Forest fire smoke detection method based on image segmentation
What is Al technical title?
Al technical title is built by PatSnap Al team. It summarizes the technical point description of the patent document.
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
This helps you quickly interpret patents by identifying the three key elements:
Problems solved by technology
Method used
Benefits of technology
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
Click on the blue label to locate the original text in one second.
Reading with bidirectional positioning of images and text.
Smart Image
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
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
Application Date:The date an application was filed.
Publication Date:The date a patent or application was officially published.
First Publication Date:The earliest publication date of a patent with the same application number.
Issue Date:Publication date of the patent grant document.
PCT Entry Date:The Entry date of PCT National Phase.
Estimated Expiry Date:The statutory expiry date of a patent right according to the Patent Law, and it is the longest term of protection that the patent right can achieve without the termination of the patent right due to other reasons(Term extension factor has been taken into account ).
Invalid Date:Actual expiry date is based on effective date or publication date of legal transaction data of invalid patent.
Login to view more 
Login to view more