Unlock instant, AI-driven research and patent intelligence for your innovation.

A Smoke Detection Approach Combining Deep Convolutional Neural Networks and Visual Change Maps

A deep convolution and neural network technology, applied in the field of smoke detection, to achieve the effect of improving reliability

Active Publication Date: 2021-07-27
NAT UNIV OF DEFENSE TECH
View PDF8 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, for objects that are particularly similar to smoke, such as clouds and fog, false alarms still exist

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
  • A Smoke Detection Approach Combining Deep Convolutional Neural Networks and Visual Change Maps
  • A Smoke Detection Approach Combining Deep Convolutional Neural Networks and Visual Change Maps
  • A Smoke Detection Approach Combining Deep Convolutional Neural Networks and Visual Change Maps

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0029]The present invention proposes a smoke detection method combining a deep convolutional neural network and a visual change map. The main innovation is to propose the concept of a visual change map to describe the physical diffusion characteristics of smoke. In the implementation process, first use the deep convolutional neural network to detect the suspected smoke area in the image, and then use the visual change map to make a second judgment on the suspected smoke area to eliminate clouds, fog and other objects that are particularly similar to smoke, reducing the cost of smoke detection. false alarm;

[0030] The implementation steps are as follows:

[0031] Step1: Input the kth frame image, the size of the image in the present invention is 240×320;

[0032] Step2: If k>1, it means that the current image is not the first frame image, and go to the next step; otherwise, cache the current frame image, k=k+1, and return to Step1;

[0033] Step3: Detection of suspected smo...

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 present invention relates to a smoke detection method combining a deep convolutional neural network and a visual change graph. First, a deep convolutional neural network is used to initially detect the suspected smoke area. On this basis, based on the physical characteristics of smoke diffusion, a visual change map is constructed based on video motion changes, and then the SVM classifier is used to realize the secondary judgment of the smoke area. The beneficial effect of the present invention is to reduce clouds, fog, etc. and smoke The false alarms caused by similar targets further improve the reliability of smoke detection.

Description

technical field [0001] The present invention mainly relates to a smoke detection method combining a deep convolutional neural network and a visual change map. [0002] technical background [0003] Among various disasters, fire is one of the main disasters that most frequently and commonly threaten public safety and social development. Fire not only destroys material property, causes chaos in social order, but also directly or indirectly endangers lives. In order to reduce fire hazards, in addition to various fire prevention measures, fire detection technology is also needed to detect and deal with fires in time. When a fire occurs, usually there is smoke first and then fire. Therefore, rapid detection of smoke is conducive to early detection of fire and reduction of fire hazards. Traditional smoke sensors require smoke to enter the sensor and the concentration reaches a certain level to be detected, which is difficult to use in outdoor open spaces. Vision-based smoke dete...

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): G06K9/00G06K9/46G06K9/62G06N3/04
CPCG06V20/52G06V10/462G06N3/045G06F18/2411G06F18/214
Inventor 程江华刘通陈朔陈明辉华宏虎张亮王洋
Owner NAT UNIV OF DEFENSE TECH