Outdoor fire source identification method

Abstract

The invention discloses an outdoor fire source identification method, and relates to the technical field of fire source identification and detection, and the method comprises the following steps: S1, data sampling: collecting video information through a camera video, transmitting the video information to a computer end, reading the video information through the computer, and obtaining a video image; s2, extracting a suspected area of a fire source, and extracting the suspected area of the fire source according to a two-frame difference method; s3, flame shape features and flame area features of the suspected fire source area are extracted, S4, whether a fire source is generated in the suspected fire source area or not is comprehensively judged according to changes of the flame shape features and the flame area features of the suspected fire source area, and if it is judged that the fire source is generated in the suspected fire source area through the flame area changes and the flame shape change features, the suspected fire source area is judged to be the suspected fire source area. And if so, judging that the area is a fire source area.

Description

Technical Field

[0001] The present invention relates to the technical field of fire source identification and detection, and specifically to an outdoor fire source identification method. Background Technique

[0002] Fire is a phenomenon that emits light and heat during the process of material combustion, and it is also a form of energy emission. Since different environments will interfere with the video by bringing different noise information during the process of fire video acquisition, affecting the image quality, the recognition efficiency and accuracy are relatively low during the process of fire recognition.

[0003] Therefore, we propose an outdoor fire source identification method to solve the problems mentioned above. Summary of the Invention

[0004] The purpose of the present invention is to provide an outdoor fire source identification method to solve the problems in the current market mentioned in the above background technique.

[0005] To achieve the above purpose, the present invention provides the following technical solution: An outdoor fire source identification method, including the following steps:

[0006] S1: Data sampling

[0007] Collect video information through camera video monitoring and transmit it to the computer terminal. The computer reads the video information to obtain video images;

[0008] S2: Extract the suspected fire source area

[0009] Extract the suspected fire source area according to the two-frame difference method;

[0010] S3: Extract the flame shape feature and flame area feature of the suspected fire source area

[0011] S4: Comprehensively judge whether a fire source is generated in the suspected fire source area according to the changes in the flame shape feature and flame area feature of the suspected fire source area

[0012] If it is judged that a fire source is generated in the suspected fire source area through both the flame area change and the flame shape change features, then judge that this area is the fire source area.

[0013] Preferably, the specific method for extracting the suspected fire source area by the two-frame difference method is:

[0014] S21: Convert the front and rear two-frame images in the obtained video stream into grayscale images, and then perform Gaussian blur to eliminate noise interference;

[0015] S22: Subtract the front and rear two-frame images to obtain the difference area between the two images;

[0016] S23: Then perform binary segmentation on the difference image, taking the difference region as the foreground and the unchanged region as the background, and perform an opening operation to eliminate some small interferences.

[0017] Preferably, the method for extracting the flame shape feature is as follows:

[0018] The similarity of the flame contour is represented by the position relationship of pixel points between images, and its expression is:

[0019]

[0020] In the formula: σ i represents the similarity of the flame contour at time t, B i (x, y) represents the pixel at the coordinate (x, y) in the i-th image. If this pixel is a bright point, then B i (x, y) = 1, otherwise B i (x, y) = 0; Φ represents the suspected fire source area;

[0021] Let σ i be the similarity of pixel point i, then the total similarity of the suspected fire source area can be calculated as:

[0022]

[0023] where N is the number of pixel points;

[0024] Define the flame similarity discrimination threshold and where If or then it indicates low similarity, and the suspected area is determined as a non-fire source area. If indicates high similarity, then it is considered that the video image of the suspected fire source area has the shape change characteristics of the flame, and the suspected fire source area is the fire source area.

[0025] Preferably, the method for extracting the flame area feature is as follows:

[0026] The growth of the flame area is represented by the flame area growth rate, and its calculation formula is as follows:

[0027] ΔS t =(S t+Δt -S t ) / Δt

[0028] In the formula, S t 、S t+ΔtThey are the flame image area values at times t and t + Δt respectively. By performing binary processing on the suspected fire source area extracted above, the pixel points with all pixel values of 255 are summed, and the result is used as the flame image area;

[0029] If the flame area shows an increasing trend, then a fire source is generated in the suspected fire source area; otherwise, no fire source is generated.

[0030] Compared with the prior art, the beneficial effects of the present invention are:

[0031] An outdoor fire source recognition method of the present invention first extracts a suspected fire source area through the two-frame difference method to initially extract the suspected area where a fire source may be generated, and then extracts the flame shape characteristics and flame area characteristics of the suspected fire source area. Through comprehensive judgment based on the change in the flame area and the change in the flame shape, when the corresponding characteristics all meet the corresponding conditions, it is determined that this area is the fire source area. The method of the present invention can quickly and effectively identify fire sources. Description of the Drawings

[0033] Figure 1 It is a flowchart of an outdoor fire source recognition method of the present invention. Detailed Embodiments

[0034] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the specification. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0035] The present invention provides a technical solution: an outdoor fire source recognition method, including the following steps:

[0036] S1: Data Sampling

[0037] Video information is collected through a camera video monitor and transmitted to the computer terminal, and the computer reads the video information to obtain video images;

[0038] S2: Extract the suspected fire source area

[0039] Extract the suspected fire source area according to the two-frame difference method;

[0040] The specific method for extracting the suspected fire source area by the two-frame difference method is as follows:

[0041] S21: Convert the front and rear two-frame images in the obtained video stream into grayscale images, and then perform Gaussian blur to eliminate noise interference;

[0042] S22: Subtract the previous and the next frames of images to obtain the difference region between the two images;

[0043] S23: Then perform binary segmentation on the difference image, taking the difference region as the foreground and the unchanged region as the background, and perform an opening operation to eliminate some minor interferences;

[0044] S3: Extract the flame shape feature and the flame area feature of the suspected fire source region

[0045] The method for extracting the flame shape feature is as follows:

[0046] The similarity of the flame contour is represented by the position relationship of pixel points between images, and its expression is:

[0047]

[0048] In the formula: σ i represents the similarity of the flame contour at time t, Bi(x, y) represents the pixel at the coordinate (x, y) in the i-th image. If the pixel is a bright point, then B i (x, y) = 1, otherwise B i (x, y) = 0; Φ represents the suspected fire source region;

[0049] Let σ i be the similarity of pixel point i, then the total similarity of the suspected fire source region can be calculated as:

[0050]

[0051] where N is the number of pixel points;

[0052] Define the discrimination threshold of the flame similarity and where If or then it indicates low similarity, and the suspected region is determined as a non-fire source region. If indicates high similarity, then it is considered that the video image of the suspected fire source region has the shape change characteristics of the flame, and the suspected fire source region is the fire source region;

[0053] The method for extracting the flame area feature is as follows:

[0054] In the initial stage of the object generating a flame, the flame spreads rapidly in all directions, and its characteristic manifestation is that the flame area continuously increases. Therefore, the change in the flame area is used as the judgment basis for the generation of the fire source, and the growth of the flame area is represented by the flame area growth rate. Its calculation formula is as follows:

[0055] ΔS t = (S t+Δt-S t ) / Δt

[0056] Wherein, S t , S t+Δt are the flame image area values at times t and t + Δt respectively. By performing binarization processing on the suspected fire source area extracted above, the pixel points with all pixel values of 255 are summed, and the result is used as the flame image area;

[0057] If the flame area shows an increasing trend, then a fire source is generated in the suspected fire source area; otherwise, no fire source is generated.

[0058] S4: Comprehensively judge whether a fire source is generated in the suspected fire source area according to the changes in the flame shape characteristics and flame area characteristics of the suspected fire source area

[0059] If it is judged that a fire source is generated in the suspected fire source area through both the change in the flame area and the change in the flame shape, then this area is judged as the fire source area.

[0060] A method for identifying outdoor fire sources according to the present invention first extracts the suspected fire source area by the two-frame difference method, initially extracts the area suspected of generating a fire source, and then extracts the flame shape characteristics and flame area characteristics of the suspected fire source area. Through comprehensive judgment of the change in the flame area and the change in the flame shape, when the corresponding characteristics all meet the corresponding conditions, it is determined that this area is the fire source area. The fire source can be quickly and effectively identified by the method of the present invention.

[0061] Although the present invention has been described in detail with reference to the foregoing embodiments, for those skilled in the art, they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements on some of the technical features. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. An outdoor fire source recognition method, characterized in that: It includes the following steps: S1: Data sampling Video information is collected through camera video monitoring and transmitted to the computer side. The computer reads the video information to obtain video images; S2: Extract the suspected fire source area The suspected fire source area is extracted according to the two-frame difference method; S3: Extract the flame shape feature and flame area feature of the suspected fire source area S4: Comprehensively judge whether a fire source is generated in the suspected fire source area according to the changes in the flame shape feature and flame area feature of the suspected fire source area If it is judged that a fire source is generated in the suspected fire source area through both the flame area change and the flame shape change features, then this area is judged as the fire source area.

2. The outdoor fire source recognition method according to claim 1, characterized in that: The specific method for extracting the suspected fire source area by the two-frame difference method is as follows: S21: Convert the front and rear two-frame images in the obtained video stream into grayscale images, and then perform Gaussian blur to eliminate noise interference; S22: Subtract the front and rear two-frame images to obtain the difference area between the two images; S23: Then perform binary segmentation on the difference image, take the difference area as the foreground and the unchanged area as the background, and perform opening operation to eliminate some minor interferences.

3. The outdoor fire source recognition method according to claim 1, characterized in that: The method for extracting the flame shape feature is as follows: The similarity of the flame contour is represented by the position relationship of pixel points between images, and its expression is: Where: σ i represents the similarity of the flame contour at time t, B i (x, y) represents the pixel at the coordinate (x, y) in the i-th image. If the pixel is a bright point, then B i (x, y) = 1, otherwise B i (x, y) = 0; Φ represents the suspected fire source area; Let σ i be the similarity of pixel point i. Then the total similarity of the suspected fire source area can be calculated as follows: where N is the number of pixel points; Define the discrimination threshold for flame similarity and where If or it indicates low similarity and determines that the suspected area is a non-fire source area. If indicates high similarity, it is considered that the video image of the suspected fire source area has the morphological change characteristics of a flame, and the suspected fire source area is a fire source area.

4. The outdoor fire source recognition method according to claim 1, characterized in that: The method for extracting the flame area feature is as follows: The growth of the flame area is represented by the flame area growth rate, and its calculation formula is as follows: ΔS t = (S t+Δt - S t ) / Δt Where S t and S t+Δt are the flame image area values at times t and t + Δt respectively. By performing binarization processing on the suspected fire source area extracted above, the pixel points with all pixel values of 255 are summed, and the result is used as the flame image area; If the flame area shows an increasing trend, then a fire source is generated in the suspected fire source area, otherwise no fire source is generated.

Images