Supercharge Your Innovation With Domain-Expert AI Agents!

Machine vision human body abnormal behavior recognition method based on multi-feature fusion

A multi-feature fusion and machine vision technology, applied in character and pattern recognition, acquisition/recognition of facial features, instruments, etc., can solve problems such as poor effect and poor robustness

Active Publication Date: 2019-12-31
BEIJING UNIV OF TECH
View PDF3 Cites 8 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

In complex real-world scenarios, the robustness is not good, and the effect is poor, and it has not been well applied.

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
  • Machine vision human body abnormal behavior recognition method based on multi-feature fusion
  • Machine vision human body abnormal behavior recognition method based on multi-feature fusion
  • Machine vision human body abnormal behavior recognition method based on multi-feature fusion

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0073] 1. Facial expression, attribute detection

[0074] The first is to detect the face in the video. The face detection uses the SFace algorithm. This algorithm designs two branches, Anchor-based and Anchor-free. Both branches use IoU Loss as Regression Loss. This adjustment It helps to unify the output methods of the two branches, optimize the combined results, and solve the multi-scale problem of faces to a certain extent.

[0075] Then, the detected face is detected by facial expression and facial attributes, and a multi-task network is designed for facial expression detection and attribute recognition, such as figure 2 As shown, the input of the model is a human face, and the features are extracted through the deep convolutional neural network. Considering the real-time requirements, Backbone uses the shuffleNetV2 network; at the same time, the trained model is compressed, that is, some parameters are removed. Convolution kernel, because these convolution kernels are ...

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 machine vision human body abnormal behavior recognition method based on multi-feature fusion. The method comprises face attribute detection, expression analysis, posture analysis and human body abnormal behavior analysis. The method comprises the following steps: firstly, performing face detection on pedestrians in a video, normalizing the detected faces, and inputting the normalized faces into a face attribute and expression detection model to obtain attributes and facial expressions of the pedestrians; performing human skeleton key point detection on pedestrians inthe video to obtain human skeleton position information; finally, fusing pedestrian attributes, the facial expression and posture features by using the feature fusion method provided by the invention,inputting the fused data into a human body abnormal behavior analysis model to analyze the abnormal behavior of the pedestrian, wherein the human body abnormal behavior analysis model is designed byadopting the proposed thought of grouping cross transfer. The method has better robustness, portability and high speed, and can be embedded into a camera to analyze the behavior of the pedestrian in the current scene; and the method has far-reaching significance in application in the field of security and protection.

Description

technical field [0001] The invention relates to a machine vision human body abnormal behavior recognition method, in particular to a machine vision human body abnormal behavior recognition method based on multi-feature fusion, which belongs to the field of intelligent security. Background technique [0002] With the development of computer technology, the Internet and artificial intelligence, the scale of video images has grown exponentially. How to let the machine "recognize" the image according to the human way of thinking, and realize the automatic understanding of the image in different scenes, has become an urgent problem in the field of machine vision. [0003] Nowadays, cameras are ubiquitous, and massive amounts of video data are generated every moment, and cameras are widely used in the field of security; however, most of the current cameras are used as video collection devices, and cannot detect abnormalities of people in the scene. Behavior recognition, usually a...

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): G06K9/00G06N3/04
CPCG06V40/174G06V40/172G06V40/20G06N3/045
Inventor 陈双叶张洪路
Owner BEIJING UNIV OF TECH
Features
  • R&D
  • Intellectual Property
  • Life Sciences
  • Materials
  • Tech Scout
Why Patsnap Eureka
  • Unparalleled Data Quality
  • Higher Quality Content
  • 60% Fewer Hallucinations
Social media
Patsnap Eureka Blog
Learn More

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

© 2025 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com