An intelligent electrical fire monitoring system based on the Internet of Things

Abstract

This invention provides an intelligent electrical fire monitoring system based on the Internet of Things (IoT), comprising: determining a target image by detecting temperature and identifying flames; segmenting the target image using a segmentation method; and then calculating the hazard intensity value of each target area by calculating the features of the pixels. The beneficial effects of this invention are: it enables the judgment of hazard intensity values ​​based on segmented areas, thereby allowing for tiered fire suppression based on the hazard intensity values ​​of each target area, thus preventing greater damage to the area where the equipment is located and greatly protecting the safety of that area.

Description

Technical Field

[0001] This invention relates to the field of artificial intelligence, and in particular to an intelligent electrical fire monitoring system based on the Internet of Things. Background Technology

[0002] In a society where mechanical intelligence is constantly advancing, electrical equipment in factories may very well be left unattended. Because electrical equipment can experience short circuits, overvoltage, and other fire hazards under certain conditions, current technology typically involves directly spraying pre-installed fire extinguishing devices at the base of the flames to lower the ignition point and extinguish the fire. However, the fire extinguishing devices in the area may not be sufficient to completely extinguish the fire, leaving other equipment in the same area at risk of fire and causing further damage. Summary of the Invention

[0003] The main objective of this invention is to provide an intelligent electrical fire monitoring system based on the Internet of Things, which aims to solve the problem that if the fire extinguishing devices in the area where the equipment is located are insufficient to completely extinguish the fire, other equipment in the area may still be at risk of fire.

[0004] This invention provides an intelligent electrical fire monitoring system based on the Internet of Things, comprising:

[0005] An acquisition module is used to acquire the temperature of each device in the area where the device is located through a preset temperature sensor array; wherein, the temperature sensor array is connected to each device in the area where the device is located, and is used to acquire the temperature of each device;

[0006] The acquisition module is used to acquire multiple photos by capturing frames of the target device at preset time intervals through a camera when the temperature of the target device exceeds the set operating temperature value of the target device; wherein the camera is positioned in a top-down shooting direction.

[0007] The detection module is used to detect whether there is firelight in the captured photo;

[0008] The marking module is used to mark the photo with firelight as the first photo if there is firelight.

[0009] The pre-identification module is used to pre-identify the position of each device in each of the first captured photos and mark the first captured photos to obtain a second captured photo with the mark;

[0010] The transformation module is used to perform two-dimensional discrete wavelet transform on each of the second captured photos and remove the low-frequency subband to obtain the target image;

[0011] The selection module is used to select K1 preset points in the horizontal direction and K2 preset points in the vertical direction on the target image to perform horizontal and vertical integral projection partitioning on the target image to obtain (K1+1)×(K2+1) target regions; where K1 and K2 are preset values;

[0012] Extraction module, used to extract formulas Extract pixel features from pixels; where This indicates a preset indicator function. , This represents the absolute value of the difference between the grayscale value of the p-th surrounding pixel of pixel g and the grayscale value of the floor color, where 'a' is a preset value. Indicates indicator function and P represents the total number of pixels. , This represents the grayscale value of pixel g, and the surrounding pixels represent pixels within a preset distance from the preset pixel.

[0013] The representation module is used to represent each target region using the pixel features to obtain a feature matrix for each target region.

[0014] The calculation module is used to calculate the sum of the values ​​in the feature matrix of each target region to obtain the feature value of each target region;

[0015] The determination module is used to determine the hazard intensity value of each target area based on the magnitude of the feature value.

[0016] Furthermore, the determining module also includes:

[0017] The identification submodule is used to identify the location of the fire in the first captured photo and mark the target area with the fire to obtain the fire area;

[0018] The calculation submodule is used to calculate the reciprocal of the distance from each target area to the nearest fire area plus 1, and multiply it by the characteristic value to obtain the danger intensity value of each target area.

[0019] Furthermore, it also includes:

[0020] The judgment module is used to determine whether the danger intensity value of each target area is less than the preset value;

[0021] The setting module is used to set the firefighting plan according to the magnitude of the danger intensity value if the case is true.

[0022] The fire suppression module is used to extinguish fires in the area where the equipment is located, based on the fire suppression plan and the location of each target area.

[0023] Furthermore, it also includes:

[0024] The first statistical module is used to calculate the hazard intensity value of each target area at each time point;

[0025] The input module is used to input the danger intensity value and target image of each target area into a preset neural network model to obtain the predicted danger intensity value of each target area after a set time. The preset neural network model is trained in a supervised manner using multiple sets of training data. Each set of training data includes the danger intensity value of the target area at multiple consecutive time points, the corresponding target image, and the danger intensity value of the target area after a set time. The danger intensity value of the target area at multiple consecutive time points and the corresponding target image are the input of the preset neural network model, and the danger intensity value of the target area after a set time is the output of the preset network model.

[0026] The second statistics module is used to count the number of target areas in each target area whose predicted danger intensity value is greater than the preset danger intensity value;

[0027] The quantity judgment module is used to determine whether the quantity in the target area is greater than the preset quantity;

[0028] The distress call module is used to call the fire department if the number of calls exceeds a preset limit.

[0029] Furthermore, it also includes:

[0030] The sorting module is used to sort the target areas from largest to smallest according to their hazard intensity values.

[0031] The control module is used to control the preset fire-fighting devices to extinguish fires in each target area in sequence according to the arranged target areas.

[0032] This invention also provides an intelligent electrical fire monitoring method based on the Internet of Things, comprising:

[0033] The temperature of each device in the area where the device is located is obtained through a preset temperature sensor array; wherein, the temperature sensor array is connected to each device in the area where the device is located, and is used to obtain the temperature of each device;

[0034] When the temperature of the target device exceeds its set operating temperature, the camera captures frames from the target device at preset time intervals to obtain multiple photos; wherein the camera is positioned in a top-down shooting direction.

[0035] Detect whether the photograph contains flames;

[0036] If there is firelight, the photo with firelight is recorded as the first photo.

[0037] The positions of each device in each of the first captured photos are pre-identified, and the first captured photos are marked to obtain the marked second captured photos;

[0038] Perform two-dimensional discrete wavelet transform on each of the second-taken photos and remove the low-frequency subband to obtain the target image;

[0039] K1 preset points are selected horizontally and K2 preset points are selected vertically on the target image. The target image is then divided into (K1+1)×(K2+1) target regions by horizontal and vertical integral projection. K1 and K2 are preset values. According to the formula... Extract pixel features from pixels; where This indicates a preset indicator function. , This represents the absolute value of the difference between the grayscale value of the p-th surrounding pixel of pixel g and the grayscale value of the floor color, where 'a' is a preset value. Indicates indicator function and P represents the total number of pixels. , This represents the grayscale value of pixel g, and the surrounding pixels represent pixels within a preset distance from the preset pixel.

[0040] Each target region is represented by a matrix using the pixel features to obtain the feature matrix of each target region;

[0041] Calculate the sum of the values ​​in the feature matrix of each target region to obtain the feature value of each target region;

[0042] The hazard intensity value of each target area is determined based on the magnitude of the characteristic value.

[0043] Furthermore, the step of determining the danger intensity value of each target area based on the magnitude of the feature value also includes:

[0044] The location of the firelight is identified in the first photograph, and the target area with the firelight is marked to obtain the firelight area;

[0045] Calculate the reciprocal of the distance from each target area to the nearest fire area plus 1, and multiply it by the characteristic value to obtain the danger intensity value of each target area.

[0046] Furthermore, after the step of determining the danger intensity value of each target area based on the magnitude of the feature value, the method further includes:

[0047] Determine whether the hazard intensity value of each target area is less than the preset value;

[0048] If so, then set up a firefighting plan according to the magnitude of the danger intensity value;

[0049] Firefighting operations were carried out in the area where the equipment was located, based on the firefighting plan and the location of each target area.

[0050] Furthermore, after the step of determining the danger intensity value of each target area based on the magnitude of the feature value, the method further includes:

[0051] Statistical analysis of the hazard intensity values ​​for each target area at each time point;

[0052] The danger intensity values ​​and target images of each target region are input into a preset neural network model to obtain the predicted danger intensity values ​​of each target region after a set time. The preset neural network model is trained in a supervised manner using multiple sets of training data. Each set of training data includes the danger intensity values ​​of the target region at multiple consecutive time points, the corresponding target images, and the danger intensity values ​​of the target region after a set time. The danger intensity values ​​of the target region at multiple consecutive time points and the corresponding target images are the inputs to the preset neural network model, and the danger intensity values ​​of the target region after a set time are the outputs of the preset network model.

[0053] Count the number of target areas whose predicted hazard intensity value is greater than the preset hazard intensity value;

[0054] Determine if the number of target areas exceeds the preset number;

[0055] If the number exceeds the preset limit, call the fire department for help.

[0056] Furthermore, after the step of determining the danger intensity value of each target area based on the magnitude of the feature value, the method further includes:

[0057] The target areas are arranged from largest to smallest according to their hazard intensity values.

[0058] The pre-set fire-fighting devices are controlled according to the arranged target areas to extinguish fires in each target area in sequence.

[0059] The beneficial effects of this invention are as follows: By detecting temperature and identifying fire, the target image is determined. The target image is segmented using a segmentation method. Then, the danger intensity value of each target area is obtained by calculating the features of the pixels. This enables the judgment of danger intensity value based on the segmented area. Subsequently, fire extinguishing can be carried out in a graded manner based on the danger intensity value of each target area, thereby avoiding greater damage to the area where the equipment is located and greatly protecting the safety of the area where the equipment is located. Attached Figure Description

[0060] Figure 1This is a schematic block diagram of an intelligent electrical fire monitoring system based on the Internet of Things according to an embodiment of the present invention;

[0061] Figure 2 This is a flowchart illustrating an embodiment of an intelligent electrical fire monitoring method based on the Internet of Things according to the present invention.

[0062] Figure 3 This is a schematic block diagram of the structure of a computer device according to an embodiment of this application.

[0063] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Implementation

[0064] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0065] It should be noted that all directional indicators (such as up, down, left, right, front, back, etc.) in the embodiments of the present invention are only used to explain the relative positional relationship and movement of each component in a certain specific posture (as shown in the figure). If the specific posture changes, the directional indicator will also change accordingly. The connection can be a direct connection or an indirect connection.

[0066] In this article, the term "and / or" is merely a description of the relationship between related objects, indicating that there can be three kinds of relationships. For example, A and B can represent three situations: A exists alone, A and B exist simultaneously, and B exists alone.

[0067] Furthermore, in this invention, descriptions involving "first," "second," etc., are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined with "first" or "second" may explicitly or implicitly include at least one of that feature. Additionally, the technical solutions of the various embodiments can be combined with each other, but only on the basis of being achievable by those skilled in the art. When the combination of technical solutions is contradictory or impossible to implement, such a combination of technical solutions should be considered non-existent and not within the scope of protection claimed by this invention.

[0068] Reference Figure 1 The present invention also provides an intelligent electrical fire monitoring system based on the Internet of Things, comprising:

[0069] The acquisition module 10 is used to acquire the temperature of each device in the area where the device is located through a preset temperature sensor array; wherein, the temperature sensor array is connected to each device in the area where the device is located, and is used to acquire the temperature of each device;

[0070] The acquisition module 20 is used to acquire multiple photos by capturing frames of the target device at preset time intervals through a camera when the temperature of the target device is greater than the set operating temperature value of the target device; wherein the camera is in a top-down shooting direction.

[0071] Detection module 30 is used to detect whether there is firelight in the captured photo;

[0072] The marking module 40 is used to mark the photograph with firelight as the first photograph if there is firelight.

[0073] The pre-identification module 50 is used to pre-identify the position of each device in each of the first captured photos and mark the first captured photos to obtain a second captured photo with the mark.

[0074] The transformation module 60 is used to perform two-dimensional discrete wavelet transform on each of the second captured photos and remove the low-frequency subband to obtain the target image;

[0075] The selection module 70 is used to select K1 preset points in the horizontal direction and K2 preset points in the vertical direction on the target image to perform horizontal and vertical integral projection partitioning on the target image to obtain (K1+1)×(K2+1) target regions; where K1 and K2 are preset values;

[0076] Extraction module 80 is used to extract formulas. Extract pixel features from pixels; where This indicates a preset indicator function. , This represents the absolute value of the difference between the grayscale value of the p-th surrounding pixel of pixel g and the grayscale value of the floor color, where 'a' is a preset value. Indicates indicator function and P represents the total number of pixels. , This represents the grayscale value of pixel g, and the surrounding pixels represent pixels within a preset distance from the preset pixel.

[0077] The representation module 90 is used to represent each target region using the pixel features to obtain a feature matrix for each target region.

[0078] The calculation module 100 is used to calculate the sum of the values ​​in the feature matrix of each target region to obtain the feature value of each target region;

[0079] The determination module 110 is used to determine the danger intensity value of each target area based on the magnitude of the feature value.

[0080] As described in the acquisition module 10, the temperature of each device in the area where the device is located is acquired through a preset temperature sensor array; wherein, the temperature sensor array is connected to each device in the area where the device is located, and is used to acquire the temperature of each device; wherein, the temperature sensor array is connected to each device, and is used to monitor the temperature of each device. When its temperature is higher than the operating temperature of the device, it is determined that it has a burning document. Therefore, it can be monitored by a camera, so that the corresponding photo can be acquired in the first time.

[0081] As described in the acquisition module 20, when the temperature of the target device is greater than the set operating temperature value of the target device, the camera acquires the shooting frames of the target device at preset time intervals to obtain multiple photos. The camera is positioned in a top-down direction, meaning it can be set on the ceiling. The preset time interval is not limited; for example, a 3-second interval can be set. It should be noted that a shot is taken at each time interval, thus obtaining multiple photos.

[0082] As described in the detection module 30, it detects whether there is firelight in the captured photo. The firelight can be identified by existing image recognition methods. Of course, since the temperature has been determined in advance by the temperature sensor array, it can be combined here. That is, if the temperature is greater than the set working temperature and the pixel values ​​of some pixels have changed, this change tends to be the pixel values ​​of firelight. Therefore, it can be determined whether there is firelight. If the pixel values ​​do not change, it can be said that there is no firelight in the photo.

[0083] As described in module 40, if there is firelight, the photograph showing firelight is recorded as the first photograph. That is, the photograph showing firelight is extracted and recorded as the first photograph. Since the presence of obvious firelight indicates a serious fire situation, this portion of the firelight is extracted.

[0084] As described in the pre-identification module 50, the positions of each device in each of the first captured photos are pre-identified, and the first captured photos are marked to obtain a second captured photo with the markings. To facilitate subsequent judgment, the positions of each device can be pre-identified here. The pre-identification method can be to obtain a distribution map of each device, and then mark the first captured photo according to the distribution map, thereby obtaining a second captured photo with the markings.

[0085] As described in the transformation module 60, each of the second captured photos undergoes a two-dimensional discrete wavelet transform, and the low-frequency sub-band is removed to obtain the target image. The two-dimensional discrete wavelet transform is a feature extraction tool. After the two-dimensional discrete wavelet transform, the photo becomes four parts. Because wavelet transform has excellent localization properties in the frequency domain, after removing the low-frequency sub-band, each sub-band can recover the original target image. That is, the target image reflects the firelight outline information in the original photo, facilitating the extraction of firelight features. Furthermore, since dense smoke is generated during a fire, it can cause interference. The two-dimensional discrete wavelet transform can reduce this interference, making the firelight outline more detailed.

[0086] As described in module 70 and module 80, K1 preset points are selected in the horizontal direction and K2 preset points are selected in the vertical direction on the target image. The target image is then divided into (K1+1)×(K2+1) target regions by horizontal and vertical integral projection. Here, K1 and K2 are preset values. According to the formula... Extract pixel features from pixels; where This indicates a preset indicator function. , This represents the absolute value of the difference between the grayscale value of the p-th surrounding pixel of pixel g and the grayscale value of the floor color, where 'a' is a preset value. Indicates indicator function and P represents the total number of pixels. , This represents the grayscale value of pixel g, and the surrounding pixels represent pixels within a preset distance from the preset pixel.

[0087] Here, K1 and K2 are preset values. Their purpose is to divide the target image into multiple regions for feature extraction, enabling subsequent determination of danger intensity based on these features. The surrounding pixels are defined relative to the central pixels. To avoid errors caused by the floor color, the grayscale value of the floor color should be subtracted. In defining this, it should be noted that... As an indicator function, when the pixel value of a point differs significantly from that of its surrounding pixels, the corresponding K value will increase, thus increasing the value of subsequent feature values. Conversely, if the difference is small, the K value will be small, resulting in small feature values. This is an indicator function that further processes the data based on the sum of the values ​​of each pixel; the sum is... ,therefore, It is a function that is positively correlated with the sum, such as a linear function, a quadratic function, an exponential function, etc. It only needs to be ensured that as the sum increases, its corresponding value also increases.

[0088] As described in the representation module 90, calculation module 100, and determination module 110, each target region is represented by a matrix using the pixel features to obtain a feature matrix for each target region; the sum of the values ​​in the feature matrix of each target region is calculated to obtain the feature value of each target region; and the danger intensity value of each target region is determined based on the magnitude of the feature value. By representing each pixel using a matrix and then calculating the sum of the values ​​to obtain the feature value of each target region, it should be noted that the larger the feature value, the higher the corresponding danger intensity value. That is, the region with the largest feature value is generally in a state of burning, with one side not burning and the other side burning. This results in the largest difference between pixel values, and therefore the greatest danger intensity. This achieves the judgment of danger intensity value based on different regions, allowing for tiered fire suppression based on the danger intensity value of each target region, thereby preventing greater damage to the area where the equipment is located and greatly protecting the safety of the area.

[0089] The Internet of Things (IoT) comprises multiple IoT devices (such as temperature sensor arrays and cameras), an IoT server, and a data processing center. When an IoT device (taking a temperature sensor array as an example) receives a data collection request from the IoT server, it responds by acquiring the temperature of each device in its area and then sending the data to the requesting IoT server. At this point, the IoT device can send the data to the data processing center via a message queue. This data processing center, such as a database, can be used to process data such as detecting fire in captured images and processing target images. In some embodiments, the IoT device may also include a fire extinguishing device, which is controlled by the calculation results processed by the data processing center to extinguish the fire.

[0090] In one embodiment, the determining module 110 further includes:

[0091] The identification submodule is used to identify the location of the fire in the first captured photo and mark the target area with the fire to obtain the fire area;

[0092] The calculation submodule is used to calculate the reciprocal of the distance from each target area to the nearest fire area plus 1, and multiply it by the characteristic value to obtain the danger intensity value of each target area.

[0093] As described in the module above, the location of the fire is identified in the first captured image, and the target areas with fire are marked to obtain the fire area. The reciprocal of the distance from each target area to the nearest fire area plus one is calculated and multiplied by a feature value to obtain the danger intensity value of each target area. Specifically, marking the location with fire is problematic because, in reality, there may be other objects on the equipment, or the equipment itself may have significant color differences, resulting in a larger feature value and thus a larger final danger intensity value. Therefore, using the reciprocal of the distance from the target area to the nearest fire area plus one as a weighting value can reduce this error and improve the accuracy of the judgment.

[0094] In one embodiment, it also includes:

[0095] The judgment module is used to determine whether the danger intensity value of each target area is less than the preset value;

[0096] The setting module is used to set the firefighting plan according to the magnitude of the danger intensity value if the case is true.

[0097] The fire suppression module is used to extinguish fires in the area where the equipment is located, based on the fire suppression plan and the location of each target area.

[0098] As described in the module above, it is determined whether the danger intensity value of each target area is less than a preset value; if so, a firefighting plan is set according to the danger intensity value; and the area where the equipment is located is extinguished according to the firefighting plan and the location of each target area. After obtaining the danger intensity value of each target area, targeted firefighting can be carried out in the target area, thereby ensuring the safety of other equipment in the area where the equipment is located. In a preferred embodiment, if the danger intensity value of each target area is greater than the set value, it can be determined that firefighting is not possible, and firefighting can be carried out according to the danger intensity value of other locations, thereby saving the dosage of firefighting equipment and achieving a better firefighting effect.

[0099] In one embodiment, it also includes:

[0100] The first statistical module is used to calculate the hazard intensity value of each target area at each time point;

[0101] The input module is used to input the danger intensity value and target image of each target area into a preset neural network model to obtain the predicted danger intensity value of each target area after a set time. The preset neural network model is trained in a supervised manner using multiple sets of training data. Each set of training data includes the danger intensity value of the target area at multiple consecutive time points, the corresponding target image, and the danger intensity value of the target area after a set time. The danger intensity value of the target area at multiple consecutive time points and the corresponding target image are the input of the preset neural network model, and the danger intensity value of the target area after a set time is the output of the preset network model.

[0102] The second statistics module is used to count the number of target areas in each target area whose predicted danger intensity value is greater than the preset danger intensity value;

[0103] The quantity judgment module is used to determine whether the quantity in the target area is greater than the preset quantity;

[0104] The distress call module is used to call the fire department if the number of calls exceeds a preset limit.

[0105] As described in the module above, the prediction of the fire situation after a set time is realized. That is, the danger intensity value of each target area at each time point is statistically analyzed. Based on the changes in the danger intensity value of each target area and the location of each device, a prediction can be made. The prediction method is to input the data into a preset network model for prediction. Specifically, the preset network model can be one of the following: convolutional neural network model, recurrent neural network model, and deep neural network model. Specifically, the training process can be represented as: acquiring multiple sets of training data in advance, where the danger intensity value of the target area after the set time is the danger intensity value acquired in real time after the set time. That is, each target area is marked by the danger intensity value. The danger intensity values ​​of the target areas at multiple consecutive time points and the corresponding target images are input into the input layer. The hidden layer predicts the danger intensity value at each location. The output layer outputs the final result. Then, the final result is compared with the preset result (i.e., the danger intensity value of the target area after the set time). The parameters in the hidden layer are continuously adjusted through the loss function to achieve the training purpose.

[0106] In one embodiment, it also includes:

[0107] The sorting module is used to sort the target areas from largest to smallest according to their hazard intensity values.

[0108] The control module is used to control the preset fire-fighting devices to extinguish fires in each target area in sequence according to the arranged target areas.

[0109] As described in the module above, automatic fire suppression is achieved for target areas. These target areas are arranged from highest to lowest hazard intensity, and pre-set fire-fighting devices are controlled sequentially to extinguish the fires in each area. Since the location of the equipment may have existing fire-fighting equipment, fire suppression can be carried out based on the hazard intensity value of each target area, thereby achieving the purpose of fire suppression.

[0110] Reference Figure 2 This invention proposes an intelligent electrical fire monitoring method based on the Internet of Things, comprising:

[0111] S1: The temperature of each device in the area where the device is located is obtained through a preset temperature sensor array; wherein, the temperature sensor array is connected to each device in the area where the device is located, and is used to obtain the temperature of each device;

[0112] S2: When the temperature of the target device is higher than the set operating temperature of the target device, the camera captures frames of the target device at preset time intervals to obtain multiple photos; wherein the camera is shooting from a top-down direction.

[0113] S3: Detects whether there is firelight in the captured photo;

[0114] S4: If there is firelight, then the photo with firelight is recorded as the first photo.

[0115] S5: Pre-identify the position of each device in each of the first captured photos, and mark the first captured photos to obtain the marked second captured photos;

[0116] S6: Perform two-dimensional discrete wavelet transform on each of the second captured photos and remove the low-frequency subband to obtain the target image;

[0117] S7: Take K1 preset points in the horizontal direction and K2 preset points in the vertical direction on the target image, and perform horizontal and vertical integral projection partitioning on the target image to obtain (K1+1)×(K2+1) target regions; where K1 and K2 are preset values;

[0118] S8: According to the formula Extract pixel features from pixels; where This indicates a preset indicator function. , This represents the absolute value of the difference between the grayscale value of the p-th surrounding pixel of pixel g and the grayscale value of the floor color, where 'a' is a preset value. Indicates indicator function and P represents the total number of pixels. , S9: Represent the grayscale value of pixel g, and the surrounding pixels represent pixels within a preset distance of a preset pixel; S9: Represent each target region using the pixel features to obtain the feature matrix of each target region;

[0119] S10: Calculate the sum of the values ​​in the feature matrix of each target region to obtain the feature value of each target region;

[0120] S11: Determine the hazard intensity value of each target area based on the magnitude of the characteristic value.

[0121] In one embodiment, step S11, which determines the danger intensity value of each target area based on the magnitude of the feature value, further includes:

[0122] S1101: Identify the location of the fire in the first photograph, and mark the target area with the fire to obtain the fire area;

[0123] S1102: Calculate the reciprocal of the distance from each target area to the nearest fire area plus 1, and multiply it by the characteristic value to obtain the danger intensity value of each target area.

[0124] In one embodiment, after step S11 of determining the danger intensity value of each target area based on the magnitude of the feature value, the method further includes:

[0125] S1201: Determine whether the danger intensity value of each target area is less than the preset value;

[0126] S1202: If so, set up the firefighting plan according to the magnitude of the danger intensity value;

[0127] S1203: Extinguish the fire in the area where the equipment is located according to the firefighting plan and the location of each target area.

[0128] In one embodiment, after step S12 of determining the danger intensity value of each target area based on the magnitude of the feature value, the method further includes:

[0129] S1211: Statistical analysis of the hazard intensity values ​​for each target area at each time point;

[0130] S1212: Input the danger intensity value and target image of each target region into a preset neural network model to obtain the predicted danger intensity value of each target region after a set time; wherein, the preset neural network model is trained in a supervised manner using multiple sets of training data. Each set of training data includes the danger intensity value of the target region at multiple consecutive time points, the corresponding target image, and the danger intensity value of the target region after a set time. The danger intensity value of the target region at multiple consecutive time points and the corresponding target image are the inputs of the preset neural network model, and the danger intensity value of the target region after a set time is the output of the preset network model.

[0131] S1213: Count the number of target areas whose predicted hazard intensity value is greater than the preset hazard intensity value;

[0132] S1214: Determine whether the number of target areas is greater than the preset number;

[0133] S1215: If the number exceeds the preset quantity, call the fire department for help.

[0134] In one embodiment, after step S12 of determining the danger intensity value of each target area based on the magnitude of the feature value, the method further includes:

[0135] S1221: Arrange the target areas from largest to smallest according to their hazard intensity values;

[0136] S1222: Control the preset fire-fighting devices according to the arranged target areas to extinguish the fire in each target area in sequence.

[0137] The beneficial effects of this invention are as follows: By detecting temperature and identifying fire, the target image is determined. The target image is segmented using a segmentation method. Then, the danger intensity value of each target area is obtained by calculating the features of the pixels. This enables the judgment of danger intensity value based on the segmented area. Subsequently, fire extinguishing can be carried out in a graded manner based on the danger intensity value of each target area, thereby avoiding greater damage to the area where the equipment is located and greatly protecting the safety of the area where the equipment is located.

[0138] Reference Figure 3 This application also provides a computer device, which may be a server, and its internal structure may be as follows: Figure 3As shown, the computer device includes a processor, memory, network interface, and database connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and database. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The database stores various photographs, etc. The network interface is used for communication with external terminals via a network connection. When the computer program is executed by the processor, it can implement the IoT-based intelligent electrical fire monitoring method described in any of the above embodiments.

[0139] Those skilled in the art will understand that Figure 3 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer equipment on which the present application is applied.

[0140] This application also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, can implement the IoT-based intelligent electrical fire monitoring method described in any of the above embodiments.

[0141] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media provided in this application and used in the embodiments can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0142] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, apparatus, article, or method that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, apparatus, article, or method. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, apparatus, article, or method that includes that element.

[0143] The embodiments of this application can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence (AI) refers to the theories, methods, technologies, and application systems that use digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0144] Foundational technologies in artificial intelligence generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing, operating / interactive systems, and mechatronics. AI software technologies mainly encompass computer vision, robotics, biometrics, speech processing, natural language processing, and machine learning / deep learning.

[0145] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of the claims of the present invention.

Claims

1. An intelligent electrical fire monitoring system based on the Internet of Things, characterized in that, include: An acquisition module is used to acquire the temperature of each device in the area where the device is located through a preset temperature sensor array; wherein, the temperature sensor array is connected to each device in the area where the device is located, and is used to acquire the temperature of each device; The acquisition module is used to acquire multiple photos by capturing frames of the target device at preset time intervals through a camera when the temperature of the target device exceeds the set operating temperature value of the target device; wherein the camera is positioned in a top-down shooting direction. The detection module is used to detect whether there is firelight in the captured photo; The marking module is used to mark the photo with firelight as the first photo if there is firelight. The pre-identification module is used to pre-identify the position of each device in each of the first captured photos and mark the first captured photos to obtain a second captured photo with the mark; The transformation module is used to perform two-dimensional discrete wavelet transform on each of the second captured photos and remove the low-frequency subband to obtain the target image; The selection module is used to select K1 preset points in the horizontal direction and K2 preset points in the vertical direction on the target image to perform horizontal and vertical integral projection partitioning on the target image to obtain (K1+1)×(K2+1) target regions; where K1 and K2 are preset values; Extraction module, used to extract formulas Extract pixel features from pixels; where This indicates a preset indicator function. , This represents the absolute value of the difference between the grayscale value of the p-th surrounding pixel of pixel g and the grayscale value of the floor color, where 'a' is a preset value. Indicates indicator function and P represents the total number of pixels. , This represents the grayscale value of pixel g, and the surrounding pixels represent pixels within a preset distance from the preset pixel. The representation module is used to represent each target region using the pixel features to obtain a feature matrix for each target region. The calculation module is used to calculate the sum of the values ​​in the feature matrix of each target region to obtain the feature value of each target region; The determination module is used to determine the hazard intensity value of each target area based on the magnitude of the feature value.

2. The IoT-based intelligent electrical fire monitoring system as described in claim 1, characterized in that, The determining module further includes: The identification submodule is used to identify the location of the fire in the first captured photo and mark the target area with the fire to obtain the fire area; The calculation submodule is used to calculate the reciprocal of the distance from each target area to the nearest fire area plus 1, and multiply it by the characteristic value to obtain the danger intensity value of each target area.

3. The IoT-based intelligent electrical fire monitoring system as described in claim 1, characterized in that, Also includes: The judgment module is used to determine whether the danger intensity value of each target area is less than the preset value; The setting module is used to set the firefighting plan according to the magnitude of the danger intensity value if the case is true. The fire suppression module is used to extinguish fires in the area where the equipment is located, based on the fire suppression plan and the location of each target area.

4. The IoT-based intelligent electrical fire monitoring system as described in claim 1, characterized in that, Also includes: The first statistical module is used to calculate the hazard intensity value of each target area at each time point; The input module is used to input the danger intensity value and target image of each target area into a preset neural network model to obtain the predicted danger intensity value of each target area after a set time. The preset neural network model is trained in a supervised manner using multiple sets of training data. Each set of training data includes the danger intensity value of the target area at multiple consecutive time points, the corresponding target image, and the danger intensity value of the target area after a set time. The danger intensity value of the target area at multiple consecutive time points and the corresponding target image are the input of the preset neural network model, and the danger intensity value of the target area after a set time is the output of the preset network model. The second statistics module is used to count the number of target areas in each target area whose predicted danger intensity value is greater than the preset danger intensity value; The quantity judgment module is used to determine whether the quantity in the target area is greater than the preset quantity; The distress call module is used to call the fire department if the number of calls exceeds a preset limit.

5. The IoT-based intelligent electrical fire monitoring system as described in claim 1, characterized in that, Also includes: The sorting module is used to sort the target areas from largest to smallest according to their hazard intensity values. The control module is used to control the preset fire-fighting devices to extinguish fires in each target area in sequence according to the arranged target areas.

6. A smart electrical fire monitoring method based on the Internet of Things, characterized in that, include: The temperature of each device in the area where the device is located is obtained through a preset temperature sensor array; wherein, the temperature sensor array is connected to each device in the area where the device is located, and is used to obtain the temperature of each device; When the temperature of the target device exceeds its set operating temperature, the camera captures frames from the target device at preset time intervals to obtain multiple photos; wherein the camera is positioned in a top-down shooting direction. Detect whether the photograph contains flames; If there is firelight, the photo with firelight is recorded as the first photo. The positions of each device in each of the first captured photos are pre-identified, and the first captured photos are marked to obtain the marked second captured photos; Perform two-dimensional discrete wavelet transform on each of the second-taken photos and remove the low-frequency subband to obtain the target image; K1 preset points are selected horizontally and K2 preset points are selected vertically on the target image. The target image is then divided into (K1+1)×(K2+1) target regions by horizontal and vertical integral projection. Here, K1 and K2 are preset values. This is used to calculate the target region according to the formula... Extract pixel features from pixels; where This indicates a preset indicator function. , This represents the absolute value of the difference between the grayscale value of the p-th surrounding pixel of pixel g and the grayscale value of the floor color, where 'a' is a preset value. Indicates indicator function and P represents the total number of pixels. , The grayscale value of pixel g is represented by the grayscale value of the surrounding pixels, and the surrounding pixels are represented by pixels within a preset length of the preset pixels. Each target region is represented by the pixel features in a matrix to obtain the feature matrix of each target region. Calculate the sum of the values ​​in the feature matrix of each target region to obtain the feature value of each target region; The hazard intensity value of each target area is determined based on the magnitude of the characteristic value.

7. The IoT-based intelligent electrical fire monitoring method as described in claim 6, characterized in that, The step of determining the danger intensity value of each target area based on the magnitude of the feature value further includes: The location of the firelight is identified in the first photograph, and the target area with the firelight is marked to obtain the firelight area; Calculate the reciprocal of the distance from each target area to the nearest fire area plus 1, and multiply it by the characteristic value to obtain the danger intensity value of each target area.

8. The IoT-based intelligent electrical fire monitoring method as described in claim 6, characterized in that, Following the step of determining the hazard intensity value of each target area based on the magnitude of the characteristic value, the method further includes: Determine whether the hazard intensity value of each target area is less than the preset value; If so, then set up a firefighting plan according to the magnitude of the danger intensity value; Firefighting operations were carried out in the area where the equipment was located, based on the firefighting plan and the location of each target area.

9. The IoT-based intelligent electrical fire monitoring method as described in claim 6, characterized in that, Following the step of determining the hazard intensity value of each target area based on the magnitude of the characteristic value, the method further includes: Statistical analysis of the hazard intensity values ​​for each target area at each time point; The danger intensity values ​​and target images of each target region are input into a preset neural network model to obtain the predicted danger intensity values ​​of each target region after a set time. The preset neural network model is trained in a supervised manner using multiple sets of training data. Each set of training data includes the danger intensity values ​​of the target region at multiple consecutive time points, the corresponding target images, and the danger intensity values ​​of the target region after a set time. The danger intensity values ​​of the target region at multiple consecutive time points and the corresponding target images are the inputs to the preset neural network model, and the danger intensity values ​​of the target region after a set time are the outputs of the preset network model. Count the number of target areas whose predicted hazard intensity value is greater than the preset hazard intensity value; Determine if the number of target areas exceeds the preset number; If the number exceeds the preset limit, call the fire department for help.

10. The IoT-based intelligent electrical fire monitoring method as described in claim 6, characterized in that, Following the step of determining the hazard intensity value of each target area based on the magnitude of the characteristic value, the method further includes: The target areas are arranged from largest to smallest according to their hazard intensity values. The pre-set fire-fighting devices are controlled according to the arranged target areas to extinguish fires in each target area in sequence.

Images