Unlock instant, AI-driven research and patent intelligence for your innovation.

A Human Behavior Recognition Algorithm Based on Normalized r-transform Hierarchical Model

A technology of hierarchical models and recognition algorithms, applied in character and pattern recognition, computing, computer parts and other directions, and can solve problems such as not considering hierarchical relationships

Inactive Publication Date: 2016-01-06
SOUTHWEAT UNIV OF SCI & TECH
View PDF2 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

It does not take into account that the recognition process caused by the similarity between behaviors may have a hierarchical relationship that is first coarsely divided and then subdivided

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
  • A Human Behavior Recognition Algorithm Based on Normalized r-transform Hierarchical Model
  • A Human Behavior Recognition Algorithm Based on Normalized r-transform Hierarchical Model

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0024] The present invention will be described in detail below with reference to specific embodiments.

[0025] MEI, normalized R transform, K-Means, star model, distance transform, etc. involved in the present invention are all mature technologies, and have been reported in public literature, and are not repeated in the present invention.

[0026] In order to utilize the discriminative ability and distinguishing ability of multiple features, and at the same time avoid the increase of feature dimension caused by the fusion of multiple features. A hierarchical recognition model of human behavior is proposed. The first layer quickly determines the classification of behavior based on a single feature, and the second layer selects different features (combinations) for different behavior classifications to realize behavior type recognition. This divide-and-conquer hierarchical identification method reduces the dimension of behavior features, and solves the problem of high complexit...

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 human behavior recognition algorithm based on a normalization R transformation hierarchical model. The human behavior recognition algorithm based on the normalization R transformation hierarchical model comprises a behavior training stage and a behavior recognition stage, according to normalization R transformation of manual of engineering instructions (MEI), the behavior training stage completes behavior first layer broad heading division, and then describing and recognition methods of divided various behavior types are respectively confirmed. A Euclidean distance method is utilized by a first layer of the behavior recognition to confirm the behavior belonging large categories, and a second layer of the behavior recognition stage conducts finial behavior recognition according to behavior describing and recognition methods which are formulated in training stages of the large categories. The behavior recognition hierarchical model is concise and high in recognition stage.

Description

technical field [0001] The invention relates to the field of machine vision research, in particular to a human action recognition algorithm based on a normalized R transform layered model. Background technique [0002] Human behavior recognition has a wide range of application value in video surveillance, content-based video retrieval, sports event analysis, human-computer interaction and virtual reality technology, and has become a hot research field in computer vision. Human behavior recognition is usually reduced to a classification problem of time-varying data, including behavior description and behavior classification and recognition. [0003] The two most commonly used features in current behavior understanding research are shape features and motion features. However, there are many kinds of human behaviors, which are complex and changeable. Usually, it is insufficient to use a certain appearance shape feature or motion feature to characterize human behavior. Appeara...

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 Patents(China)
IPC IPC(8): G06K9/66G06K9/00
Inventor 范勇高琳欧阳寒李郁峰李绘卓
Owner SOUTHWEAT UNIV OF SCI & TECH