Method for extracting human body features of robust time-space domain

A technology of human body features and extraction methods, which is applied in the fields of computer vision and pattern recognition, and can solve problems such as large divergence within the class of human body data

Inactive Publication Date: 2013-12-18
NANJING UNIV OF SCI & TECH
View PDF3 Cites 3 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to try to study the human body and its movement patterns from the perspective of joint spatio-temporal features, and propose a spatio-temporal A human body feature extraction method based on multi-level description of appearance information and motion information

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 for extracting human body features of robust time-space domain
  • Method for extracting human body features of robust time-space domain
  • Method for extracting human body features of robust time-space domain

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0056] The overall operation process of the present invention is as figure 1 shown. The specific implementation manners of the present invention will be further described in detail below in conjunction with the accompanying drawings.

[0057] step 1: Input a set of videos containing human body data from which to select a spatiotemporal volume.

[0058] First, select a group of adjacent frames from the video, and the number of frame sequences usually takes 5 to 10 frames. If the number of frame sequences is too small, the motion information will not be obvious when the human body is moving slowly; otherwise, the amount of calculation will be large and the subsequent processing speed will be slow.

[0059] Then, select a rectangular area image of the same size at the same position of each frame image, and the image sequence composed of all rectangular area images is a space-time volume, as shown in Figure 21. The single-frame image coordinates of the spatiotemporal volume ...

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 extracting human body features of a robust time-space domain. The method comprises the steps that a set of video containing human body data is input and a time-space body is selected from the video containing the human body data; the gradient of each frame of an image of the time-space body is calculated so that a gradient time-space body can be obtained; according to the direction quantifying step size and the direction quantifying step size , direction partition is conduced on the gradient time-space body so that channels in different directions can be obtained; according to the space quantifying step size , space partition is conducted on each channel; a feature descriptor of each channel is calculated; the feature descriptors of all the channels are connected in series so that a set of feature descriptors can be formed. According to the different step sizes, the parameters of , and are adjusted and the steps of the method are repeated until the feature descriptors with different description capacities in different groups in a preset number are generated. The method obviously improves the robustness of human body detection so that the problem that the signal to noise ratio is low is solved to a certain degree.

Description

technical field [0001] The present invention relates to the field of computer vision and pattern recognition, more specifically, relates to a human body feature extraction method that comprehensively utilizes appearance information and multi-level description of motion information in time-space domain. Background technique [0002] Human body detection refers to the process of determining the position and size of all human bodies in an input image or video sequence. The computer can further realize the response to human behavior, and finally realize the purpose of human-computer interaction or computer automatic processing and recognition. Effective feature extraction methods can significantly improve the robustness of human detectors and reduce false alarms. [0003] According to the different types of features, it can be divided into grayscale-based features, gradient-based features and methods based on multi-feature fusion. Regarding the differences in the appearance of...

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): G06K9/46
Inventor 刘亚洲张艳孙权森
Owner NANJING UNIV OF SCI & TECH
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