016-06-01
BEIJING JIAOTONG UNIV +1
View PDF3 Cites 8 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Although, the model is insensitive to perspective and scale changes and does not need to track the human body, it has its own limitations

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
  • Behavior identification method through combination of confidence and contribution degree on the basis of space-time context
  • Behavior identification method through combination of confidence and contribution degree on the basis of space-time context
  • Behavior identification method through combination of confidence and contribution degree on the basis of space-time context

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0024] The present invention will be described below in conjunction with the accompanying drawings and specific embodiments.

[0025] The spatio-temporal interest points involved in the present invention, HOG / HOF descriptors, BOW model, K-Means clustering, K-nearest neighbor classification, etc. are all mature technologies, all of which have been published in open literature, and will not be repeated in the present invention.

[0026] refer to figure 1 , at step S 5 In mining the confidence of each word from the association rules, in order to obtain the potential association relationship within the feature, the statistical method is used to mine the association rules between each word and each behavior. an association rule Indicates such a pattern that when X happens, Y also happens at the same time. If X is a word set and Y is a behavior set, then this association rule can describe the association between words and behaviors. The confidence and support of association rul...

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 present invention provides a behavior identification method through combination of confidence and contribution degree on the basis of a space-time context. The deficiency of human body behavior identification through adoption of classic local characteristics and a word bag model is made up. The behavior identification method describes the context interaction information of the local characteristics in the spatial domain and the time domain and considers the relationship between the characteristics and the behavior. The method comprises: excavating the confidence of words from association rules; learning the contribution degree of the words from a linear SVM; calculating an association weight of the words and the corresponding behavior through combination of the confidence and the contribution degree; calculating context association coefficients according to the space-time interaction relationship and the association weight; accumulating the association coefficients of all the points in a neighborhood to obtain a local context descriptor; and generating a characteristic sequence for behavior classification through adoption of a context descriptor cumulative histogram. Through adoption of the machine learning and data excavation correlation technology, the behavior identification method through combination of confidence and contribution degree on the basis of a space-time context is able to allow a human body behavior identification algorithm to have the discrimination capability and the robustness.

Description

technical field [0001] The invention relates to the fields of content-based video analysis, intelligent monitoring and human behavior recognition, in particular to a behavior recognition method based on spatio-temporal context combined with confidence and contribution. Background technique [0002] Human behavior analysis based on vision has broad application prospects, because more than 80% of the information obtained by the human body is visual information. Faced with such a huge amount of information, manual processing is obviously impractical, so there is an urgent need for research and development that can replace manpower. computer-related abilities. Human action recognition has a wide range of applications in video surveillance, content-based retrieval and human-computer interaction, and has become a popular research field in computer vision. Behavior analysis has two key issues: behavior description and behavior recognition. Behavior description is to express behavi...

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
Patent Type & Authority Applications(China)
IPC IPC(8): G06K9/62
Inventor 苗振江许万茹张强刘汝杰
Owner BEIJING JIAOTONG 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
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