Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

An adaptive video object behavior track analysis method based on a convolutional neural network

A convolutional neural network and video object technology, applied in the field of behavior analysis of video objects, to achieve the effect of ensuring reliability, reducing algorithm complexity, and ensuring real-time performance

Pending Publication Date: 2019-06-25
JILIN UNIV
View PDF10 Cites 6 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

The difficulty of detection and tracking still lies in complex background situations such as occlusion and scale changes

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
  • An adaptive video object behavior track analysis method based on a convolutional neural network
  • An adaptive video object behavior track analysis method based on a convolutional neural network
  • An adaptive video object behavior track analysis method based on a convolutional neural network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0066] The specific embodiment of the present invention is described below in conjunction with accompanying drawing:

[0067] The basic framework of the present invention's adaptive video object behavior trajectory analysis method based on convolutional neural network is as follows: figure 1 shown;

[0068] 1. Video collection: retrieve the surveillance video of the overall environment in the real scene, such as the surveillance video of a bank business hall;

[0069] 2. Detection and tracking: Use the detection-based tracking method and kernel correlation filter algorithm to accurately detect and track moving targets in the video. During the tracking process, use the kernel correlation filter to train the target detector, and select the target area as a positive sample. The background area is a negative sample, and when the filter template acts on the tracking target, the position of the maximum response value obtained is the target position;

[0070] The specific algorithm i...

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 an adaptive video object behavior track analysis method based on a convolutional neural network, and belongs to the technical field of image engineering and machine vision. Fora monitoring video in a real scene, a kernel correlation filtering algorithm is adopted to effectively detect and track a video object, so that a track graph of a directed curve is generated, a convolutional neural network is utilized based on deep learning to adaptively train and learn a normal behavior track in the current scene, and finally suspicious behaviors are distinguished; According tothe invention, deep learning is applied to behavior analysis of the video object; the normal behavior track in the scene can be learned in a self-adaptive manner; The method has the advantages that manual definition is not needed, so that the suspicious behavior judgment reliability is guaranteed, the use background can be continuously expanded in a real scene, the detection tracking part utilizesthe characteristics of a cyclic matrix, the deep learning input is a two-dimensional track graph, the space-time characteristics of a video are reserved, and the real-time performance is guaranteed to the maximum extent.

Description

technical field [0001] The invention belongs to the technical fields of image engineering and machine vision, and in particular relates to a behavior analysis method of video objects. Background technique [0002] With the advent of the era of artificial intelligence, intelligent video surveillance systems have become an important part. Under the surveillance camera, for the video objects in the real scene, it can automatically detect and track in real time, and conduct effective behavior analysis and understanding to realize intelligent video surveillance in the true sense. It needs to be integrated with mathematics, computer, electronics, control, psychology and other disciplines, and can play a key role in many fields such as national security, traffic management, and social services. This topic has high theoretical research value and broad practical application prospects. [0003] Detection and tracking of moving targets: From the perspective of research objects, it ca...

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): G06K9/00
Inventor 王世刚赵文婷卢洋赵岩韦健杨楚皙
Owner JILIN UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products