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

Forest fire smoke detection method based on image segmentation

A technology of forest fire and image segmentation, applied in the direction of instruments, character and pattern recognition, computer parts, etc., to achieve the effect of solving the problem of over-segmentation

Inactive Publication Date: 2019-09-10
HARBIN UNIV OF SCI & TECH
View PDF6 Cites 13 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, most of the traditional forest fire smoke detection algorithms only consider the fire detection in an ideal background and analyze the image at the pixel level, which is not necessarily suitable for smoke recognition in a complex environment such as a forest.

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
  • Forest fire smoke detection method based on image segmentation
  • Forest fire smoke detection method based on image segmentation
  • Forest fire smoke detection method based on image segmentation

Examples

Experimental program
Comparison scheme
Effect test

specific Embodiment approach 1

[0048] The forest fire smoke detection method based on image segmentation of the present embodiment, described method comprises the following steps:

[0049] SLIC superpixel segmentation

[0050] A superpixel is a small area composed of a series of adjacent pixels with similar characteristics. SLIC is an improvement to the k-means clustering algorithm. It defines the distance between pixels according to color and space, and reduces the amount of calculation by limiting the search space. The computational complexity is linearly related to the number of pixels N, and is independent of The number K of superpixels.

[0051] Step 1. Initialization:

[0052] For color images in CIELAB space, initialize K cluster centers; move the cluster centers to the lowest gradient of the 3×3 neighborhood to avoid placing the cluster centers on edges or noise points;

[0053] Step 2. Assignment:

[0054] In the allocation process, according to the similarity of the measurement, each pixel i i...

specific Embodiment approach 2

[0093] The difference from Embodiment 1 is that in the forest fire smoke detection method based on image segmentation in this embodiment, in step 6, the process of binary classification of superpixel blocks is that there are many relevant features available for research, such as Spectral features, texture features, geometric features, etc., but considering the complexity of the forest environment, according to the uncertainty of the shape of the smoke and the limitation of the monitoring distance, analyze the spectral information features of the pixel block; the smoke itself and the forest background have different spectral information. Obvious difference; in RGB and HIS color spaces, the mean value of each superpixel block R, G, B, M-N, S and I, and the mean square error of each pixel block gray value are extracted as the input features of the pattern classifier. Among them, M and N are the maximum and minimum values ​​of R, G, and B; 46 smoke pixel blocks and 53 background pi...

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 forest fire smoke detection method based on image segmentation, and belongs to the field. Most traditional forest fire hazard smoke detection algorithms only consider fire hazard detection under an ideal background, analyze images on a pixel level, and are not necessarily suitable for smoke identification under complex environments such as forests. A forest fire smoke detection method based on image segmentation sequentially comprises the following steps of initializing a color image in a CIELAB space a clustering center; moving the clustering center to the lowest gradient of a 3 * 3 neighborhood; performing distribution; updating; when each pixel is classified to the nearest cluster center, updating the cluster center by using the average value of vectors of allpixels in the region; merging the isolated pixel points into the nearest super pixel; combining the superpixels; segmenting the sky and the ground; performing smoke recognition. The limitation that acamera needs to be fixed in a traditional forest fire smoke detection algorithm is broken through, and the method is suitable for dynamic panoramic sampling of the camera on a forest.

Description

technical field [0001] The invention relates to a forest fire smoke detection method based on image segmentation. Background technique [0002] Intelligent image processing technology of forest fire is a new research field. In order to avoid the spread of fire, the real-time and accuracy of fire detection is crucial; in addition, the prediction of fire spread and the location of fire are also conducive to timely fire fighting. In the forest, due to the occlusion of trees, the appearance of smoke is often earlier than that of flames. Therefore, the study of smoke detection based on digital images can give early warning of forest fires and intuitively locate the fire location. The existing technology has made a more comprehensive summary of forest fire smoke detection algorithms based on video images: generally, it can be detected from the spectral characteristics, shape characteristics and motion characteristics of smoke, and then assisted by other detection methods, such as...

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): G06K9/00G06K9/62G06K9/40G06K9/34
CPCG06V20/41G06V10/26G06V10/30G06F18/22G06F18/2411
Inventor 张玉萍于家博
Owner HARBIN UNIV OF SCI & TECH
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