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

Monitoring system and method for judging video behavior based on deep learning

A deep learning and monitoring system technology, applied in neural learning methods, CCTV systems, TV system components, etc., can solve problems such as poor video behavior analysis ability and small video acquisition range

Active Publication Date: 2018-09-28
广州飞宇智能科技有限公司
View PDF8 Cites 28 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0007] The purpose of the present invention is to provide a monitoring system and method for judging video behavior based on deep learning, so as to solve the problems of small video acquisition range and poor video behavior analysis ability of existing video monitoring equipment

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
  • Monitoring system and method for judging video behavior based on deep learning
  • Monitoring system and method for judging video behavior based on deep learning
  • Monitoring system and method for judging video behavior based on deep learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0041] The monitoring system for judging video behavior based on deep learning in this embodiment 1 includes a monitoring terminal and a server; the monitoring terminal includes a camera and a front-end alarm module; the server includes a cloud platform video database, an offline video database, and a deep learning judgment module and abnormal event alarm module;

[0042] The camera is used to collect video information in real time and sent to the cloud platform video database for storage; the cloud platform video database is used to store the video information sent by the camera and forward the deep learning judgment module for behavior analysis; the offline The video database is used to store video information for the deep learning judgment module to construct a behavior judgment model; the deep learning judgment module is used to construct a behavior judgment model according to the video in the offline video database, and according to the behavior judgment model to receive ...

Embodiment 2

[0053] Further, on the basis of embodiment 1:

[0054] The feature extraction module is used to extract object, scene, behavior action and optical flow feature respectively from the video to be described; the feature extraction module includes object scene behavior action feature extraction submodule, 3D convolution feature extraction submodule and optical flow feature extraction submodule module; the object scene behavior action feature extraction sub-module includes a frame-by-frame image extraction unit and a feature vector generation unit; the frame-by-frame image extraction unit is used to divide the video to be described according to the specified frame rate fps, and randomly extract the The 80 frames of images are used for feature extraction in the next step; the feature vector generation unit is used to input the sampled frames to ImageNet, Places365, UCF-101 respectively. The pre-trained GoogleNet model of these three data sets extracts the features of the pool5 layer,...

Embodiment 3

[0114] Further, on the basis of embodiment 2:

[0115] The monitoring method of this embodiment 3 is used for a monitoring system based on deep learning to judge video behavior, including steps:

[0116] Step 1: the DSP controller 1 presets the first target rotation angle a, the second target rotation angle b and the target moving distance c;

[0117] Step 2: the DSP controller 1 sends a shooting command to the image acquisition part 2 to collect video, and the image acquisition part 2 sends the collected video to the DSP controller 1;

[0118] Step 3: the DSP controller 1 sends the received video to the cloud platform video database to store and forward the deep learning judgment module;

[0119] Step 4: The deep learning judgment module performs behavior judgment on the received video, and sends corresponding alarm commands to the front-end alarm module and the abnormal event alarm module according to the judgment result;

[0120] Step 5: the DSP controller 1 sends a rotat...

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 provides a monitoring system for judging a video behavior based on deep learning. The system comprises a monitoring terminal and a server; the monitoring terminal comprises a camera anda front end alarm module; the server comprises a cloud platform video database, an offline video database, a deep learning judgment module and an abnormal event alarm module; the camera is used for collecting video information in real time; the cloud platform video database stores the video information sent by the camera; the offline video database stores the video information for the deep learning judgment module to construct a behavior judgment model; the deep learning judgment module constructs the behavior judgment model according to the videos in the offline video database and judges thereceived video behavior according to the behavior judgment model; the front end alarm module is an alarm light or a horn; the abnormal event alarm mode is an alarm light or a horn; and the deep learning judgment module comprises a feature extraction module, a feature fusing and splicing module, a semantic coding determining modules, a feature decoding modules and a classification alarm module.

Description

technical field [0001] The invention relates to the technical field of video monitoring, in particular to a monitoring system and method for judging video behavior based on deep learning. Background technique [0002] The lens of the traditional video surveillance system can only be rotated within a certain angle range, and the range of its image acquisition is only limited to a small range; for the case of large-scale image acquisition or comprehensive video surveillance , the traditional video surveillance system is no longer applicable; [0003] The traditional video surveillance system only completes the collection of video information, but cannot accurately analyze the collected video information, and cannot effectively judge the video content. [0004] With the explosive growth of Internet multimedia data such as pictures and videos, computer vision has become a hot research field today. In the past, tasks that completely relied on manual annotation and description 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
Patent Type & Authority Applications(China)
IPC IPC(8): H04N7/18H04N5/76G08B13/196G06N3/08
CPCG06N3/08G08B13/19602H04N5/76H04N7/18
Inventor 陈劲全田菁余卫宇林俊科
Owner 广州飞宇智能科技有限公司
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

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com