Fire detection method based on convolutional neural network

A convolutional neural network and detection method technology, applied in the field of machine vision applications, can solve problems such as the inability to guarantee detection accuracy and reliability, and achieve the effect of preliminary positioning

Pending Publication Date: 2020-05-19
浙江中创天成科技有限公司
View PDF5 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

The above detection methods have certain limitations in the face of complex scenes, less data, and light occlusion, and cannot guarantee the accuracy and reliability of detection.

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
  • Fire detection method based on convolutional neural network
  • Fire detection method based on convolutional neural network
  • Fire detection method based on convolutional neural network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0033] (1) Collect videos with fire scenes and normal scenes without fire from search engines such as Baidu and Google, and intercept them into two types of pictures, 7,500 each, a total of 15,000.

[0034] (2) These pictures are input into the convolutional neural network model that the present invention builds as data set and train.

[0035] (3) After the training, input a new video with a fire scene as a test. The results show that the occurrence of a fire can basically be detected in the fire scene, and the superpixel outline intuitively depicts the shape and position of the flame. According to statistics Results The true class of the experiment has 7223 images and the TPR is about 96.3%. The experiment shows that the proposed method has high accuracy and robustness, and it can detect and describe the general shape of the fire well. The feasibility of the method is proved.

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 fire detection method based on a convolutional neural network, and the method comprises the steps: constructing the convolutional neural network for training, enabling the input of the convolutional neural network to be a collected fire image, and enabling the output of the convolutional neural network to be a probability value of whether an input image is a fire image; the method comprises the steps of verifying whether an input to-be-predicted picture is a fire picture, performing detection in a superpixel detection mode, performing superpixel segmentation on the to-be-predicted picture, performing binary classification detection on pixel points in each segmented superpixel region, and obtaining the shape of a fire on the to-be-predicted picture according to a detection result of each grid. Compared with a traditional target detection and positioning method, the method has the advantages that target positioning is achieved without a labeled training set, andin the aspect of accuracy, fire binary classification detection has the highest 93% accuracy in pictures and is low in complexity.

Description

technical field [0001] The invention belongs to the field of machine vision applications, and in particular relates to a picture and video fire detection technology based on a convolutional neural network Background technique [0002] Fires often occur in various scenarios, and timely detection of whether there is a fire has important application value in real life. It is widely used in video surveillance, traffic monitoring, public safety, urban planning, and the construction of smart supermarkets, such as monitoring the security of an important meeting place to ensure the smooth progress of the meeting. [0003] As a commonly used means of public security prevention and control, computer vision is used to judge scene pictures and videos, and most of them currently use big data-based target detection methods. The above detection methods have certain limitations in the face of complex scenes, less data, and light occlusion, and cannot guarantee the accuracy and reliability ...

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): G06K9/00G06K9/62G06N3/04
CPCG06V20/41G06V20/52G06N3/045G06F18/214G06F18/241
Inventor 苏宏业马龙华陆哲明虞斌超
Owner 浙江中创天成科技有限公司
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

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap