Identification method of fire source in charging place based on automatic tracking and positioning of machine vision

Abstract

The present invention provides a method for identifying the fire source point of a charging place fire based on automatic tracking and positioning by machine vision, which relates to the technical field of fire fighting equipment. A thermal imaging image and a visible light image of a target area are simultaneously acquired and preprocessed, and a suspicious area is located by threshold segmentation. Flame thermal features and flame color features are respectively extracted from the suspicious areas of the thermal imaging image and the visible light image; a flame visual recognition coefficient is obtained by analyzing the flame color features; an edge temperature coefficient and an internal temperature coefficient are obtained by mathematically analyzing the flame thermal features; a fire source recognition coefficient is obtained by mathematically analyzing the flame visual recognition coefficient, the edge temperature coefficient and the internal temperature coefficient, and a fire source is identified by comparing the fire source recognition coefficient with a preset threshold.

Description

Technical Field

[0001] The present invention relates to the technical field of fire fighting equipment, and in particular to a method for identifying a fire source point in a charging place based on automatic tracking and positioning by machine vision. Background Art

[0002] With the rapid development and popularization of electric vehicles, the number of charging facilities has increased significantly, and with it comes the potential risk of fire during the charging process. Fire may not only cause property loss, but also pose a more serious threat to personal safety. At present, conventional fire-fighting robots are increasingly used in fire rescue, but their intelligence level is still low, especially when operating the fire cannon of the fire-fighting robot to aim at the fire source, the efficiency is low and highly dependent on the skill level of the operator. Due to the complex and changeable environment at the fire scene, the judgment and operation of the operator are crucial to the effect of fire extinguishing. Skilled operators can respond quickly, while novices or inexperienced personnel may hesitate at critical moments, resulting in reduced fire extinguishing timeliness and unnecessary delays in rescue. In addition, traditional fire-fighting robots often rely on manual command when locating the fire source, and cannot achieve automatic tracking and precise positioning, which affects the accuracy and effectiveness of fire-fighting operations.

[0003] In recent years, the automatic tracking and positioning technology of machine vision has made significant progress. This technology can quickly and accurately identify the fire source by analyzing the on-site images, improving the robot's response ability and positioning accuracy in complex fire environments. Compared with traditional methods, firefighting robots using automated visual recognition can not only reduce the dependence on operator skills, but also quickly lock the fire source in the early stages of a fire, greatly improving the timeliness and efficiency of firefighting and rescue.

[0004] In the prior art, the publication number CN111738082A discloses a method and device for automatically tracking and locating the fire source point based on machine vision. The prior art uses a binocular infrared thermal imager to collect fire scene images; pre-processes the fire scene images, and segments the fire source point image information according to the image segmentation algorithm; clusters the fire source point images and extracts the fire source point features according to the clustering-based image automatic threshold algorithm; constructs a fire source point identifier based on the extracted fire source point features by using a deep learning method, and identifies the fire source point in the fire scene image through the fire source point identifier; and determines the three-dimensional spatial coordinates of the fire source point by triangulation principle according to the parallax of the identified fire source point in the two fire scene images of the binocular infrared thermal imager. However, the prior art still has defects. The prior art only collects thermal imaging images of the fire scene, and the data type is too single. This limitation makes the accuracy of the recognition result insufficient, which may waste precious rescue time. For complex fire scenes, it is difficult to fully reflect the on-site situation by relying solely on thermal imaging technology, which in turn affects timely and effective response measures.

[0005] The above information disclosed in this Background section is only for enhancement of understanding of the background of the present disclosure and therefore it may contain information that does not constitute the prior art that is already known to one of ordinary skill in the art. Summary of the invention

[0006] The purpose of the present invention is to provide a method for identifying the fire source of a charging place fire based on automatic tracking and positioning by machine vision, so as to solve the problems raised in the above-mentioned background technology.

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] A method for identifying the fire source of a charging place fire based on automatic tracking and positioning by machine vision, the specific steps comprising:

[0009] Step 1: Synchronously acquire the thermal imaging image and visible light image of the target area, convert the visible light image from the RGB color space to the HSV color space to obtain a color recognition image, grayscale the thermal imaging image to obtain a temperature recognition image, and map each pixel position of the temperature recognition image and the color recognition image one by one;

[0010] Step 2: Use the maximum inter-class variance method to divide the temperature recognition image into a high temperature part and a normal temperature part, extract the area of ​​the color recognition image corresponding to the high temperature part to obtain a high temperature color image, perform color threshold processing on the high temperature color image, divide the high temperature color image into a flame color part and other color parts, extract the flame color part as a flame color image, and extract the area of ​​the temperature recognition image corresponding to the flame color part to obtain a flame thermal image;

[0011] Step 3: Extract features of each flame color part and flame heat part respectively to obtain flame color feature data and flame heat feature data, and obtain the flame visual recognition coefficient by analyzing the flame color feature data; obtain the edge temperature coefficient and the internal temperature coefficient by mathematically analyzing the flame heat feature data;

[0012] Step 4: The fire source identification coefficient is obtained by mathematically analyzing the flame visual identification coefficient, the edge temperature coefficient and the internal temperature coefficient, and the fire source identification coefficient is compared with the preset fire source identification threshold. The area greater than the fire source identification threshold is identified as the fire source area, and the area less than or equal to the fire source identification threshold is identified as the non-fire source area.

[0013] Furthermore, the coordinates of the pixels of the temperature recognition image and the color recognition image are annotated, with the bottom pixel row being the first row and the leftmost pixel column being the first column. The pixel positions of the temperature recognition image and the color recognition image are mapped one-to-one through the coordinates so that each pixel has a unique coordinate value.

[0014] Furthermore, the specific logic of using the maximum inter-class variance method to divide the temperature recognition image into a high temperature part and a normal temperature part is as follows: the frequency of the pixel points of each level in the 256 gray levels is counted to form a gray histogram, and the inter-class variance of each threshold in the histogram is calculated, and the optimal separation threshold is selected through the inter-class variance, and the temperature recognition image is divided into a high temperature part and a normal temperature part through the optimal separation threshold;

[0015] The specific logic for calculating the optimal separation threshold is: traverse all possible thresholds, divide the construction site image into normal temperature part and high temperature part according to the value of each threshold, count the number of pixels in the normal temperature part and the high temperature part, calculate the mean of the gray value of the pixels in the normal temperature part and the high temperature part, and calculate the inter-class variance of the corresponding threshold. The specific formula is:

[0016] σ(t) 2 =w B (t)*w F (t)*(μ B (t)-μ F (t)) 2

[0017] Among them, σ(t) 2 represents the inter-class variance when the threshold is t, t∈[0, 255], and t∈N + ,w B (t), w F (t) are the number of pixels in the normal temperature part and the high temperature part when the threshold is t, μ B (t), μ F (t) are the mean grayscale values ​​of the pixels in the normal temperature part and the high temperature part when the threshold is t;

[0018] Calculate the threshold value when the inter-class variance reaches the maximum, and define this value as the optimal separation threshold.

[0019] Furthermore, the specific logic for dividing the high temperature color image into the flame color part and the other color parts is as follows:

[0020] The color threshold processing is performed on the high temperature color image. The specific logic for dividing the high temperature color image into the flame color part and the other color part is as follows: setting the hue threshold range, saturation threshold range and brightness threshold range, defining the area in the high temperature color image that meets the above three threshold ranges as the flame color part, and defining the other parts as the other color part. The threshold range is:

[0021]

[0022] Among them, H is hue, S is saturation, and V is brightness.

[0023] Furthermore, the specific logic for generating the flame visual recognition coefficient is: mathematically analyzing the flame color characteristics to generate the flame visual recognition coefficient, wherein the flame color characteristics include internal average saturation, internal saturation variance, and edge circularity; and the specific formula for generating the flame visual recognition coefficient is:

[0024]

[0025] Where PS is the flame visual recognition coefficient, is the internal average saturation, σ B is the internal saturation variance, and Y is the edge circularity.

[0026] Furthermore, the specific logic for generating the edge temperature coefficient and the internal temperature coefficient is: mathematically analyzing the flame thermal characteristic data to obtain the edge temperature coefficient and the internal temperature coefficient. The flame thermal characteristics include the edge temperature gradient variance, the edge average temperature gradient, the edge minimum temperature gradient, the internal temperature variance, the internal average temperature and the internal maximum temperature gradient; the specific formula for generating the edge temperature coefficient is:

[0027]

[0028] Where Te is the edge temperature coefficient, σ e is the edge temperature gradient variance, T e is the average edge temperature gradient, T min is the minimum temperature gradient at the edge;

[0029] The specific formula used to generate the internal temperature coefficient is:

[0030]

[0031] Where Tn is the internal temperature coefficient, σ n is the internal temperature variance, T n is the average internal temperature, T max is the maximum internal temperature gradient;

[0032] Furthermore, the specific logic for generating the fire source identification coefficient is: the fire source identification coefficient is generated by mathematically analyzing the flame visual identification coefficient, the edge temperature coefficient and the internal temperature coefficient; the specific formula for generating the fire source identification coefficient is:

[0033]

[0034] Among them, Ev is the fire source identification coefficient, PS is the flame visual identification coefficient, Te is the edge temperature coefficient, and Tn is the internal temperature coefficient.

[0035] Compared with the prior art, the present invention has the following beneficial effects:

[0036] The present invention collects thermal imaging image data and visible light image data, analyzes the two and generates a fire source identification coefficient that reflects the comprehensive similarity between the internal temperature and visual characteristics of the corresponding area and the fire source, and accurately identifies the fire source through the fire source identification coefficient. This comprehensive method can more comprehensively capture the multi-dimensional information of the fire scene and prevent misjudgment or missed judgment caused by single thermal imaging. By combining visual features, it can effectively make up for the shortcomings of thermal imaging and achieve more reliable fire source positioning. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 It is a schematic diagram of the overall method flow of the present invention. DETAILED DESCRIPTION

[0038] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with specific embodiments.

[0039] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the present invention should be understood by people with ordinary skills in the field to which the present invention belongs. The words "first", "second" and similar words used in the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. "Include" or "comprise" and similar words mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connect" or "connected" and similar words are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0040] Example:

[0041] See also Figure 1 , the present invention provides a technical solution:

[0042] A method for identifying the fire source of a charging place fire based on automatic tracking and positioning by machine vision, the specific steps comprising:

[0043] Step 1: Synchronously acquire the thermal imaging image and visible light image of the target area, convert the visible light image from the RGB color space to the HSV color space to obtain a color recognition image, grayscale the thermal imaging image to obtain a temperature recognition image, and map each pixel position of the temperature recognition image and the color recognition image one by one;

[0044] The specific formula used in the grayscale processing method of this embodiment is:

[0045] Y=0.298*R+0.587*G+0.114*B

[0046] Wherein, Y is the gray value of the pixel in the first recognition image, and R, G, and B are the red component value, green component value, and blue component value of the pixel of the leaf image after image scaling, respectively.

[0047] In this embodiment, HSV is a common color model, which represents three attributes of color: hue (H), saturation (S) and brightness (V). Hue represents the basic attribute of color. In the HSV color space, the value range of hue is 0 to 360 degrees, starting from red and encoding other colors in a clockwise direction. Saturation represents the purity or depth of color. A higher saturation means that the color is brighter, while a lower saturation means that the color is darker. In the HSV color space, the value range of saturation is 0% to 100%, where 0% represents a grayscale image and 100% represents a fully saturated color. Brightness represents the lightness and darkness of the color. A higher brightness value means that the color is brighter, while a lower brightness value means that the color is darker. In the HSV color space, the value range of brightness is 0% to 100%, where 0% represents black and 100% represents white. Compared with the common RGB color model, the HSV color model is closer to human perception of color and is easier to understand and use.

[0048] The pixel points of the temperature recognition image and the color recognition image are marked with coordinates. The bottom pixel row is the first row, and the leftmost pixel column is the first column. The pixel positions of the temperature recognition image and the color recognition image are mapped one by one through coordinates, so that each pixel has a unique coordinate value. Step 2: Use the maximum inter-class variance method to divide the temperature recognition image into a high temperature part and a normal temperature part, extract the area of ​​the color recognition image corresponding to the high temperature part to obtain a high temperature color image, perform color threshold processing on the high temperature color image, divide the high temperature color image into a flame color part and other color parts, extract the flame color part as a flame color image, and extract the area of ​​the temperature recognition image corresponding to the flame color part to obtain a flame thermal image.

[0049] The specific logic of using the maximum inter-class variance method to divide the temperature recognition image into a high temperature part and a normal temperature part is as follows: the frequency of the pixel points of each level in the 256 gray levels is counted to form a gray histogram, and the inter-class variance of each threshold in the histogram is calculated, and the optimal separation threshold is selected through the inter-class variance, and the temperature recognition image is divided into a high temperature part and a normal temperature part through the optimal separation threshold;

[0050] The specific logic for calculating the optimal separation threshold is: traverse all possible thresholds, divide the construction site image into normal temperature part and high temperature part according to the value of each threshold, count the number of pixels in the normal temperature part and the high temperature part, calculate the mean of the gray value of the pixels in the normal temperature part and the high temperature part, and calculate the inter-class variance of the corresponding threshold. The specific formula is:

[0051] σ(t) 2 =w B (t)*w F (t)*(μ B (t)-μ F (t)) 2

[0052] Among them, σ(t) 2 represents the inter-class variance when the threshold is t, t∈[0, 255], and t∈N + ,w B (t), w F (t) are the number of pixels in the normal temperature part and the high temperature part when the threshold is t, μ B (t), μ F (t) are the mean grayscale values ​​of the pixels in the normal temperature part and the high temperature part when the threshold is t;

[0053] Calculate the threshold value when the inter-class variance reaches the maximum, and define this value as the optimal separation threshold.

[0054] The color threshold processing is performed on the high temperature color image. The specific logic for dividing the high temperature color image into the flame color part and the other color part is as follows: setting the hue threshold range, saturation threshold range and brightness threshold range, defining the area in the high temperature color image that meets the above three threshold ranges as the flame color part, and defining the other parts as the other color part. The threshold range is:

[0055]

[0056] Among them, H is hue, S is saturation, and V is brightness.

[0057] High-temperature objects are not necessarily fire sources. In charging places, in addition to flames, common facilities such as wires, charging facilities in use, and lights also radiate high heat and are divided into high-temperature parts in thermal imaging images. By performing color threshold processing on the high-temperature color image, areas whose colors are obviously not flames can be removed, facilitating accurate identification of fire sources.

[0058] Step 3: Extract features of each flame color part and flame heat part respectively to obtain flame color feature data and flame heat feature data, and obtain the flame visual recognition coefficient by analyzing the flame color feature data; obtain the edge temperature coefficient and the internal temperature coefficient by mathematically analyzing the flame heat feature data;

[0059] The specific logic for generating the flame visual recognition coefficient is: mathematically analyzing the flame color characteristics to generate the flame visual recognition coefficient, wherein the flame color characteristics include internal average saturation, internal saturation variance, and edge circularity; the specific formula for generating the flame visual recognition coefficient is:

[0060]

[0061] Where PS is the flame visual recognition coefficient, is the internal average saturation, and Y is the edge roundness. The flame of the fire source is brighter and has a higher saturation. The higher the value, the greater the probability that the corresponding area is the fire source, and the internal saturation variance σ B Reflects the uniform distribution of internal saturation. The internal saturation of the fire source is uneven, so the internal saturation variance σ B The smaller it is, the greater the probability that the corresponding area is the fire source. The edge circularity Y reflects the regularity of the outline of the corresponding area. Since the outline of the fire source flame is irregular, the smaller the edge circularity Y is, the greater the probability that the corresponding area is the fire source. The flame visual recognition coefficient PS reflects the similarity between the corresponding area and the fire source from the visual angle. The larger its value is, the higher the similarity is, and the greater the probability that the area is the fire source. The generation of this coefficient can provide an important basis for the identification of the fire source.

[0062] The specific logic for generating the edge temperature coefficient and the internal temperature coefficient is: mathematically analyzing the flame thermal characteristic data to obtain the edge temperature coefficient and the internal temperature coefficient. The flame thermal characteristics include the edge temperature gradient variance, the edge average temperature gradient, the edge minimum temperature gradient, the internal temperature variance, the internal average temperature and the internal maximum temperature gradient; the specific formula for generating the edge temperature coefficient is:

[0063]

[0064] Where Te is the edge temperature coefficient, σ e is the edge temperature gradient variance, T e is the average edge temperature gradient, T min is the minimum temperature gradient at the edge; the average temperature gradient at the edge is T eIt reflects the comprehensive changes of the edge area temperature and other parts. The larger the value, the greater the comprehensive changes of the edge area temperature and other parts. Due to the irregularity of the flame combustion, the edge temperature gradient variance σ e The larger the value, the more likely the corresponding area is to be a fire source. The edge temperature coefficient Te reflects the similarity between the edge temperature of the corresponding area and the fire source. The larger the value, the more similar the edge temperature of the corresponding area is to the fire source. The generation of this coefficient can provide an important basis for the identification of fire sources.

[0065] The specific formula used to generate the internal temperature coefficient is:

[0066]

[0067] Where Tn is the internal temperature coefficient, σ n is the internal temperature variance, T n is the average internal temperature, T max is the maximum internal temperature gradient; relative to facilities with higher radiation temperature and color characteristics similar to flames, the internal temperature of the fire source is higher and its distribution is relatively uneven; the internal temperature coefficient Tn reflects the similarity between the internal temperature characteristics of the corresponding area and the fire source. The larger its value is, the more similar the edge temperature of the corresponding area is to the fire source. The generation of this coefficient can provide an important basis for the identification of the fire source.

[0068] Step 4: The fire source identification coefficient is obtained by mathematically analyzing the flame visual identification coefficient, the edge temperature coefficient and the internal temperature coefficient, and the fire source identification coefficient is compared with the preset fire source identification threshold. The area greater than the fire source identification threshold is identified as the fire source area, and the area less than or equal to the fire source identification threshold is identified as the non-fire source area.

[0069] The specific logic for generating the fire source identification coefficient is: the fire source identification coefficient is generated by mathematically analyzing the flame visual identification coefficient, the edge temperature coefficient and the internal temperature coefficient; the specific formula for generating the fire source identification coefficient is:

[0070]

[0071] Among them, Ev is the fire source identification coefficient, PS is the flame visual identification coefficient, Te is the edge temperature coefficient, and Tn is the internal temperature coefficient. The fire source identification coefficient Ev reflects the comprehensive similarity between the internal temperature and visual characteristics of the corresponding area and the fire source. The larger the value, the more similar the edge temperature of the corresponding area is to the fire source. The generation of this coefficient can provide an important basis for the identification of the fire source.

[0072] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0073] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination thereof. When implemented by software, the above embodiments may be implemented in whole or in part in the form of a computer program product. Those skilled in the art may appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein may be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed by hardware or software methods depends on the specific application and design constraints of the technical solution.

[0074] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, and may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0075] The above description is only a specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application.

Claims

1. A method for identifying the fire source of a charging place fire based on automatic tracking and positioning by machine vision, characterized in that: The specific steps include: Step 1: Synchronously acquire the thermal imaging image and visible light image of the target area, convert the visible light image from the RGB color space to the HSV color space to obtain a color recognition image, grayscale the thermal imaging image to obtain a temperature recognition image, and map each pixel position of the temperature recognition image and the color recognition image one by one; Step 2: Use the maximum inter-class variance method to divide the temperature recognition image into a high temperature part and a normal temperature part, extract the area of ​​the color recognition image corresponding to the high temperature part to obtain a high temperature color image, perform color threshold processing on the high temperature color image, divide the high temperature color image into a flame color part and other color parts, extract the flame color part as a flame color image, and extract the area of ​​the temperature recognition image corresponding to the flame color part to obtain a flame thermal image; Step 3: Extract features of each flame color part and flame heat part respectively to obtain flame color feature data and flame heat feature data, and obtain the flame visual recognition coefficient by analyzing the flame color feature data; obtain the edge temperature coefficient and the internal temperature coefficient by mathematically analyzing the flame heat feature data; Step 4: The fire source recognition coefficient is obtained by mathematically analyzing the flame visual recognition coefficient, the edge temperature coefficient and the internal temperature coefficient, and the fire source recognition coefficient is compared with a preset fire source recognition threshold, and the area greater than the fire source recognition threshold is identified as the fire source area, and the area less than or equal to the fire source recognition threshold is identified as the non-fire source area; The specific logic for generating the flame visual recognition coefficient is: mathematically analyzing the flame color characteristics to generate the flame visual recognition coefficient, wherein the flame color characteristics include internal average saturation, internal saturation variance, and edge circularity; the specific formula for generating the flame visual recognition coefficient is: Where PS is the flame visual recognition coefficient, is the internal average saturation, σ B is the internal saturation variance, Y is the edge circularity; The specific logic for generating the edge temperature coefficient and the internal temperature coefficient is as follows: the edge temperature coefficient and the internal temperature coefficient are obtained by mathematically analyzing the flame thermal characteristic data, wherein the flame thermal characteristics include the edge temperature gradient variance, the edge average temperature gradient, the edge minimum temperature gradient, the internal temperature variance, the internal average temperature and the internal maximum temperature gradient; the specific formula for generating the edge temperature coefficient is as follows: Where Te is the edge temperature coefficient, σ e is the edge temperature gradient variance, T e is the average edge temperature gradient, T min is the minimum temperature gradient at the edge; The specific formula used to generate the internal temperature coefficient is: Where Tn is the internal temperature coefficient, σ n is the internal temperature variance, T n is the average internal temperature, T max is the maximum internal temperature gradient; The specific logic for generating the fire source identification coefficient is: the fire source identification coefficient is generated by mathematically analyzing the flame visual identification coefficient, the edge temperature coefficient and the internal temperature coefficient; the specific formula for generating the fire source identification coefficient is: Among them, Ev is the fire source identification coefficient, PS is the flame visual identification coefficient, Te is the edge temperature coefficient, and Tn is the internal temperature coefficient.

2. According to claim 1, a method for identifying the fire source of a charging place fire based on automatic tracking and positioning by machine vision, characterized in that: The coordinates of the pixels of the temperature recognition image and the color recognition image are marked, with the bottom pixel row as the first row and the leftmost pixel column as the first column. The pixel positions of the temperature recognition image and the color recognition image are mapped one by one through the coordinates, so that each pixel has a unique coordinate value.

3. According to claim 1, a method for identifying the fire source of a charging place fire based on automatic tracking and positioning by machine vision, characterized in that: The specific logic of using the maximum inter-class variance method to divide the temperature recognition image into a high temperature part and a normal temperature part is as follows: the frequency of the pixel points of each level in the 256 gray levels is counted to form a gray histogram, and the inter-class variance of each threshold in the histogram is calculated, and the optimal separation threshold is selected through the inter-class variance, and the temperature recognition image is divided into a high temperature part and a normal temperature part through the optimal separation threshold; The specific logic for calculating the optimal separation threshold is: traverse all possible thresholds, divide the construction site image into normal temperature part and high temperature part according to the value of each threshold, count the number of pixels in the normal temperature part and the high temperature part, calculate the mean of the gray value of the pixels in the normal temperature part and the high temperature part, and calculate the inter-class variance of the corresponding threshold. The specific formula is: σ(t) 2 =w B (t)*w F (t)*(μ B (t)-m F (t)) 2 Among them, σ(t) 2 represents the inter-class variance when the threshold is t, t∈[0, 255], and t∈N + ,w B (t), w F (t) are the number of pixels in the normal temperature part and the high temperature part when the threshold is t, μ B (t), μ F (t) are the mean grayscale values ​​of the pixels in the normal temperature part and the high temperature part when the threshold is t; Calculate the threshold value when the inter-class variance reaches the maximum, and define this value as the optimal separation threshold.

4. According to claim 1, a method for identifying the fire source of a charging place fire based on automatic tracking and positioning by machine vision, characterized in that: The color threshold processing is performed on the high temperature color image. The specific logic for dividing the high temperature color image into the flame color part and the other color part is as follows: setting the hue threshold range, saturation threshold range and brightness threshold range, defining the area in the high temperature color image that meets the above three threshold ranges as the flame color part, and defining the other parts as the other color part. The threshold range is: Among them, H is hue, S is saturation, and V is brightness.

Images