An Online Adaptive Anomaly Event Detection Method in Video Scenarios

A technology for abnormal events and video scenes, applied in computer parts, instruments, computing, etc., can solve the problems of low detection accuracy, insufficient research, and insufficient description of foreground information.

Active Publication Date: 2019-04-23
HANGZHOU DIANZI UNIV
View PDF3 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, the research of this method is not deep enough, and its description of foreground information is not accurate enough, which leads to the reduction of detection accuracy.

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 Online Adaptive Anomaly Event Detection Method in Video Scenarios
  • An Online Adaptive Anomaly Event Detection Method in Video Scenarios
  • An Online Adaptive Anomaly Event Detection Method in Video Scenarios

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0026] Attached below figure 1 , the specific embodiment of the present invention is further described in detail. Since the detection model used in the present invention is based on online processing, that is to say, it is necessary to learn the motion information of the foreground object online, to obtain which are the mainstream motion patterns, and which are the motion patterns with significant differences from the mainstream motion patterns, and think that they are different from the mainstream motion patterns. A movement pattern with a significant difference from the dominant movement pattern is the movement pattern of an abnormal event. This method needs to use a video frame at the beginning of the video stream as the initialization data to learn the motion pattern of the foreground object, and then calculate a threshold according to the motion information in the initialization data, and calculate the motion pattern of the foreground object in the new incoming video fram...

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 online adaptive abnormal event detection method in a video scene. The present invention is based on the time-space domain characteristics of the three-dimensional optical flow histogram and an online adaptive abnormal event detection method. In the feature extraction stage, the method uses the three-dimensional optical flow histogram as the description of the foreground object movement based on the time-space domain in the video scene. In the abnormal event detection stage, an online adaptive method is used for detection. The present invention not only uses the time-space domain information based on the three-dimensional optical flow histogram in the feature extraction stage, but also uses an online self-adaptive method in the detection stage, which improves the adaptability of the abnormal event detection model to different scenarios.

Description

technical field [0001] The invention relates to a method for detecting abnormal events in a video scene, in particular to an online self-adaptive abnormal event detection method in a video scene. The invention is an abnormal event detection method based on the time-space domain characteristics of the three-dimensional optical flow histogram and online clustering. Background technique [0002] With more and more monitoring scenes in the real world, scene analysis under video surveillance has also attracted more and more scholars' attention. Although the existing detection methods have good accuracy and time performance, they can only detect abnormal events in specific scenarios. However, surveillance videos in real scenarios are constantly changing and cannot be estimated. We cannot judge in advance which situations are abnormal events and which situations are normal events. Therefore, this requires us to adaptively learn the motion pattern of the foreground in this scene f...

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): G06K9/00
CPCG06V20/41
Inventor 徐向华吕艳艳李平
Owner HANGZHOU DIANZI 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