Active flame identification and positioning method, system and equipment

Abstract

The invention discloses an active flame identification and positioning method, system and device, and relates to the technical field of fire identification. The active flame identification and positioning method comprises the following steps: S1, regional feature evaluation; s2, flame authenticity evaluation; and S3, flame positioning accuracy evaluation. According to the method, the region feature evaluation value is obtained through the obtained flame feature data, whether the to-be-detected flame target exists in the to-be-monitored region is judged based on the obtained region feature evaluation value, then the flame authenticity evaluation value is obtained according to the obtained flame dynamic data, and whether flame positioning is carried out or not is judged based on the obtained flame authenticity evaluation value. And finally, obtaining a flame positioning accurate evaluation value according to the obtained flame positioning accurate data, and judging whether flame positioning optimization is performed or not based on the obtained flame positioning accurate evaluation value, thereby achieving the effect of improving the flame identification and positioning accuracy, and solving the problem of low data accuracy in the flame identification and positioning process in the prior art.

Description

Technical Field

[0001] The present invention relates to the technical field of fire identification, and in particular to an active flame identification and positioning method, system and equipment. Background Art

[0002] In various warehouses, forests and other environments, timely detection and accurate positioning of fires are crucial to reducing losses. Traditional fire detection methods such as smoke detectors have problems such as long response time and susceptibility to environmental interference. Thermal imaging technology can identify heat sources by detecting the thermal radiation of objects, but monocular thermal imaging has limited accuracy in flame positioning. Therefore, a more advanced flame identification and positioning system is needed to improve the efficiency and accuracy of fire detection. With the advancement of infrared imaging and other technologies, the accuracy and response speed of sensors have been greatly improved, making flame identification and positioning possible. The application of new sensing technology has greatly improved the accuracy of fire source identification and positioning, especially in complex environments, and can provide more accurate and timely fire warnings.

[0003] In the prior art, flame characteristic signals are collected by various sensors and then analyzed by algorithms to achieve accurate judgment of the presence and position of the flame.

[0004] For example, the invention patent with announcement number: CN117876874B discloses a forest fire detection and positioning method and system based on high-point monitoring videos, including: constructing a forest fire detection data set based on the acquired high-point monitoring videos, and training the constructed multi-scale and multi-dimensional feature extraction network that integrates video spatial features and temporal features to identify the detection frames of smoke and flames in the high-point monitoring videos to be identified, and finally obtaining the detection frames of smoke objects and flame objects and the center two-dimensional pixel coordinates of the flame objects in each frame image; accurately positioning forest fires driven by a video stereo grid based on the detection frames of smoke objects and flame objects and the center two-dimensional pixel coordinates of the flame objects in each frame image.

[0005] For example, a patent application with publication number: CN116778423A discloses an artificial intelligence-based forest fire detection method, which includes the following steps: S1, respectively obtaining a video image and an infrared thermal image at the same position in the forest; S2, using a convolutional neural network to extract features of the video image; S3, classifying and identifying flames through a least squares support vector machine to obtain a first flame position; S4, using an FCN fully convolutional neural network to perform flame judgment on the highest temperature point of the infrared thermal image to obtain a second flame position; S5, judging the first flame position and the second flame position through a Gaussian distribution to obtain the true position of the flame, thereby completing forest fire detection.

[0006] However, in the process of implementing the technical solution of the invention in the embodiments of the present application, the present application found that the above technology has at least the following technical problems:

[0007] In the prior art, in complex environments, for example, in a forest, sunlight may cause interference to the sensor, affecting the accurate identification of the flame, and the sensor may not be able to directly capture the flame signal because the flame may be blocked by an object or the sensor signal may be reflected, making the position identification of the flame more difficult, resulting in the problem of low data accuracy in the flame identification and positioning process. Summary of the invention

[0008] The embodiments of the present application provide an active flame identification and positioning method, system and device, thereby solving the problem of low data accuracy in the flame identification and positioning process in the prior art and improving the accuracy of flame identification and positioning.

[0009] The embodiment of the present application provides an active flame identification and positioning method, comprising the following steps: evaluating a to-be-monitored area according to acquired flame characteristic data to obtain a regional characteristic evaluation value, judging whether there is a to-be-detected flame target in the to-be-monitored area based on the acquired regional characteristic evaluation value, and the regional characteristic evaluation value is used to quantitatively evaluate the probability that the to-be-monitored area contains a flame; if there is a to-be-detected flame target, evaluating the to-be-detected flame target according to acquired flame dynamic data to obtain a flame authenticity evaluation value, judging whether to perform flame positioning based on the acquired flame authenticity evaluation value, and the flame authenticity evaluation value is used to quantitatively evaluate the authenticity of the existence of the to-be-detected flame target; if flame positioning is to be performed, evaluating the acquired flame positioning accuracy data to obtain a flame positioning accuracy evaluation value, judging whether to perform flame positioning optimization based on the acquired flame positioning accuracy evaluation value, and the flame positioning accuracy evaluation value is used to quantitatively evaluate the accuracy of flame positioning.

[0010] Furthermore, the flame characteristic data include temperature gradient, ultraviolet radiation intensity, ultraviolet wavelength, smoke concentration, smoke particles, carbon monoxide concentration and carbon dioxide concentration; the flame dynamic data include flame flickering frequency, flame area change rate and flame shape complexity; the flame positioning accuracy data include image signal-to-noise ratio, parallax, baseline distance and matching error value; the flame flickering frequency is obtained by extracting and analyzing the flame flickering signal after analyzing the brightness change of the flame target to be detected by image processing software; the flame shape complexity is obtained by extracting the contour of the high-temperature area through an edge detection algorithm and then analyzing the complexity of the contour.

[0011] Furthermore, the specific method for evaluating the monitored area according to the acquired flame characteristic data to obtain the regional characteristic evaluation value is as follows: obtaining the temperature gradient deviation according to the relative relationship between the temperature gradient and the preset temperature gradient in the database; obtaining the ultraviolet coefficient according to the relative relationship between the ultraviolet wavelength factor, the ultraviolet radiation intensity and the preset ultraviolet radiation intensity in the database; obtaining the smoke concentration deviation according to the relative relationship between the smoke concentration and the preset minimum smoke concentration in the database; obtaining the smoke particle deviation according to the relative relationship between the smoke particles and the preset minimum smoke particles in the database, and processing the smoke concentration deviation and the smoke particle deviation to obtain the smoke coefficient; obtaining the carbon monoxide concentration deviation according to the relative relationship between the carbon monoxide concentration and the preset carbon monoxide concentration range in the database; obtaining the carbon dioxide concentration deviation according to the relative relationship between the carbon dioxide concentration and the preset carbon dioxide concentration in the database, and processing the carbon monoxide concentration deviation and the carbon dioxide concentration deviation to obtain the gas coefficient; processing the temperature gradient deviation, ultraviolet coefficient, smoke coefficient and gas coefficient to obtain the regional characteristic evaluation value.

[0012] Furthermore, the specific process of judging whether there is a flame target to be detected in the monitored area based on the acquired regional characteristic evaluation value is as follows: judging whether the regional characteristic evaluation value is greater than the preset regional characteristic threshold in the database: if the regional characteristic evaluation value is greater than the preset regional characteristic threshold in the database, there is a flame target to be detected in the monitored area, and the flame target to be detected is evaluated according to the acquired flame dynamic data to obtain a flame authenticity evaluation value; if the regional characteristic evaluation value is not greater than the preset regional characteristic threshold in the database, there is no flame target to be detected.

[0013] Furthermore, the specific method for evaluating the flame target to be detected based on the acquired flame dynamic data to obtain a flame authenticity evaluation value is as follows: the flame flicker frequency deviation is obtained based on the relative relationship between the flame flicker frequency and the preset flame flicker frequency in the database; the flame area change rate deviation is obtained based on the relative relationship between the flame area change rate and the preset flame area change rate in the database; the flame shape complexity deviation is obtained based on the relative relationship between the flame shape complexity and the preset flame shape complexity in the database; the regional feature evaluation value, the flame flicker frequency deviation, the flame area change rate deviation and the flame shape complexity deviation are processed to obtain the flame authenticity evaluation value.

[0014] Furthermore, the specific process of judging whether to perform flame positioning based on the obtained flame authenticity evaluation value is as follows: judging whether the flame authenticity evaluation value is greater than a preset flame authenticity threshold in the database: if the flame authenticity evaluation value is greater than the preset flame authenticity threshold in the database, flame positioning is performed; if the flame authenticity evaluation value is not greater than the preset flame authenticity threshold in the database, flame positioning is not performed.

[0015] Further, the specific method for evaluating the flame positioning accuracy data obtained to obtain the flame positioning accuracy evaluation value is as follows: the image signal-to-noise ratio deviation is obtained according to the relative relationship between the image signal-to-noise ratio and the preset image signal-to-noise ratio in the database; the parallax deviation is obtained according to the relative relationship between the parallax and the preset parallax in the database; the baseline distance deviation is obtained according to the relative relationship between the baseline distance and the preset baseline distance in the database; the matching error value coefficient is obtained according to the relative relationship between the matching error value and the preset matching error value in the database; the image signal-to-noise ratio deviation, the parallax deviation, the baseline distance deviation and the matching error value coefficient are processed to obtain the flame positioning accuracy evaluation value; the method for obtaining the flame positioning accuracy evaluation value is:

[0016]

[0017] Where m represents the number of the flame target to be detected, m = 1, 2, ..., M, M represents the total number of flame targets to be detected, DWZ m It represents the flame location accuracy evaluation value of the mth flame target to be detected, ZSh m Represents the flame authenticity evaluation value of the mth flame target to be detected, TXZ m SCa represents the image signal-to-noise ratio of the mth flame target to be detected. m represents the parallax of the mth flame target to be detected, JXJ m Indicates the baseline distance of the mth flame target to be detected, PWC m represents the matching error value of the mth flame target to be detected, TXZ0 represents the preset image signal-to-noise ratio, SCa0 represents the preset parallax, JXJ0 represents the preset baseline distance, and PWC0 represents the preset matching error value.

[0018] Furthermore, the specific process of judging whether to perform flame positioning optimization based on the obtained flame positioning accuracy evaluation value is as follows: judging whether the flame positioning accuracy evaluation value is greater than a preset positioning accuracy threshold in a database: if the flame positioning accuracy evaluation value is greater than the preset positioning accuracy threshold in a database, flame positioning optimization is not performed; if the flame positioning accuracy evaluation value is not greater than the preset positioning accuracy threshold in a database, flame positioning optimization is performed; the flame positioning optimization includes replacing a binocular thermal imaging camera, calibrating a binocular thermal imaging camera, optimizing a binocular thermal imaging camera layout, performing image preprocessing, and optimizing flame feature extraction and matching.

[0019] The embodiment of the present application provides an active flame recognition and positioning system, including: a regional feature evaluation module, a flame authenticity evaluation module and a flame positioning accuracy evaluation module; wherein the regional feature evaluation module is used to evaluate the monitored area according to the acquired flame feature data to obtain a regional feature evaluation value, and judge whether there is a flame target to be detected in the monitored area based on the acquired regional feature evaluation value, and the regional feature evaluation value is used to quantitatively evaluate the probability that the monitored area contains a flame; the flame authenticity evaluation module is used to evaluate the flame target to be detected according to the acquired flame dynamic data to obtain a flame authenticity evaluation value if there is a flame target to be detected, and judge whether to perform flame positioning based on the acquired flame authenticity evaluation value, and the flame authenticity evaluation value is used to quantitatively evaluate the authenticity of the existence of the flame target to be detected; the flame positioning accuracy evaluation module is used to evaluate according to the acquired flame positioning accuracy data to obtain a flame positioning accuracy evaluation value if flame positioning is performed, and judge whether to perform flame positioning optimization based on the acquired flame positioning accuracy evaluation value, and the flame positioning accuracy evaluation value is used to quantitatively evaluate the accuracy of flame positioning.

[0020] An embodiment of the present application provides an electronic device, comprising a memory for storing computer program instructions and a processor for executing the program instructions, wherein when the computer program instructions are executed by the processor, the electronic device is triggered to execute the active flame identification and positioning method.

[0021] One or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages:

[0022] 1. The regional characteristic evaluation value obtained by the acquired flame characteristic data is used to determine whether there is a flame target to be detected in the monitored area, and then the flame authenticity evaluation value obtained based on the acquired flame dynamic data is used to determine whether flame positioning is to be performed. Finally, the flame positioning accuracy evaluation value obtained based on the acquired flame positioning accuracy data is used to determine whether flame positioning optimization is to be performed, thereby improving the accuracy of the flame dynamic data, and further achieving an improvement in the accuracy of flame recognition and positioning, effectively solving the problem of low data accuracy in the flame recognition and positioning process in the prior art.

[0023] 2. The flame flicker frequency deviation is obtained by the flame flicker frequency and the preset flame flicker frequency, and then the flame area change rate deviation is obtained according to the flame area change rate and the preset flame area change rate. Then the flame shape complexity deviation is obtained according to the flame shape complexity and the preset flame shape complexity. Finally, the regional feature evaluation value, flame flicker frequency deviation, flame area change rate deviation and flame shape complexity deviation are processed to obtain the flame authenticity evaluation value, thereby quantitatively evaluating the authenticity of the flame target to be detected, and providing a reliable basis for subsequent flame positioning.

[0024] 3. The image signal-to-noise ratio deviation is obtained through the image signal-to-noise ratio and the preset image signal-to-noise ratio, and then the parallax deviation is obtained according to the parallax and the preset parallax, and then the baseline distance deviation is obtained according to the baseline distance and the preset baseline distance, and then the matching error value coefficient is obtained according to the matching error value and the preset matching error value, and finally the image signal-to-noise ratio deviation, parallax deviation, baseline distance deviation and matching error value coefficient are processed to obtain the accurate evaluation value of flame positioning, thereby quantitatively evaluating the accuracy of flame positioning, and then achieving the improvement of the accuracy of flame positioning. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 A flow chart of an active flame identification and positioning method provided in an embodiment of the present application;

[0026] Figure 2 A schematic diagram showing the variation of the flame location accuracy evaluation value with the image signal-to-noise ratio provided in an embodiment of the present application;

[0027] Figure 3 A schematic diagram showing the variation of the flame positioning accuracy evaluation value with the variation of parallax provided in an embodiment of the present application;

[0028] Figure 4 A schematic diagram showing the variation of the flame positioning accuracy evaluation value with the baseline distance provided in the embodiment of the present application;

[0029] Figure 5 A schematic diagram showing the variation of the flame positioning accuracy evaluation value with the matching error value provided in the embodiment of the present application;

[0030] Figure 6 A schematic diagram of the structure of an active flame recognition and positioning system provided in an embodiment of the present application. DETAILED DESCRIPTION

[0031] The embodiments of the present application solve the problem of low data accuracy in the flame identification and positioning process in the prior art by providing an active flame identification and positioning method, system and device. A regional feature evaluation value is obtained by obtaining flame feature data. Based on the obtained regional feature evaluation value, it is judged whether there is a flame target to be detected in the monitored area. Then, a flame authenticity evaluation value is obtained according to the obtained flame dynamic data. Based on the obtained flame authenticity evaluation value, it is judged whether flame positioning is to be performed. Finally, a flame positioning accuracy evaluation value is obtained according to the obtained flame positioning accurate data. Based on the obtained flame positioning accuracy evaluation value, it is judged whether flame positioning optimization is to be performed, thereby improving the accuracy of flame identification and positioning.

[0032] The technical solution in the embodiment of the present application is to solve the problem of low data accuracy in the flame recognition and positioning process. The overall idea is as follows:

[0033] The obtained regional feature evaluation value is used to determine whether there is a flame target to be detected in the monitored area, and then the flame authenticity evaluation value is used to determine whether to locate the flame. Finally, the flame positioning accuracy evaluation value is used to determine whether to optimize the flame positioning, thereby achieving the effect of improving the accuracy of flame recognition and positioning.

[0034] In order to better understand the above technical solution, the above technical solution will be described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0035] like Figure 1 As shown, it is a flow chart of an active flame identification and positioning method provided by an embodiment of the present application, and the method includes the following steps: S1, regional feature evaluation: evaluating the monitored area according to the acquired flame feature data to obtain a regional feature evaluation value, judging whether there is a flame target to be detected in the monitored area based on the acquired regional feature evaluation value, and the regional feature evaluation value is used to quantitatively evaluate the probability that the monitored area contains a flame; S2, flame authenticity evaluation: if there is a flame target to be detected, evaluating the flame target to be detected according to the acquired flame dynamic data to obtain a flame authenticity evaluation value, judging whether to perform flame positioning based on the acquired flame authenticity evaluation value, and the flame authenticity evaluation value is used to quantitatively evaluate the authenticity of the existence of the flame target to be detected; S3, flame positioning accuracy evaluation: if flame positioning is performed, evaluating according to the acquired flame positioning accurate data to obtain a flame positioning accurate evaluation value, judging whether to perform flame positioning optimization based on the acquired flame positioning accurate evaluation value, and the flame positioning accurate evaluation value is used to quantitatively evaluate the accuracy of flame positioning.

[0036] It should be added that flame characteristic data include temperature gradient, ultraviolet radiation intensity, ultraviolet wavelength, smoke concentration, smoke particles, carbon monoxide concentration and carbon dioxide concentration; flame dynamic data include flame flickering frequency, flame area change rate and flame shape complexity; flame positioning accuracy data include image signal-to-noise ratio, parallax, baseline distance and matching error value; flame flickering frequency is obtained by extracting and analyzing the flame flickering signal after analyzing the brightness changes of the flame target to be detected through image processing software; flame shape complexity is obtained by extracting the contour of the high-temperature area through the edge detection algorithm and then analyzing the complexity of the contour.

[0037] It should be added that the temperature gradient is obtained by a thermal imager; the ultraviolet radiation intensity is obtained by an ultraviolet radiometer; the ultraviolet wavelength is obtained by a spectrometer; the smoke concentration is obtained by a smoke detector; the smoke particles are obtained by a laser particle size analyzer; the carbon monoxide concentration and the carbon dioxide concentration are obtained by a gas detector.

[0038] The flame area change rate is obtained by identifying and extracting the outline of the flame area through image processing software (eg, the open source computer vision library OpenCV), and calculating the ratio of the areas of the flame areas at adjacent time points in a continuous time series.

[0039] The image signal-to-noise ratio represents the sum of the signal-to-noise ratios obtained by processing the ratio of the pixel value mean and standard deviation of the two flame images.

[0040] The parallax is obtained by calculating the difference in the position of the flame in the two flame images; the baseline distance refers to the distance between the two camera lenses in the binocular thermal imaging camera.

[0041] The matching error value is represented by the matching result obtained by matching the feature points in the flame image captured simultaneously by the binocular thermal imaging camera.

[0042] In this embodiment, the flame flicker frequency refers to the frequency of the light and dark changes of the flame caused by unstable combustion; the flame shape complexity refers to the complexity of the flame shape. The shape of the flame is usually irregular and has no fixed geometric shape; the flame shape complexity can reflect the combustion state and stability of the flame; through the comprehensive evaluation of multiple sensor data, the existence and authenticity of the flame can be accurately judged, and accurate flame positioning can be achieved; through the flame positioning accuracy evaluation, the positioning results can be optimized to improve the reliability and accuracy of flame detection and positioning, thereby achieving an improvement in the accuracy of flame recognition and positioning.

[0043] Further, the specific method for evaluating the monitored area according to the acquired flame characteristic data to obtain the regional characteristic evaluation value is as follows: the temperature gradient deviation (i.e. ), the temperature gradient deviation is obtained by ratio processing the temperature gradient and the preset temperature gradient in the database; the ultraviolet coefficient (i.e. ZWa) is obtained according to the relative relationship between the ultraviolet wavelength factor, the ultraviolet radiation intensity and the preset ultraviolet radiation intensity in the database. i ), the ultraviolet coefficient is obtained by processing the ultraviolet radiation intensity and the preset ultraviolet radiation intensity in the database; the smoke concentration deviation (i.e. ), the smoke concentration deviation is obtained by ratio processing between the smoke concentration and the preset minimum smoke concentration in the database; the smoke particle deviation (i.e. ), the smoke particle deviation is obtained by processing the ratio of the smoke particles to the preset minimum smoke particles in the database, and the smoke concentration deviation and the smoke particle deviation are processed to obtain the smoke coefficient (i.e. YWu i); obtain the carbon monoxide concentration deviation according to the relative relationship between the carbon monoxide concentration and the preset carbon monoxide concentration range in the database (when the carbon monoxide concentration is less than the preset minimum carbon monoxide concentration, the carbon monoxide concentration deviation is When the carbon monoxide concentration is greater than the preset maximum carbon monoxide concentration, the carbon monoxide concentration deviation is When carbon monoxide is within the preset carbon monoxide concentration range, the carbon monoxide concentration deviation is 0), and the carbon monoxide concentration deviation is obtained by processing the carbon monoxide concentration and the preset carbon monoxide concentration range in the database; the carbon dioxide concentration deviation is obtained according to the relative relationship between the carbon dioxide concentration and the preset carbon dioxide concentration in the database (i.e. ), the carbon monoxide concentration deviation and the carbon dioxide concentration deviation are processed to obtain the gas coefficient; the temperature gradient deviation, ultraviolet coefficient, smoke coefficient and gas coefficient are processed to obtain the regional characteristic evaluation value.

[0044] Among them, the specific method for obtaining the regional characteristic evaluation value is:

[0045]

[0046] Where i represents the number of the area to be monitored, i = 1, 2, ..., N, N represents the total number of areas to be monitored, QYu i represents the regional characteristic evaluation value of the i-th monitored area, WDT i represents the temperature gradient of the ith monitored area, ZWa i represents the ultraviolet coefficient of the ith area to be monitored, YWu i Indicates the smoke coefficient of the i-th monitored area, QTI i represents the gas coefficient of the i-th area to be monitored, WDT0 represents the preset temperature gradient, ZFQ i represents the ultraviolet radiation intensity of the i-th monitored area, ZFQ0 represents the preset ultraviolet radiation intensity, α represents the ultraviolet wavelength factor, YND i represents the smoke concentration of the ith monitored area, YLZ i represents the smoke particles in the i-th area to be monitored, YND min Indicates the preset minimum smoke density, YLZ min Indicates the preset minimum smoke particles, YYD i Indicates the carbon monoxide concentration in the i-th monitored area, EYD i represents the carbon dioxide concentration in the i-th area to be monitored, YYD min Indicates the preset minimum carbon monoxide concentration, YYD max Indicates the preset maximum carbon monoxide concentration, and EYD0 indicates the preset carbon dioxide concentration.

[0047] In this embodiment, the preset temperature gradient is set according to the safety standard, for example, the preset temperature gradient is 50%; the preset ultraviolet radiation intensity is represented by the average value of the ultraviolet radiation intensity in the historical time period; the preset minimum smoke concentration is represented by the minimum value of the smoke concentration in the historical time period; the preset minimum smoke particles are represented by the minimum value of the smoke particles in the historical time period; the preset carbon monoxide concentration range and the preset carbon dioxide concentration are set according to the application scenario and safety standards, for example, the preset maximum carbon monoxide concentration range is between 50-1000ppm, then the preset minimum carbon monoxide concentration is 50ppm, and the preset maximum carbon monoxide concentration is 1000ppm; the preset carbon dioxide concentration is 5%.

[0048] The ultraviolet wavelength factor is obtained from the database and represents the influence of the ultraviolet wavelength on the regional characteristic evaluation value. For example, a mapping set of ultraviolet wavelength and its corresponding influencing factor is constructed based on the relationship between historical ultraviolet wavelength and flame characteristic data (such as temperature gradient, ultraviolet radiation intensity), and the real-time ultraviolet wavelength is input into the mapping set to obtain the corresponding ultraviolet wavelength factor.

[0049] The regional characteristic evaluation value in this algorithm involves processing multiple independent variables (temperature gradient, ultraviolet radiation intensity, ultraviolet wavelength, smoke concentration, smoke particles, carbon monoxide concentration and carbon dioxide concentration). There is a mutual influence relationship between these independent variables; the higher the flame temperature, the greater the temperature gradient may be; there is a certain correlation between the flame temperature and ultraviolet radiation intensity. The higher the flame temperature, the stronger the ultraviolet radiation intensity generated; the intensity of ultraviolet radiation is closely related to the wavelength. Generally, the shorter the ultraviolet wavelength (such as UV-C and UV-B bands), the stronger the ultraviolet radiation intensity; the higher the flame temperature, the shorter the ultraviolet wavelength; when the smoke concentration increases, it is often It means that more organic matter is burned and produced, so the increase in smoke concentration is usually accompanied by an increase in carbon monoxide concentration and carbon dioxide concentration; the higher the smoke concentration, the intensity of ultraviolet radiation may be scattered or absorbed by smoke particles, resulting in a decrease in ultraviolet radiation intensity; when the flame temperature increases, the carbon dioxide concentration usually increases as well; in summary, the regional characteristic evaluation value is positively correlated with temperature gradient, ultraviolet radiation intensity, smoke concentration, carbon monoxide concentration and carbon dioxide concentration, and is negatively correlated with smoke particles. When the carbon monoxide concentration is not within the preset carbon monoxide concentration range, the closer the carbon monoxide concentration is to the preset carbon monoxide concentration range, the greater the regional characteristic evaluation value.

[0050] Through the above steps, the probability of the monitored area containing flames is quantitatively evaluated, and then it is determined whether there is a fire risk in the monitored area.

[0051] Furthermore, the specific process of judging whether there is a flame target to be detected in the monitored area based on the obtained regional characteristic evaluation value is as follows: judge whether the regional characteristic evaluation value is greater than the preset regional characteristic threshold in the database: if the regional characteristic evaluation value is greater than the preset regional characteristic threshold in the database, there is a flame target to be detected in the monitored area, and the flame target to be detected is evaluated according to the acquired flame dynamic data to obtain a flame authenticity evaluation value; if the regional characteristic evaluation value is not greater than the preset regional characteristic threshold in the database, there is no flame target to be detected.

[0052] In this embodiment, the preset regional feature threshold is represented by the maximum value of the regional feature evaluation value in the historical time period; through the above steps, the accuracy and reliability of flame detection and identification are achieved.

[0053] Furthermore, the specific method for evaluating the flame target to be detected based on the acquired flame dynamic data to obtain the flame authenticity evaluation value is as follows: the flame flicker frequency deviation (i.e. ); The flame area change rate deviation (i.e. ); obtain the flame shape complexity deviation (i.e. ); the regional feature evaluation value, flame flicker frequency deviation, flame area change rate deviation and flame shape complexity deviation are processed to obtain the flame authenticity evaluation value.

[0054] Among them, the specific method for obtaining the flame authenticity evaluation value is:

[0055]

[0056] Where m represents the number of the flame target to be detected, m = 1, 2, ..., M, M represents the total number of flame targets to be detected, ZSh m represents the flame authenticity evaluation value of the mth flame target to be detected, QYu i represents the regional characteristic evaluation value of the i-th monitored area, SSP m Indicates the flame flicker frequency of the mth flame target to be detected, MJB m represents the flame area change rate of the mth flame target to be detected, XZF m represents the flame shape complexity of the mth flame target to be detected, SSP0 represents the preset flame flickering frequency, MJB0 represents the preset flame area change rate, and XZF0 represents the preset flame shape complexity.

[0057] In this embodiment, the preset flame flickering frequency is represented by the minimum value of the flame flickering frequency in the historical time period; the preset flame area change rate is represented by the minimum value of the flame area change rate in the historical time period; and the preset flame shape complexity is represented by the minimum value of the flame shape complexity in the historical time period.

[0058] The total number of flame targets to be detected is 1, the regional feature evaluation value is 1.435, the preset flame flicker frequency is 1Hz, and the preset flame area change rate is 0.01m 2 / s, taking the preset flame shape complexity as 1.2 as an example, the statistical table of changes in flame authenticity evaluation values ​​is shown in Table 1:

[0059] Table 1 Statistics of changes in flame authenticity evaluation values

[0060]

[0061] It can be seen from the first and second groups of data in the table that when the flame flickering frequency increases, the flame authenticity evaluation value increases accordingly; it can be seen from the second and third groups of data that when the flame area change rate increases, the flame authenticity evaluation value increases accordingly; it can be seen from the third and fourth groups of data that when the flame shape complexity increases, the flame authenticity evaluation value increases accordingly.

[0062] The flame authenticity evaluation value in this algorithm involves processing multiple independent variables (flame flickering frequency, flame area change rate, flame shape complexity), and there is a mutual influence relationship between these independent variables; the flickering of the flame may cause slight changes in the flame area. For example, when the flame flickering frequency increases, the flame area change rate may also increase accordingly; the complexity of the flame shape may affect the stability of the flame area. If the complexity of the flame shape is greater, the flame area change rate may be higher, because the complex shape is more susceptible to external factors; under unstable combustion conditions, the flame flickers; the frequency increases, which may cause slight changes in the flame shape, thereby increasing the complexity of the flame shape; in summary, the flame authenticity evaluation value is positively correlated with the flame flickering frequency, the flame area change rate, and the flame shape complexity.

[0063] Through the above steps, the authenticity of the flame target to be detected is quantitatively evaluated, thereby providing a reliable basis for subsequent flame positioning.

[0064] Furthermore, the specific process of determining whether to perform flame positioning based on the obtained flame authenticity evaluation value is as follows: Determine whether the flame authenticity evaluation value is greater than a preset flame authenticity threshold in the database: if the flame authenticity evaluation value is greater than the preset flame authenticity threshold in the database, flame positioning is performed; if the flame authenticity evaluation value is not greater than the preset flame authenticity threshold in the database, flame positioning is not performed.

[0065] In this embodiment, the preset flame authenticity threshold is represented by the maximum value of the flame authenticity evaluation value in the historical time period.

[0066] The steps of flame positioning are as follows: use a binocular thermal imaging camera to capture the flame image, apply edge detection algorithm and shape analysis algorithm (such as contour extraction algorithm) to the flame image, extract the contour and shape features of the flame (such as area, perimeter, aspect ratio, etc.); match the contour and shape features of the flame in the two thermal imaging images through a feature point matching algorithm (such as scale-invariant feature transform SIFT) to obtain the corresponding positions (x1, y1) and (x2, y2) of the flame in the two flame images; use the principle of triangulation to obtain the flame in the three-dimensional image according to the known positional relationship between the two camera lenses in the binocular thermal imaging camera and the parallax of the flame. The coordinates (X, Y, Z) in the three-dimensional space are as follows: X=B*f / d; Y=(y1-cy1)*B*f / d; Z=f*B / d, wherein X is the horizontal coordinate of the flame in the three-dimensional space, B is the baseline distance between the two binocular thermal imaging cameras, f is the focal length of the binocular thermal imaging camera, d is the parallax, Y is the vertical coordinate of the flame in the three-dimensional space, y1 is the vertical coordinate of the flame in the flame image, cy1 is the vertical coordinate of the optical axis of the binocular thermal imaging camera in the image, and Z is the vertical coordinate of the flame in the three-dimensional space.

[0067] Through the above steps, accurate positioning of the flame target to be detected is achieved.

[0068] Further, the specific method for evaluating the flame positioning accuracy data obtained is as follows: the image signal-to-noise ratio deviation (i.e. ), the image signal-to-noise ratio deviation is obtained by processing the image signal-to-noise ratio and the preset image signal-to-noise ratio in the database; the disparity deviation (i.e. ), the parallax deviation is obtained by processing the parallax and the preset parallax in the database; the baseline distance deviation (i.e. ), the baseline distance deviation is obtained by processing the baseline distance and the preset baseline distance in the database; the matching error value coefficient (i.e. ), the matching error value coefficient is obtained by processing the matching error value and the preset matching error value in the database; the image signal-to-noise ratio deviation, parallax deviation, baseline distance deviation and matching error value coefficient are processed to obtain the flame positioning accuracy evaluation value.

[0069] The method to obtain an accurate assessment value of flame location is:

[0070]

[0071] Where m represents the number of the flame target to be detected, m = 1, 2, ..., M, M represents the total number of flame targets to be detected, DWZ m It represents the flame location accuracy evaluation value of the mth flame target to be detected, ZSh m represents the flame authenticity evaluation value of the mth flame target to be detected (the flame authenticity evaluation value greater than the preset flame authenticity threshold in the database), TXZ m SCa represents the image signal-to-noise ratio of the mth flame target to be detected. m represents the parallax of the mth flame target to be detected, JXJ m Indicates the baseline distance of the mth flame target to be detected, PWC m represents the matching error value of the mth flame target to be detected, TXZ0 represents the preset image signal-to-noise ratio, SCa0 represents the preset parallax, JXJ0 represents the preset baseline distance, and PWC0 represents the preset matching error value.

[0072] In this embodiment, the preset image signal-to-noise ratio is represented by the average value of the image signal-to-noise ratio in the historical time period; the preset disparity is represented by the average value of the disparity in the historical time period; the preset baseline distance is represented by the average value of the baseline distance in the historical time period; and the preset matching error value is represented by the average value of the matching error value in the historical time period.

[0073] In this algorithm, the accurate evaluation value of flame location involves processing multiple independent variables (image signal-to-noise ratio, parallax, baseline distance and matching error value), and there is a mutual influence relationship between these independent variables; the lower the image signal-to-noise ratio, the more noise will interfere with the recognition and matching of feature points, resulting in an increase in the matching error value; an increase in baseline distance may lead to an increase in parallax, making the binocular vision system more sensitive to depth changes of objects; with the increase of baseline distance, the matching error value may increase, because the increase in baseline distance makes the position difference between the projection points on the imaging planes of the two cameras more obvious, and also increases the influence of noise and interference factors on the matching algorithm; in summary, the accurate evaluation value of flame location is positively correlated with the image signal-to-noise ratio, and negatively correlated with the parallax, baseline distance and matching error value.

[0074] For example, the flame authenticity evaluation value is 1, the preset image signal-to-noise ratio is 20 dB, the preset parallax is 5 cm, the preset baseline distance is 10 cm, and the preset matching error value is 0.5 pixels. Figure 2As shown, it is a schematic diagram of the change of the flame positioning accuracy evaluation value provided in the embodiment of the present application with the change of the image signal-to-noise ratio. When the parallax is 5 cm, the baseline distance is 10 cm, and the matching error value is 0.5 pixels, the flame positioning accuracy evaluation value increases with the increase of the image signal-to-noise ratio.

[0075] like Figure 3 The figure shows a schematic diagram of the change of the flame positioning accuracy evaluation value with the change of parallax provided by the embodiment of the present application. When the image signal-to-noise ratio is 50dB, the baseline distance is 10cm, and the matching error value is 0.5 pixels, the flame positioning accuracy evaluation value decreases with the increase of parallax.

[0076] like Figure 4 The figure shows a schematic diagram of the variation of the flame positioning accuracy evaluation value with the baseline distance provided in the embodiment of the present application. When the image signal-to-noise ratio is 50 dB, the parallax is 5 cm, and the matching error value is 0.5 pixels, the flame positioning accuracy evaluation value decreases with the increase of the baseline distance.

[0077] like Figure 5 The figure shows a schematic diagram of the change of the flame positioning accuracy evaluation value with the matching error value provided in the embodiment of the present application. When the image signal-to-noise ratio is 50dB, the parallax is 5cm, and the baseline distance is 10cm, the flame positioning accuracy evaluation value decreases with the increase of the matching error value.

[0078] Through the above steps, the accuracy of flame positioning is quantitatively evaluated, thereby achieving improvement in the accuracy of flame positioning.

[0079] Furthermore, the specific process of judging whether to perform flame positioning optimization based on the obtained flame positioning accuracy evaluation value is as follows: judging whether the flame positioning accuracy evaluation value is greater than the preset positioning accuracy threshold in the database: if the flame positioning accuracy evaluation value is greater than the preset positioning accuracy threshold in the database, flame positioning optimization is not performed; if the flame positioning accuracy evaluation value is not greater than the preset positioning accuracy threshold in the database, flame positioning optimization is performed; flame positioning optimization includes replacing the binocular thermal imaging camera, calibrating the binocular thermal imaging camera, optimizing the binocular thermal imaging camera layout, performing image preprocessing, and optimizing flame feature extraction and matching.

[0080] In this embodiment, the preset positioning accuracy threshold is represented by the maximum value of the flame positioning accuracy evaluation value in the historical time period.

[0081] Specifically, replace the binocular thermal imaging camera: the binocular thermal imaging camera currently in use may have insufficient performance problems, and needs to be replaced with a binocular thermal imaging camera with higher resolution, or use a long-wave infrared thermal imaging sensor; for example, the binocular thermal imaging camera currently used in the factory is a 320x240 resolution binocular thermal imaging camera, resulting in poor image quality, so it needs to be replaced with a 640x480 resolution binocular thermal imaging camera, or use a long-wave infrared thermal imaging sensor.

[0082] Specifically, calibrate the binocular thermal imaging camera: use the checkerboard calibration method to perform geometric calibration on the binocular thermal imaging camera to correct the image distortion caused by the lens; adjust the color temperature of the binocular thermal imaging camera through a black body radiation source to ensure that the displayed temperature is consistent with the actual temperature.

[0083] Specifically, optimize the layout of binocular thermal imaging cameras: select a suitable installation location based on the characteristics and needs of the target monitoring area to ensure that the binocular thermal imaging cameras can cover key areas. For example, for a factory workshop, it can be installed at a high place (ceiling or overhead platform, depending on the application scenario) to ensure comprehensive monitoring; for a forest environment, it can be installed on a high observation tower to ensure comprehensive monitoring; adjust the relative position and angle of the two binocular thermal imaging cameras to have an appropriate baseline distance and field of view overlap area. The baseline distance can be set between 2-4 meters; increase the number of binocular thermal imaging cameras.

[0084] Specifically, image preprocessing: Select appropriate image preprocessing algorithms for different application scenarios and environmental conditions. For example, for environments with high noise, denoising algorithms such as median filtering and Gaussian filtering can be used; for images with low contrast, enhancement algorithms such as histogram equalization and contrast stretching can be used.

[0085] Optimize flame feature extraction and matching: Use machine learning (for example, support vector machines are used to classify and identify the extracted flame feature data and flame dynamic data) and deep learning algorithms (for example, convolutional neural networks learn the deep features of flames, such as color, shape, texture, etc., through structures such as convolutional layers and pooling layers) to train a large number of flame images to improve the accuracy of flame feature extraction and image matching.

[0086] Through the above steps, the flame positioning accuracy is improved.

[0087] like Figure 6As shown, it is a structural schematic diagram of an active flame identification and positioning system provided by an embodiment of the present application. An active flame identification and positioning system provided by an embodiment of the present application includes: a regional feature evaluation module, a flame authenticity evaluation module and a flame positioning accuracy evaluation module; wherein the regional feature evaluation module is used to evaluate the monitored area according to the acquired flame feature data to obtain a regional feature evaluation value, and judge whether there is a flame target to be detected in the monitored area based on the acquired regional feature evaluation value, and the regional feature evaluation value is used to quantitatively evaluate the probability that the monitored area contains a flame; the flame authenticity evaluation module is used to evaluate the flame target to be detected according to the acquired flame dynamic data to obtain a flame authenticity evaluation value if there is a flame target to be detected, and judge whether to perform flame positioning based on the acquired flame authenticity evaluation value, and the flame authenticity evaluation value is used to quantitatively evaluate the authenticity of the existence of the flame target to be detected; the flame positioning accuracy evaluation module is used to evaluate according to the acquired flame positioning accuracy data to obtain a flame positioning accuracy evaluation value if flame positioning is performed, and judge whether to perform flame positioning optimization based on the acquired flame positioning accuracy evaluation value, and the flame positioning accuracy evaluation value is used to quantitatively evaluate the accuracy of flame positioning.

[0088] Among them, an embodiment of the present application also provides an electronic device, which includes a memory for storing computer program instructions and a processor for executing the program instructions, wherein when the computer program instructions are executed by the processor, the electronic device is triggered to execute an active flame identification and positioning method.

[0089] In this embodiment, through regional feature evaluation and flame authenticity evaluation, false alarms and interference can be eliminated more effectively and the accuracy of flame recognition can be improved; through flame positioning accuracy evaluation, positioning errors can be discovered and corrected in time and the accuracy of flame positioning can be improved; thereby achieving an improvement in the accuracy of flame recognition and positioning and improving the efficiency of fire warning and emergency response.

[0090] When the flame authenticity assessment module confirms the existence of flame and the flame location accuracy assessment module ensures the accuracy of flame location, the fire extinguishing strategy is executed; the best fire extinguishing strategy is formulated based on factors such as the flame area size, flame spread speed, distance from surrounding objects and possible fire source types.

[0091] Fire-fighting strategies generally include selecting appropriate fire-extinguishing agents (such as water, foam, dry powder, etc.), determining the type and quantity of fire-fighting equipment, and planning the action routes of fire-fighting personnel. Select fire-fighting agents according to the type of fire source, for example, foam fire extinguishers or dry powder fire extinguishers are selected for oil fires. Select the type of fire-fighting equipment according to the application scenario. For example, fire sprinklers are suitable for scenarios with large spaces and fast fire spread (such as warehouses, large factories, etc.), and can quickly cover a large area. Determine the number and layout of fire-fighting equipment according to the size, spread speed and direction of the flame. Plan the action routes of fire-fighting personnel according to the location of the fire, the size and spread speed of the flame.

[0092] Obtain flame characteristic data and flame dynamic data in real time to determine whether the fire extinguishing strategy is effective. If not, change the type of fire extinguishing agent, increase fire extinguishing equipment and fire extinguishing personnel.

[0093] To summarize, the embodiment of the present application determines whether there is a flame target to be detected in the monitored area through the regional feature evaluation value obtained by the acquired flame feature data, and then determines whether to perform flame positioning based on the flame authenticity evaluation value obtained based on the acquired flame dynamic data. Finally, it determines whether to perform flame positioning optimization based on the flame positioning accuracy evaluation value obtained based on the acquired flame positioning accuracy data, thereby improving the accuracy of the flame dynamic data, and further achieving an improvement in the accuracy of flame recognition and positioning, effectively solving the problem of low data accuracy in the flame recognition and positioning process in the prior art.

[0094] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0095] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0096] These computer program instructions may also be stored in a computer readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0097] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0098] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0099] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. An active flame identification and positioning method, characterized in that: The following steps are involved: S1, evaluating the monitored area according to the acquired flame characteristic data to obtain a regional characteristic evaluation value, and judging whether there is a flame target to be detected in the monitored area based on the acquired regional characteristic evaluation value, wherein the regional characteristic evaluation value is used to quantitatively evaluate the probability that the monitored area contains a flame; S2, if there is a flame target to be detected, the flame target to be detected is evaluated according to the acquired flame dynamic data to obtain a flame authenticity evaluation value, and whether to perform flame positioning is determined based on the acquired flame authenticity evaluation value, wherein the flame authenticity evaluation value is used to quantitatively evaluate the authenticity of the existence of the flame target to be detected; S3, if flame positioning is performed, an evaluation is performed based on the acquired flame positioning accuracy data to obtain a flame positioning accuracy evaluation value, and based on the acquired flame positioning accuracy evaluation value, it is determined whether flame positioning optimization is performed, wherein the flame positioning accuracy evaluation value is used to quantitatively evaluate the accuracy of flame positioning.

2. An active flame identification and positioning method as claimed in claim 1, characterized in that: The flame characteristic data include temperature gradient, ultraviolet radiation intensity, ultraviolet wavelength, smoke concentration, smoke particles, carbon monoxide concentration and carbon dioxide concentration; The flame dynamic data includes flame flicker frequency, flame area change rate and flame shape complexity; The flame positioning accurate data includes image signal-to-noise ratio, parallax, baseline distance and matching error value; The flame flicker frequency is obtained by analyzing the brightness change of the flame target to be detected through image processing software and then extracting and analyzing the flame flicker signal; The flame shape complexity is obtained by extracting the contour of the high temperature area through an edge detection algorithm and then analyzing the complexity of the contour.

3. An active flame identification and positioning method as claimed in claim 2, characterized in that: The specific method of evaluating the monitored area according to the acquired flame characteristic data to obtain the regional characteristic evaluation value is as follows: The temperature gradient deviation is obtained according to the relative relationship between the temperature gradient and the preset temperature gradient in the database; The ultraviolet coefficient is obtained according to the relative relationship between the ultraviolet wavelength factor, the ultraviolet radiation intensity and the preset ultraviolet radiation intensity in the database; The smoke concentration deviation is obtained according to the relative relationship between the smoke concentration and the preset minimum smoke concentration in the database; The smoke particle deviation is obtained according to the relative relationship between the smoke particles and the preset minimum smoke particles in the database, and the smoke concentration deviation and the smoke particle deviation are processed to obtain the smoke coefficient; Obtaining a carbon monoxide concentration deviation according to a relative relationship between the carbon monoxide concentration and a preset carbon monoxide concentration range in a database; The carbon dioxide concentration deviation is obtained according to the relative relationship between the carbon dioxide concentration and the preset carbon dioxide concentration in the database, and the carbon monoxide concentration deviation and the carbon dioxide concentration deviation are processed to obtain the gas coefficient; The temperature gradient deviation, ultraviolet coefficient, smoke coefficient and gas coefficient are processed to obtain the regional characteristic evaluation value.

4. An active flame identification and positioning method as claimed in claim 3, characterized in that: The specific process of judging whether there is a flame target to be detected in the monitored area based on the acquired regional feature evaluation value is as follows: Determine whether the regional feature evaluation value is greater than the preset regional feature threshold in the database: If the regional feature evaluation value is greater than the preset regional feature threshold in the database, there is a flame target to be detected in the monitored area, and the flame target to be detected is evaluated according to the acquired flame dynamic data to obtain a flame authenticity evaluation value; If the regional feature evaluation value is not greater than the preset regional feature threshold in the database, there is no flame target to be detected.

5. The active flame identification and positioning method according to claim 2, characterized in that: The specific method of evaluating the flame target to be detected based on the acquired flame dynamic data to obtain the flame authenticity evaluation value is as follows: Obtaining a flame flicker frequency deviation according to a relative relationship between the flame flicker frequency and a preset flame flicker frequency in a database; Obtaining a flame area change rate deviation according to a relative relationship between the flame area change rate and a preset flame area change rate in a database; Obtaining a flame shape complexity deviation according to a relative relationship between the flame shape complexity and a preset flame shape complexity in a database; The regional feature evaluation value, flame flicker frequency deviation, flame area change rate deviation and flame shape complexity deviation are processed to obtain the flame authenticity evaluation value.

6. An active flame identification and positioning method as claimed in claim 5, characterized in that: The specific process of determining whether to perform flame positioning based on the obtained flame authenticity evaluation value is as follows: Determine whether the flame authenticity evaluation value is greater than the preset flame authenticity threshold in the database: If the flame authenticity evaluation value is greater than the preset flame authenticity threshold in the database, flame positioning is performed; If the flame authenticity evaluation value is not greater than the preset flame authenticity threshold in the database, flame positioning is not performed.

7. An active flame identification and positioning method as claimed in claim 2, characterized in that: The specific method for evaluating the flame positioning accuracy data obtained to obtain the flame positioning accuracy evaluation value is as follows: The image signal-to-noise ratio deviation is obtained according to the relative relationship between the image signal-to-noise ratio and the preset image signal-to-noise ratio in the database; Obtaining a parallax deviation according to a relative relationship between the parallax and a preset parallax in a database; Obtaining a baseline distance deviation according to a relative relationship between the baseline distance and a preset baseline distance in a database; Obtaining a matching error value coefficient according to a relative relationship between the matching error value and a preset matching error value in a database; The image signal-to-noise ratio deviation, parallax deviation, baseline distance deviation and matching error value coefficient are processed to obtain the accurate evaluation value of flame location; The method to obtain an accurate assessment value of flame location is: Where m represents the number of the flame target to be detected, m = 1, 2, ..., M, M represents the total number of flame targets to be detected, DWZ m It represents the flame location accuracy evaluation value of the mth flame target to be detected, ZSh m Represents the flame authenticity evaluation value of the mth flame target to be detected, TXZ m SCa represents the image signal-to-noise ratio of the mth flame target to be detected. m represents the parallax of the mth flame target to be detected, JXJ m Indicates the baseline distance of the mth flame target to be detected, PWC m represents the matching error value of the mth flame target to be detected, TXZ0 represents the preset image signal-to-noise ratio, SCa0 represents the preset parallax, JXJ0 represents the preset baseline distance, and PWC0 represents the preset matching error value.

8. An active flame identification and positioning method as claimed in claim 7, characterized in that: The specific process of judging whether to perform flame positioning optimization based on the acquired flame positioning accuracy evaluation value is as follows: Determine whether the flame positioning accuracy assessment value is greater than the preset positioning accuracy threshold in the database: If the flame positioning accuracy evaluation value is greater than the preset positioning accuracy threshold in the database, the flame positioning optimization is not performed; If the flame positioning accuracy evaluation value is not greater than the preset positioning accuracy threshold in the database, flame positioning optimization is performed; The flame positioning optimization includes replacing the binocular thermal imaging camera, calibrating the binocular thermal imaging camera, optimizing the layout of the binocular thermal imaging camera, performing image preprocessing, and optimizing flame feature extraction and matching.

9. An active flame identification and positioning system, characterized in that: The active flame recognition and positioning system comprises: a regional feature evaluation module, a flame authenticity evaluation module and a flame positioning accuracy evaluation module; The regional feature evaluation module is used to evaluate the monitored area according to the acquired flame feature data to obtain a regional feature evaluation value, and to determine whether there is a flame target to be detected in the monitored area based on the acquired regional feature evaluation value. The regional feature evaluation value is used to quantitatively evaluate the probability that the monitored area contains a flame. The flame authenticity assessment module is used to evaluate the flame target to be detected according to the acquired flame dynamic data to obtain a flame authenticity assessment value if there is a flame target to be detected, and judge whether to perform flame positioning based on the acquired flame authenticity assessment value, and the flame authenticity assessment value is used to quantitatively evaluate the authenticity of the existence of the flame target to be detected; The flame positioning accuracy evaluation module is used to evaluate flame positioning accuracy data to obtain a flame positioning accuracy evaluation value when flame positioning is performed, and to determine whether to perform flame positioning optimization based on the obtained flame positioning accuracy evaluation value. The flame positioning accuracy evaluation value is used to quantitatively evaluate the accuracy of flame positioning.

10. An electronic device, characterized in that: The electronic device comprises a memory for storing computer program instructions and a processor for executing the program instructions, wherein when the computer program instructions are executed by the processor, the electronic device is triggered to execute an active flame identification and positioning method as claimed in any one of claims 1 to 8.

Images