Device, method and equipment for detecting forest fire around cable based on dual-channel detection model

Abstract

The invention discloses a cable surrounding forest fire detection device, method and equipment based on a dual-channel detection model, and belongs to the technical field of electric power facilities. The device comprises an image acquisition module for acquiring a visible light image and an infrared image of a cable; the detection result acquisition module inputs the visible light image into a first channel to acquire a first detection result; inputting the infrared image into a second channel to obtain a second detection result; the fire source verification module is used for determining the fire source abnormal range of the visible light image and verifying the fire source abnormal range of the visible light image by using the second detection result; or, determining an infrared image fire source abnormal range, and verifying the infrared image fire source abnormal range by using the first detection result; and the verification result output module is used for determining a fire source inference position and generating forest fire alarm information. According to the technical scheme, the visible light image and the infrared image around the cable can be subjected to dual recognition, a fire source can be found in time, the recognition precision is improved, the recognition error probability is reduced, and unnecessary economic losses are avoided.

Description

Technical Field

[0001] The present application belongs to the technical field of electric power facilities, and specifically relates to a cable perimeter wildfire detection device, method and equipment based on a dual-channel detection model. Background Art

[0002] When laying cables for long-distance power transmission, they often pass through mountainous areas or grasslands and are laid using high-altitude towers.

[0003] At present, due to the large number of combustible materials in mountainous areas or grasslands, once the cable is interrupted or fails, the current increases sharply, and the large amount of heat generated is likely to cause a wildfire. Moreover, in sparsely populated areas, if wildfires caused by cable faults are not discovered and handled in time, they will cause the wildfires to spread over a large area. If wildfires are misidentified and fires are extinguished in locations where no wildfires have occurred, serious economic losses will also be caused.

[0004] Therefore, how to accurately identify wildfires caused by cable faults is a technical problem that needs to be solved urgently by those skilled in the art. Summary of the invention

[0005] The embodiments of the present application provide a device, method and equipment for detecting wildfires around cables based on a dual-channel detection model, the purpose of which is to improve the accuracy of fire source identification while ensuring timely discovery of fire sources, reduce the probability of identification errors, and avoid unnecessary economic losses by performing dual identification of visible light images and infrared images around the cables.

[0006] In a first aspect, an embodiment of the present application provides a cable surrounding mountain fire detection device based on a dual-channel detection model, the device comprising:

[0007] An image acquisition module, used to acquire visible light images and infrared images within a preset range of the cable;

[0008] A detection result acquisition module, used for inputting the visible light image into the first channel of the dual-channel detection model to obtain a first detection result; and inputting the infrared image into the second channel of the dual-channel detection model to obtain a second detection result;

[0009] A fire source verification module, for determining an abnormal range of the fire source in the visible light image when the first detection result indicates that a fire source exists, and verifying the abnormal range of the fire source in the visible light image using the second detection result; or, for determining an abnormal range of the fire source in the infrared image when the second detection result indicates that a fire source exists, and verifying the abnormal range of the fire source in the infrared image using the first detection result;

[0010] The verification result output module is used to determine the abnormal range of the successful verification as the inferred location of the fire source when the verification is successful, and generate wildfire alarm information with the inferred location of the fire source.

[0011] Furthermore, the detection result acquisition module is specifically used to:

[0012] Extracting image features from the visible light image, matching the image features with pre-acquired fire source features, and if the match is successful, the first detection result is that there is a fire source;

[0013] as well as,

[0014] Identify the temperature distribution data of each area in the infrared image, and compare the temperature distribution data of each area with a preset temperature threshold. If the temperature value of at least one area exceeds the preset temperature threshold, determine that the second detection result is the presence of a fire source.

[0015] Furthermore, the fire source verification module is specifically used for:

[0016] When the first detection result indicates that there is a fire source, fire source feature detection is performed on the visible light image within a preset range of the fire source location, and the area with fire source features within the preset range is used as the fire source abnormal range of the visible light image;

[0017] or,

[0018] When the second detection result indicates that there is a fire source, temperature data within a preset range of the fire source location is obtained, and an area where the temperature data is within a preset temperature threshold is used as an abnormal fire source range of the infrared image.

[0019] Furthermore, the fire source verification module is specifically used for:

[0020] When the first detection result indicates that there is a fire source, determining the abnormal range of the fire source in the visible light image, obtaining the fire source confidence in the area corresponding to the abnormal range of the fire source in the visible light image in the second detection result, and if the fire source confidence is greater than the confidence threshold, the verification is successful;

[0021] or,

[0022] When the second detection result indicates that there is a fire source, determine the abnormal range of the fire source in the infrared image, and obtain the fire source confidence in the area corresponding to the abnormal range of the fire source in the infrared image in the first detection result. If the fire source confidence is greater than the confidence threshold, the verification is successful.

[0023] Furthermore, the detection result acquisition module is also used to:

[0024] Inputting the visible light image into a first channel of a dual-channel detection model to extract a first feature;

[0025] as well as,

[0026] Inputting the infrared image into the second channel of the dual-channel detection model to extract the second feature;

[0027] Fusing the first feature with the second feature to obtain a result feature;

[0028] Matching the result features with the fire source features to determine the candidate area;

[0029] Accordingly, the fire source verification module is also used for:

[0030] Identify whether there are multiple candidate areas overlapping and pointing to the same fire source in the candidate area. If so, filter out the redundant areas and retain the remaining candidate areas as the abnormal range of successful verification.

[0031] Furthermore, the detection result acquisition module is specifically used to:

[0032] Obtaining a preset first weight of a first channel and a second weight of a second channel;

[0033] The first feature and the second feature are fused according to a ratio of the first weight to the second weight to obtain a result feature.

[0034] In a second aspect, an embodiment of the present application provides a method for detecting mountain fires around cables based on a dual-channel detection model, the method comprising:

[0035] Obtain visible light images and infrared images within the preset range of the cable;

[0036] Inputting the visible light image into a first channel in a dual-channel detection model to obtain a first detection result; and inputting the infrared image into a second channel in the dual-channel detection model to obtain a second detection result;

[0037] When the first detection result indicates that a fire source exists, determining the abnormal range of the fire source in the visible light image, and verifying the abnormal range of the fire source in the visible light image using the second detection result; or, when the second detection result indicates that a fire source exists, determining the abnormal range of the fire source in the infrared image, and verifying the abnormal range of the fire source in the infrared image using the first detection result;

[0038] When the verification is successful, the abnormal range of the successful verification is determined as the inferred location of the fire source, and wildfire alarm information with the inferred location of the fire source is generated.

[0039] Further, inputting the visible light image into a first channel in a dual-channel detection model to obtain a first detection result; and inputting the infrared image into a second channel in the dual-channel detection model to obtain a second detection result, including:

[0040] Extracting image features from the visible light image, matching the image features with pre-acquired fire source features, and if the match is successful, the first detection result is that there is a fire source;

[0041] as well as,

[0042] Identify the temperature distribution data of each area in the infrared image, and compare the temperature distribution data of each area with a preset temperature threshold. If the temperature value of at least one area exceeds the preset temperature threshold, determine that the second detection result is the presence of a fire source.

[0043] In a third aspect, an embodiment of the present application provides an electronic device, which includes a processor, a memory, and a program or instruction stored in the memory and executable on the processor, wherein the program or instruction, when executed by the processor, implements the steps of the method described in the first aspect.

[0044] In a fourth aspect, an embodiment of the present application provides a readable storage medium, on which a program or instruction is stored, and when the program or instruction is executed by a processor, the steps of the method described in the first aspect are implemented.

[0045] In a fifth aspect, an embodiment of the present application provides a chip, comprising a processor and a communication interface, wherein the communication interface is coupled to the processor, and the processor is used to run a program or instruction to implement the method described in the first aspect.

[0046] In an embodiment of the present application, a gas concentration monitoring module is arranged at a preset position in the cable trench, and is used to detect the concentration of at least one gas component at the preset position to obtain a concentration detection result of the gas component; a data transmission activation module is used to receive an activation instruction from an acquisition terminal to determine a target concentration detection result; a detection result sending module is used to encode the target concentration detection result and send it to the acquisition terminal at a preset frequency; the acquisition terminal is used to receive the target concentration detection result; a harmful gas analysis module is used to analyze the harmful gas in the cable trench according to the gas concentration detection result to obtain a harmful gas analysis result.

[0047] Through the above-mentioned cable surrounding wildfire detection device based on the dual-channel detection model, the gas situation in the cable trench can be monitored in real time, and the detection results can be transmitted through a contactless communication method, which is conducive to timely identification of harmful gas anomalies and ensuring the safety of workers entering the cable trench. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 It is a structural schematic diagram of a cable surrounding mountain fire detection device based on a dual-channel detection model provided in Example 1 of the present application;

[0049] Figure 2 It is a structural schematic diagram of a cable surrounding mountain fire detection device based on a dual-channel detection model provided in Example 2 of the present application;

[0050] Figure 3 It is a structural schematic diagram of a cable surrounding mountain fire detection device based on a dual-channel detection model provided in Example 3 of the present application;

[0051] Figure 4 It is a structural schematic diagram of a cable surrounding mountain fire detection device based on a dual-channel detection model provided in Example 4 of the present application;

[0052] Figure 5 It is a structural schematic diagram of a cable surrounding mountain fire detection device based on a dual-channel detection model provided in Example 5 of the present application;

[0053] Figure 6 It is a structural schematic diagram of a cable surrounding mountain fire detection device based on a dual-channel detection model provided in Example 6 of the present application;

[0054] Figure 7 It is a structural schematic diagram of a cable surrounding mountain fire detection device based on a dual-channel detection model provided in Example 7 of the present application;

[0055] Figure 8 It is a flow chart of a cable surrounding wildfire detection method based on a dual-channel detection model provided in Example 8 of the present application;

[0056] Fig. 9 It is a structural diagram of an electronic device provided in Example 9 of the present application. DETAILED DESCRIPTION

[0057] In order to make the purpose, technical scheme and advantages of the present application clearer, the specific embodiments of the present application are further described in detail below in conjunction with the accompanying drawings. It is understood that the specific embodiments described herein are only used to explain the present application, rather than to limit the present application. It should also be noted that, for the convenience of description, only the part related to the present application but not all the contents are shown in the accompanying drawings. Before discussing the exemplary embodiments in more detail, it should be mentioned that some exemplary embodiments are described as processes or methods depicted as flow charts. Although the flow chart describes each operation (or step) as a sequential process, many of the operations therein can be implemented in parallel, concurrently or simultaneously. In addition, the order of each operation can be rearranged. The process can be terminated when its operation is completed, but it can also have additional steps not included in the accompanying drawings. The process can correspond to a method, a function, a procedure, a subroutine, a subprogram, etc.

[0058] The following will be combined with the drawings in the embodiments of the present application to clearly describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments in the present application belong to the scope of protection of this application.

[0059] The terms "first", "second", etc. in the specification and claims of the present application are used to distinguish similar objects, and are not used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable under appropriate circumstances, so that the embodiments of the present application can be implemented in an order other than those illustrated or described here, and the objects distinguished by "first", "second", etc. are generally of one type, and the number of objects is not limited. For example, the first object can be one or more. In addition, "and / or" in the specification and claims represents at least one of the connected objects, and the character " / " generally indicates that the objects associated with each other are in an "or" relationship.

[0060] In conjunction with the accompanying drawings, the cable surrounding wildfire detection device, method and equipment based on the dual-channel detection model provided in the embodiments of the present application are described in detail through specific embodiments and their application scenarios.

[0061] Embodiment 1

[0062] Figure 1 Schematic diagram of the structure of the cable surrounding mountain fire detection device based on the dual-channel detection model provided in the first embodiment of the present application. Figure 1 As shown, the specific steps include:

[0063] An image acquisition module 101 is used to acquire a visible light image and an infrared image within a preset range of the cable;

[0064] The detection result acquisition module 102 is used to input the visible light image into the first channel of the dual-channel detection model to obtain a first detection result; and input the infrared image into the second channel of the dual-channel detection model to obtain a second detection result.

[0065] The fire source verification module 103 is used to determine the abnormal range of the fire source in the visible light image when the first detection result is that there is a fire source, and use the second detection result to verify the abnormal range of the fire source in the visible light image; or, when the second detection result is that there is a fire source, determine the abnormal range of the fire source in the infrared image, and use the first detection result to verify the abnormal range of the fire source in the infrared image.

[0066] The verification result output module 104 is used to determine the abnormal range of the successful verification as the location of the inferred fire source when the verification is successful, and generate wildfire warning information with the inferred fire source location.

[0067] The application scenario of this solution can be the monitoring of wildfires around the cable installation location.

[0068] Based on the above scenario, it can be understood that the executor of this scenario can be a detection terminal that identifies the occurrence of fire sources, such as desktop computers, laptops, tablet computers, and interactive multimedia devices, etc., and no excessive limitations are made here.

[0069] The image acquisition module 101 may be a camera and an infrared camera, which may be arranged in parallel on a high-altitude iron tower to collect visible light images and infrared images of the cables on the wire rack. Specifically, the preset range may be the surrounding range that may cause a wildfire when a cable fails. For example, when a permanent grounding discharge fault occurs in a cable, the aluminum metal of the wire body melts, the insulating material catches fire, and the weeds and shrubs on the ground at the base of the pole can be ignited during the scattering process.

[0070] Visible light images can be RGB images composed of the three color spectra of red light, green light, and blue light reflected or emitted by objects. The colors of such images are very close to the colors observable by the human eye and can provide rich visual information.

[0071] Infrared images can be images obtained through infrared radiation (light with a wavelength longer than visible light). Infrared radiation is generated by the heat emitted by objects, and this radiation can penetrate some objects or materials that visible light cannot penetrate. Therefore, infrared imaging technology is particularly suitable for applications such as night observation, target tracking, temperature measurement, and detection of hidden objects.

[0072] The method of obtaining the visible light image within the preset range of the cable can be to use a camera to shoot the cable and the image within the preset range, wherein the external light is obtained through the lens of the camera, and the external light is irradiated on the pixel array of the image sensor in the camera, a photoelectric effect occurs, and a corresponding charge is generated in the pixel unit. The row selection logic unit selects the corresponding row pixel unit as needed. The image signal in the row pixel unit is transmitted to the corresponding analog signal processing unit and the A / D converter through the signal bus of the respective column, and converted into a digital image signal output, thereby realizing the acquisition of the image within the preset range of the cable. Among them, the row selection logic unit can scan the pixel array row by row or interlaced. The row selection logic unit and the column selection logic unit can be used together to realize the window extraction function of the image. The main function of the analog signal processing unit is to amplify the signal and improve the signal-to-noise ratio.

[0073] The infrared image within the preset range of the cable can be obtained by using an infrared camera to shoot the cable and the images within the preset range, wherein the infrared sensor in the infrared camera mainly uses the sensitivity of semiconductor materials to infrared radiation to detect infrared signals. Specifically, the infrared light of the object enters the infrared sensor and is absorbed by the semiconductor material, causing the conductivity of the semiconductor material to change and generating a tiny current. By amplifying and filtering the tiny current signal, the infrared light signal emitted by the target object can be extracted.

[0074] The detection result acquisition module 102 may be a program design of the detection terminal for detecting visible light images and infrared images through a dual-channel model. Specifically, the dual-channel model may be a system or model with two parallel channels or paths. The two channels may work independently or interact or fuse information at certain points to achieve specific functions or goals.

[0075] The first channel may be one of the branches in a dual-channel model capable of processing visible light images.

[0076] The second channel can be another branch of the dual-channel model that can process infrared images. The first channel and the second channel can process the visible light image and the infrared image of the same shooting position in parallel, and then achieve the purpose of fire source verification through the mutual verification of the visible light image and the infrared image of the same fire source location.

[0077] The first detection result may be a detection result of whether a fire source exists in the visible light image obtained after being processed by the first channel.

[0078] The second detection result may be a detection result of whether a fire source exists in the infrared image obtained after being processed by the second channel.

[0079] The first detection result can be obtained by pre-acquiring multiple groups of visible light images in which it is known whether there is a fire source, using the presence of a fire source as a label and the visible light image as an input variable, performing supervised training, learning the discrimination mechanism of the fire source in the visible light image, and then constructing the first channel in the dual-channel detection model. The currently acquired visible light image is input into the first channel in the dual-channel detection model, and the first channel identifies whether there is a fire source in the current visible light image according to the pre-learned discrimination mechanism, and the output result of the first channel is used as the first detection result.

[0080] The second detection result can be obtained by pre-acquiring multiple groups of infrared images in which it is known whether there is a fire source, taking the presence of a fire source as a label and the infrared image as an input variable, performing supervised training, learning the discrimination mechanism of the fire source in the infrared image, and then constructing the second channel in the dual-channel detection model. The currently acquired infrared image is input into the second channel in the dual-channel detection model, and the second channel identifies whether there is a fire source in the current infrared image according to the pre-learned discrimination mechanism, and the output result of the second channel is used as the second detection result.

[0081] The fire source verification module 103 may be a program design for the detection terminal to verify the detection result of the existence of the fire source through the dual-channel model. Specifically, the abnormal range of the fire source in the visible light image may be the range where the fire source may exist in the visible light image. Since there may be flammable materials such as grass and shrubs at the cable installation location, after the fire source is generated, it will spread to the surroundings and form a certain burning range. The burning range detected in the visible light image is the abnormal range of the fire source in the visible light image.

[0082] The abnormal range of the infrared image may be a range where a fire source may exist in the infrared image. Since there may be flammable materials such as grass and shrubs at the location where the cable is installed, after a fire source is generated, it will spread to the surrounding area and form a certain burning range. The burning range detected in the infrared image is the abnormal range of the fire source in the infrared image.

[0083] The method for determining the abnormal range of the fire source in the visible light image can be to identify whether there is a fire source in the current visible light image according to the first channel in the dual-channel model, use the visible light image with a known abnormal range of the fire source as an input variable, and use the corresponding abnormal range of the fire source as a label to perform supervised training, obtain the correlation between the visible light image and the abnormal range of the fire source, and then obtain the first channel that can predict the abnormal range of the fire source according to the visible light image with the fire source. After determining that there is a fire source in the current visible light image, the first channel predicts the abnormal range of the fire source in the current visible light image with the fire source according to the pre-learned correlation between the visible light image and the abnormal range of the fire source, and uses the prediction result of the first channel as the abnormal range of the fire source in the visible light image.

[0084] The method for determining the abnormal range of the fire source in the infrared image can be to identify whether there is a fire source in the current infrared image according to the first channel in the dual-channel model, use the infrared image with a known abnormal range of the fire source as an input variable, and use the corresponding abnormal range of the fire source as a label to perform supervised training, obtain the correlation between the infrared image and the abnormal range of the fire source, and then obtain a second channel that can predict the abnormal range of the fire source according to the infrared image with the fire source. After determining that there is a fire source in the current infrared image, the second channel predicts the abnormal range of the fire source in the current infrared image with the fire source according to the pre-learned correlation between the infrared image and the abnormal range of the fire source, and uses the prediction result of the second channel as the abnormal range of the fire source in the infrared image.

[0085] The method for verifying the abnormal fire source range of the visible light image is to obtain the abnormal fire source range of the infrared image of the second channel after determining the abnormal fire source range of the visible light image, identify the degree of overlap between the abnormal fire source range of the visible light image and the abnormal fire source range of the infrared image, and when the degree of overlap reaches a preset threshold, determine that the verification of the abnormal fire source range of the visible light image is successful.

[0086] The method for verifying the abnormal fire source range of the infrared image is to obtain the abnormal fire source range of the visible light image of the first channel after determining the abnormal fire source range of the infrared image, identify the degree of overlap between the abnormal fire source range of the infrared image and the abnormal fire source range of the visible light image, and when the degree of overlap reaches a preset threshold, determine that the verification of the abnormal fire source range of the infrared image is successful.

[0087] The verification result output module 104 may be a program design of the detection terminal for determining the inferred location of the fire source and generating wildfire warning information. Specifically, the inferred location of the fire source may be a location where the verification successfully determines the existence of the fire source or is within the burning range of the fire source.

[0088] The method of generating wildfire warning information is to obtain the successfully verified abnormal range as the inferred location of the fire source after the abnormal range of the visible light image fire source or the abnormal range of the infrared image fire source is successfully verified, and write the inferred location of the fire source into the wildfire warning information to prompt relevant personnel to identify and control the wildfire at the inferred location of the fire source.

[0089] This solution provides a program code for processing visible light images and infrared images using a dual-channel detection model to confirm whether there is a fire source, for reference:

[0090] import torch

[0091] import torch.nn as nn

[0092] import torch.optim as optim

[0093] class DualChannelFireDetector(nn.Module):

[0094] def __init__(self):

[0095] super(DualChannelFireDetector,self).__init__()

[0096] #Each input image first passes through the same convolutional layer

[0097] self.shared_conv=nn.Sequential(

[0098] nn.Conv2d(1,16,kernel_size=3,padding=1),

[0099] nn.ReLU(),

[0100] nn.MaxPool2d(2) )

[0102] #Branch 1: Processing visible light images

[0103] self.visible_branch=nn.Sequential(

[0104] nn.Conv2d(16,32,kernel_size=3,padding=1),

[0105] nn.ReLU(),

[0106] nn.MaxPool2d(2),

[0107] nn.Flatten(),

[0108] nn.Linear(32*(input size / / 4)**2,64),

[0109] nn.ReLU() )

[0111] #Branch 2: Processing infrared images

[0112] self.infrared_branch=nn.Sequential(

[0113] nn.Conv2d(16,32,kernel_size=3,padding=1),

[0114] nn.ReLU(),

[0115] nn.MaxPool2d(2),

[0116] nn.Flatten(),

[0117] nn.Linear(32*(input size / / 4)**2,64),

[0118] nn.ReLU() )

[0120] #Merge two branches

[0121] self.fc = nn.Sequential(

[0122] nn.Linear(128,64),

[0123] nn.ReLU(),

[0124] nn.Linear(64,1),

[0125] nn.Sigmoid() )

[0127] def forward(self,visible,infrared):

[0128] visible=self.shared_conv(visible)

[0129] infrared=self.shared_conv(infrared)

[0130] visible_features=self.visible_branch(visible)

[0131] infrared_features=self.infrared_branch(infrared)

[0132] combined_features=torch.cat((visible_features,infrared_features),dim=1)

[0133] output=self.fc(combined_features)

[0134] return output

[0135] In an embodiment of the present application, a gas concentration monitoring module is arranged at a preset position in the cable trench, and is used to detect the concentration of at least one gas component at the preset position to obtain a concentration detection result of the gas component; a data transmission activation module is used to receive an activation instruction from an acquisition terminal to determine a target concentration detection result; a detection result sending module is used to encode the target concentration detection result and send it to the acquisition terminal at a preset frequency; the acquisition terminal is used to receive the target concentration detection result; a harmful gas analysis module is used to analyze the harmful gas in the cable trench according to the gas concentration detection result to obtain a harmful gas analysis result.

[0136] Through the above-mentioned cable surrounding wildfire detection device based on the dual-channel detection model, the gas situation in the cable trench can be monitored in real time, and the detection results can be transmitted through a contactless communication method, which is conducive to timely identification of harmful gas anomalies and ensuring the safety of workers entering the cable trench.

[0137] Embodiment 2

[0138] Figure 2 It is a structural diagram of a cable surrounding mountain fire detection device based on a dual-channel detection model provided in Example 2 of the present application. This solution makes a better improvement on the above-mentioned embodiment, and the specific improvement is as follows: the detection result acquisition module is specifically used to: extract image features in the visible light image, match the image features with the pre-acquired fire source features, if the match is successful, the first detection result is the existence of a fire source; and, identify the temperature distribution data of each area in the infrared image, and compare the temperature distribution data of each area with the preset temperature threshold respectively. If there is at least one area whose temperature value exceeds the preset temperature threshold, the second detection result is determined to be the existence of a fire source.

[0139] like Figure 2 As shown, specifically including the following:

[0140] An image acquisition module 201 is used to acquire a visible light image and an infrared image within a preset range of the cable;

[0141] The detection result acquisition module 202 is used to input the visible light image into the first channel of the dual-channel detection model to obtain a first detection result; and input the infrared image into the second channel of the dual-channel detection model to obtain a second detection result.

[0142] The fire source verification module 203 is used to determine the abnormal range of the fire source in the visible light image when the first detection result is that there is a fire source, and use the second detection result to verify the abnormal range of the fire source in the visible light image; or, when the second detection result is that there is a fire source, determine the abnormal range of the fire source in the infrared image, and use the first detection result to verify the abnormal range of the fire source in the infrared image.

[0143] The verification result output module 204 is used to determine the abnormal range where the verification is successful as the location where the fire source is inferred, and generate wildfire warning information with the location of the fire source inferred, if the verification is successful.

[0144] The detection result acquisition module 202 is specifically used to extract image features in the visible light image, and match the image features with the pre-acquired fire source features. If the match is successful, the first detection result is that there is a fire source.

[0145] as well as,

[0146] Identify the temperature distribution data of each area in the infrared image, and compare the temperature distribution data of each area with a preset temperature threshold. If the temperature value of at least one area exceeds the preset temperature threshold, determine that the second detection result is the presence of a fire source.

[0147] Image features can be the most basic features that distinguish one image from another and can be used as the iconic attribute of an image. Ideal feature descriptors should have the characteristics of repeatability, distinguishability, concentration, and efficiency; they also need to be able to cope with changes in image brightness, scale, rotation, and affine transformation.

[0148] The pre-acquired fire source features may be obtained by acquiring multiple groups of visible light images of historically known fire sources, performing feature extraction on the images of the fire source locations, and acquiring common features of the visible light images of the fire sources as the pre-acquired fire source features.

[0149] The temperature distribution data may be the distribution of the determined temperature data in each area of ​​the infrared image by displaying different colors of each area in the infrared image.

[0150] The preset temperature threshold can be obtained by pre-acquiring multiple groups of infrared images of historical fire sources, collecting statistics on the temperature data of the fire source locations, and then setting the minimum temperature data critical value that can determine whether there is a fire source in the infrared image based on the statistical results.

[0151] The method of extracting image features in a visible light image can be to preprocess the collected visible light image data, filter out noise, obtain edge information of objects in the visible light image by calculating the gradient of pixel values ​​in the visible light image data, and extract texture features in the visible light image using a gray level co-occurrence matrix or the like, and use the obtained edge information and texture features as image features of the current visible light image.

[0152] The method for determining that the first detection result is the existence of a fire source can be to analyze the image features of the current visible light image to identify whether the image features contain feature information consistent with the pre-acquired fire source features. If so, it is determined that the image features match the fire source features, indicating that there is a fire source in the current visible light image. The first detection result can be determined as the existence of a fire source.

[0153] The method of identifying the temperature distribution data of each area in the infrared image can be to correspond different colors in the infrared image to specific temperature values ​​according to the temperature scale marked in the infrared image by the infrared camera. For example, bright colors or warm tones (such as red and orange) are usually high-temperature areas, while dark colors or cold tones (such as blue and purple) are low-temperature areas. By observing the color differences of each area in the infrared image, the temperature distribution of each area can be determined.

[0154] The method for determining that the second detection result is the existence of a fire source can be to compare the temperature data of each area in the infrared image with a preset temperature threshold. If any data in the temperature data of each area exceeds the preset temperature threshold, the second detection result is determined to be the existence of a fire source; if no data in the temperature data of each area exceeds the preset temperature threshold, the second detection result is determined to be the absence of a fire source.

[0155] The advantage of this setting of the scheme is that it can determine whether there is a fire source in the visible light image and the infrared image through image feature matching and regional temperature value comparison, providing a judgment basis for the identification of the fire source and making the detection result more accurate.

[0156] Embodiment 3

[0157] Figure 3It is a structural diagram of a cable surrounding mountain fire detection device based on a dual-channel detection model provided in Example 3 of the present application. This solution makes a better improvement on the above-mentioned embodiment, and the specific improvement is: the fire source verification module is specifically used to: when the first detection result is that there is a fire source, perform fire source feature detection on the visible light image within the preset range of the fire source position, and use the area with fire source features within the preset range as the fire source abnormal range of the visible light image; or, when the second detection result is that there is a fire source, obtain temperature data within the preset range of the fire source position, and use the area where the temperature data is within the preset temperature threshold as the fire source abnormal range of the infrared image.

[0158] like Figure 3 As shown, specifically including the following:

[0159] An image acquisition module 301 is used to acquire a visible light image and an infrared image within a preset range of the cable;

[0160] The detection result acquisition module 302 is used to input the visible light image into the first channel of the dual-channel detection model to obtain a first detection result; and input the infrared image into the second channel of the dual-channel detection model to obtain a second detection result.

[0161] The fire source verification module 303 is used to determine the abnormal range of the fire source in the visible light image when the first detection result is that there is a fire source, and use the second detection result to verify the abnormal range of the fire source in the visible light image; or, when the second detection result is that there is a fire source, determine the abnormal range of the fire source in the infrared image, and use the first detection result to verify the abnormal range of the fire source in the infrared image.

[0162] The verification result output module 304 is used to determine the abnormal range where the verification is successful as the location where the fire source is inferred, and generate wildfire warning information with the location of the fire source inferred when the verification is successful.

[0163] The fire source verification module 303 is specifically used to perform fire source feature detection on the visible light image within a preset range of the fire source position when the first detection result is that there is a fire source, and to use the area with fire source features within the preset range as the fire source abnormal range of the visible light image;

[0164] or,

[0165] When the second detection result indicates that there is a fire source, temperature data within a preset range of the fire source location is obtained, and an area where the temperature data is within a preset temperature threshold is used as an abnormal fire source range of the infrared image.

[0166] The preset range of the fire source location can be that the staff predicts the fire source spread range in advance according to the fire source spread trend, and sets the fire source spread range as the preset range of the fire source location.

[0167] The method for obtaining the abnormal range of the fire source in the visible light image is that, when the first detection result is that there is a fire source, the fire source position is determined in the visible light image, and with the fire source position as the center, the image features within a preset range of the fire source position are matched with the fire source features, the matching visible light image areas are marked, and finally the marked areas are connected as the abnormal range of the fire source in the visible light image.

[0168] The method for obtaining the abnormal range of the fire source in the infrared image is that, when the second detection result is that there is a fire source, the fire source position is determined in the infrared image, and the temperature data within a preset range of the fire source position is compared with a preset temperature threshold with the fire source position as the center, and the infrared image area corresponding to the temperature data exceeding the preset temperature threshold is marked, and finally the marked areas are connected as the abnormal range of the fire source in the infrared image.

[0169] The advantage of this arrangement of the present scheme is that after confirming the existence of a fire source, the abnormal range of the fire source can be determined, which is conducive to improving the accuracy of subsequent verification of the detection results and providing an approximate range for the subsequent determination of the inferred location of the fire source.

[0170] Embodiment 4

[0171] Figure 4 It is a structural diagram of a cable surrounding mountain fire detection device based on a dual-channel detection model provided in Example 4 of the present application. This solution makes a better improvement on the above-mentioned embodiment, and the specific improvement is: the fire source verification module is specifically used to: when the first detection result is that there is a fire source, determine the abnormal range of the fire source in the visible light image, obtain the fire source confidence in the area corresponding to the abnormal range of the fire source in the visible light image in the second detection result, if the fire source confidence is greater than the confidence threshold, the verification is successful; or, when the second detection result is that there is a fire source, determine the abnormal range of the fire source in the infrared image, obtain the fire source confidence in the area corresponding to the abnormal range of the fire source in the infrared image in the first detection result, if the fire source confidence is greater than the confidence threshold, the verification is successful.

[0172] like Figure 4 As shown, specifically including the following:

[0173] An image acquisition module 401 is used to acquire a visible light image and an infrared image within a preset range of the cable;

[0174] The detection result acquisition module 402 is used to input the visible light image into the first channel of the dual-channel detection model to obtain a first detection result; and input the infrared image into the second channel of the dual-channel detection model to obtain a second detection result.

[0175] The fire source verification module 403 is used to determine the abnormal range of the fire source in the visible light image when the first detection result is that there is a fire source, and use the second detection result to verify the abnormal range of the fire source in the visible light image; or, when the second detection result is that there is a fire source, determine the abnormal range of the fire source in the infrared image, and use the first detection result to verify the abnormal range of the fire source in the infrared image.

[0176] The verification result output module 404 is used to determine the abnormal range where the verification is successful as the location where the fire source is inferred, and generate wildfire warning information with the location of the fire source inferred when the verification is successful.

[0177] The fire source verification module 403 is specifically used to determine the abnormal range of the fire source in the visible light image when the first detection result shows that there is a fire source, and obtain the fire source confidence in the area corresponding to the abnormal range of the fire source in the visible light image in the second detection result. If the fire source confidence is greater than the confidence threshold, the verification is successful.

[0178] or,

[0179] When the second detection result indicates that there is a fire source, determine the abnormal range of the fire source in the infrared image, and obtain the fire source confidence in the area corresponding to the abnormal range of the fire source in the infrared image in the first detection result. If the fire source confidence is greater than the confidence threshold, the verification is successful.

[0180] The fire source confidence may be the accuracy of the existence of the fire source in each area within the abnormal range of the fire source in the visible light image in the current detection result after verification.

[0181] The confidence threshold may be a preset confidence threshold for determining whether the verification is successful. For example, if the confidence threshold is set to 90%, only areas with a fire source confidence greater than 90% can be determined as having been successfully verified.

[0182] The method for obtaining the fire source confidence in the area corresponding to the fire source abnormal range of the visible light image in the second detection result can be, after determining the fire source abnormal range of the visible light image, matching the fire source abnormal range of the infrared image determined in the second detection result with the fire source abnormal range of the current visible light image, and taking the degree of matching between each area within the fire source abnormal range of the visible light image and the corresponding area in the infrared image as the fire source confidence in the area corresponding to the fire source abnormal range of the visible light image.

[0183] The verification success may be determined by comparing the fire source confidence of each area with the confidence threshold. If there is an area where the fire source confidence exceeds the confidence threshold, the verification is determined to be successful.

[0184] The method for obtaining the fire source confidence in the area corresponding to the fire source abnormality range of the infrared image in the first detection result can be, after determining the fire source abnormality range of the infrared image, matching the fire source abnormality range of the visible light image determined in the first detection result with the fire source abnormality range of the current infrared image, and taking the degree of matching between each area within the fire source abnormality range of the infrared image and the corresponding area in the visible light image as the fire source confidence in the area corresponding to the fire source abnormality range of the infrared image.

[0185] The advantage of this arrangement of the present scheme is that the detection result of another channel can be used to verify the detection result of the current channel, making the judgment of the fire source and the location of the fire source more accurate and avoiding economic losses caused by misjudgment.

[0186] Embodiment 5

[0187] Figure 5 This is a schematic diagram of the structure of a cable surrounding wildfire detection device based on a dual-channel detection model provided in Example 5 of the present application. This solution makes a better improvement on the above-mentioned embodiment, specifically: the harmful gas analysis module is specifically used for: the verification result output module is specifically used for: obtaining the area where the fire source confidence in the first detection result or the second detection result is greater than the confidence threshold as the inferred location of the fire source.

[0188] like Figure 5 As shown, specifically including the following:

[0189] An image acquisition module 501 is used to acquire a visible light image and an infrared image within a preset range of the cable;

[0190] The detection result acquisition module 502 is used to input the visible light image into the first channel of the dual-channel detection model to obtain a first detection result; and input the infrared image into the second channel of the dual-channel detection model to obtain a second detection result.

[0191] The fire source verification module 503 is used to determine the abnormal range of the fire source in the visible light image when the first detection result is that there is a fire source, and use the second detection result to verify the abnormal range of the fire source in the visible light image; or, when the second detection result is that there is a fire source, determine the abnormal range of the fire source in the infrared image, and use the first detection result to verify the abnormal range of the fire source in the infrared image.

[0192] The verification result output module 504 is used to determine the abnormal range where the verification is successful as the location where the fire source is inferred, and generate wildfire warning information with the location of the fire source inferred when the verification is successful.

[0193] The fire source verification module 503 is specifically used to determine the abnormal range of the fire source in the visible light image when the first detection result shows that there is a fire source, and obtain the fire source confidence in the area corresponding to the abnormal range of the fire source in the visible light image in the second detection result. If the fire source confidence is greater than the confidence threshold, the verification is successful.

[0194] or,

[0195] When the second detection result indicates that there is a fire source, determine the abnormal range of the fire source in the infrared image, and obtain the fire source confidence in the area corresponding to the abnormal range of the fire source in the infrared image in the first detection result. If the fire source confidence is greater than the confidence threshold, the verification is successful.

[0196] The verification result output module 505 is specifically used to obtain an area in the first detection result or the second detection result where the fire source confidence is greater than a confidence threshold as the inferred location of the fire source.

[0197] The method for determining the inferred location of the fire source is that, after successful verification, the corresponding area within the fire source abnormal range of the visible light image in the first detection result where the fire source confidence is greater than the confidence threshold is used as the inferred location of the fire source; or the corresponding area within the fire source abnormal range of the infrared image in the second detection result where the fire source confidence is greater than the confidence threshold is used as the inferred location of the fire source.

[0198] The advantage of this setting is that the inferred location of the fire source can be determined through the fire source confidence level, so that the staff can have a more comprehensive and clear understanding of the spread of the fire source, which is conducive to the staff to take corresponding measures to deal with the fire.

[0199] Embodiment 6

[0200] Figure 6It is a structural diagram of a cable surrounding mountain fire detection device based on a dual-channel detection model provided in Example 6 of the present application. This solution makes a better improvement on the above-mentioned embodiment, and the specific improvement is: the detection result acquisition module is also used to: input the visible light image into the first channel in the dual-channel detection model to extract the first feature; and input the infrared image into the second channel in the dual-channel detection model to extract the second feature; fuse the first feature with the second feature to obtain the result feature; match the result feature with the fire source feature to determine the candidate area; correspondingly, the fire source verification module is also used to: identify whether there are multiple candidate areas in the candidate area that overlap and point to the same fire source. If so, the redundant areas are filtered out and the remaining candidate areas are retained as the abnormal range of successful verification.

[0201] like Figure 6 As shown, specifically including the following:

[0202] An image acquisition module 601 is used to acquire a visible light image and an infrared image within a preset range of the cable;

[0203] The detection result acquisition module 602 is used to input the visible light image into the first channel of the dual-channel detection model to obtain a first detection result; and input the infrared image into the second channel of the dual-channel detection model to obtain a second detection result.

[0204] The fire source verification module 603 is used to determine the abnormal range of the fire source in the visible light image when the first detection result is that there is a fire source, and use the second detection result to verify the abnormal range of the fire source in the visible light image; or, when the second detection result is that there is a fire source, determine the abnormal range of the fire source in the infrared image, and use the first detection result to verify the abnormal range of the fire source in the infrared image.

[0205] The verification result output module 604 is used to determine the abnormal range of the successful verification as the location of the inferred fire source when the verification is successful, and generate wildfire warning information with the inferred fire source location.

[0206] The detection result acquisition module 602 is further used to input the visible light image into the first channel of the dual-channel detection model to extract the first feature;

[0207] as well as,

[0208] Inputting the infrared image into the second channel of the dual-channel detection model to extract the second feature;

[0209] Fusing the first feature with the second feature to obtain a result feature;

[0210] Matching the result features with the fire source features to determine the candidate area;

[0211] The fire source verification module 603 is also used to identify whether there are multiple candidate areas overlapping and pointing to the same fire source. If so, the redundant areas are filtered out and the remaining candidate areas are retained as abnormal ranges for successful verification.

[0212] The first feature may be an image feature of a visible light image.

[0213] The second feature may be an image feature of the infrared image.

[0214] The result feature can be a feature obtained by fusing the first feature with the second feature in some way, such as concatenation, addition, or attention mechanism. The fused feature contains information from two channels, which helps to describe the target more comprehensively.

[0215] The candidate area may be an area preliminarily determined to have a fire source.

[0216] The first feature and the second feature can be extracted by inputting the visible light image and the infrared image into the first channel and the second channel respectively, and the first channel and the second channel respectively extract features of the visible light image and the infrared image through a series of convolution layers. The convolution layer uses different convolution kernels to capture local features in the image, such as edges, textures, etc. It is understandable that since the data attributes of the first channel and the second channel are different, the convolution layer parameters of each channel, such as the convolution kernel size, step size, and padding, may also be different.

[0217] The result feature can be obtained by concatenating the feature map with the first feature from the first channel with the feature map with the second feature from the second channel in a specific dimension (usually the channel dimension) to form a new feature map as the result feature. This fusion method retains the information of all the original features and increases the number of channels of the feature map.

[0218] The candidate area can be determined by matching the result features of each area with the fire source features acquired in advance. If the matches are consistent, the area is used as a candidate area; if the matches are inconsistent, the current area is not a candidate area.

[0219] The way to determine the abnormal range of successful verification can be to determine whether the candidate area contains the fire source for each candidate area. This is usually achieved through a fully connected layer or a convolutional layer followed by a softmax classifier. Adjust the position and size of the candidate area so that it more accurately surrounds the location of the fire source. This is usually achieved through a fully connected layer or a convolutional layer followed by a bounding box regressor. After obtaining the processing results of multiple candidate areas, there may be a situation in the results where multiple candidate areas overlap and point to the same target. At this time, it is necessary to use a non-maximum suppression algorithm to eliminate redundant candidate areas and retain the optimal candidate area as the abnormal range of the final successful verification.

[0220] The advantage of this setting is that the dual-channel detection model can be used to fuse the visible light image features and infrared image features of the same location and verify the abnormal range, which is conducive to more accurate judgment of whether a fire source exists and the spread of the fire source, and improves the credibility of wildfire detection results.

[0221] Embodiment 7

[0222] Figure 7 This is a structural diagram of a cable surrounding mountain fire detection device based on a dual-channel detection model provided in Example 7 of the present application. This solution makes a better improvement on the above-mentioned embodiment, and the specific improvement is: the detection result acquisition module is specifically used to: obtain the first weight of the first channel and the second weight of the second channel that are preset; the first feature and the second feature are feature-fused according to the ratio of the first weight to the second weight to obtain the result feature.

[0223] like Figure 7 As shown, specifically including the following:

[0224] An image acquisition module 701 is used to acquire a visible light image and an infrared image within a preset range of the cable;

[0225] The detection result acquisition module 702 is used to input the visible light image into the first channel of the dual-channel detection model to obtain a first detection result; and input the infrared image into the second channel of the dual-channel detection model to obtain a second detection result.

[0226] The fire source verification module 703 is used to determine the abnormal range of the fire source in the visible light image when the first detection result is that there is a fire source, and use the second detection result to verify the abnormal range of the fire source in the visible light image; or, when the second detection result is that there is a fire source, determine the abnormal range of the fire source in the infrared image, and use the first detection result to verify the abnormal range of the fire source in the infrared image.

[0227] The verification result output module 704 is used to determine the abnormal range of the successful verification as the location of the inferred fire source when the verification is successful, and generate wildfire warning information with the inferred fire source location.

[0228] The detection result acquisition module 702 is specifically used to acquire a preset first weight of the first channel and a second weight of the second channel;

[0229] The first feature and the second feature are fused according to a ratio of the first weight to the second weight to obtain a result feature.

[0230] The first weight of the first channel may be the importance or contribution of the output result of the first channel to the fusion result when the first feature and the second feature are fused.

[0231] The second weight of the second channel may be the importance or contribution of the output result of the second channel to the fusion result when the first feature and the second feature are fused.

[0232] The method for obtaining the first weight and the second weight is that the staff can determine the degree of influence of the visible light image and the infrared image on the fire source detection results based on the accuracy of the results when the visible light image and the infrared image are separately performed for fire source detection, and then determine the first weight of the first channel for processing the visible light image and the second weight of the second channel for processing the infrared image.

[0233] The first feature and the second feature are fused in accordance with the ratio of the first weight to the second weight to obtain the result feature. After obtaining the first feature of the visible light image and the second feature of the infrared image, the first feature in the visible light image is multiplied by the first weight to obtain a feature weight map of the visible light image. After multiplying the second feature in the infrared image by the second weight, the second feature is summed with the corresponding value on the feature weight map to obtain the fused result feature.

[0234] The advantage of this arrangement of the present invention is that the characteristic values ​​in the visible light image or the infrared image can be recalibrated, so that the dual-channel detection model can pay more attention to important features related to the fire source in subsequent processing, thereby improving the accuracy of fire source determination.

[0235] Embodiment 8

[0236] Figure 8 It is a flow chart of a cable surrounding wildfire detection method based on a dual-channel detection model provided in Example 8 of the present application.

[0237] like Figure 8 As shown, the specific steps include:

[0238] S801, obtaining a visible light image and an infrared image within a preset range of the cable;

[0239] S802, inputting the visible light image into a first channel in a dual-channel detection model to obtain a first detection result; and inputting the infrared image into a second channel in the dual-channel detection model to obtain a second detection result;

[0240] S803, when the first detection result indicates that a fire source exists, determining an abnormal range of the fire source in the visible light image, and verifying the abnormal range of the fire source in the visible light image using the second detection result; or, when the second detection result indicates that a fire source exists, determining an abnormal range of the fire source in the infrared image, and verifying the abnormal range of the fire source in the infrared image using the first detection result;

[0241] S804. When the verification is successful, the abnormal range where the verification is successful is determined as the location where the fire source is inferred, and wildfire warning information with the location of the fire source inferred is generated.

[0242] Optionally, inputting the visible light image into a first channel in a dual-channel detection model to obtain a first detection result; and inputting the infrared image into a second channel in the dual-channel detection model to obtain a second detection result, including:

[0243] Extracting image features from the visible light image, matching the image features with pre-acquired fire source features, and if the match is successful, the first detection result is that there is a fire source;

[0244] as well as,

[0245] Identify the temperature distribution data of each area in the infrared image, and compare the temperature distribution data of each area with a preset temperature threshold. If the temperature value of at least one area exceeds the preset temperature threshold, determine that the second detection result is the presence of a fire source.

[0246] In an embodiment of the present application, a visible light image and an infrared image within a preset range of the cable are obtained; the visible light image is input into a first channel in a dual-channel detection model to obtain a first detection result; and the infrared image is input into a second channel in a dual-channel detection model to obtain a second detection result; when the first detection result is that a fire source exists, the abnormal range of the fire source in the visible light image is determined, and the abnormal range of the fire source in the visible light image is verified using the second detection result; or, when the second detection result is that a fire source exists, the abnormal range of the fire source in the infrared image is determined, and the abnormal range of the fire source in the infrared image is verified using the first detection result; when the verification is successful, the abnormal range with successful verification is determined as the inferred location of the fire source, and a forest fire alarm message with the inferred location of the fire source is generated.

[0247] Through the cable periphery wildfire detection method based on the above-mentioned dual-channel detection model, the gas situation in the cable trench can be monitored in real time, and the detection results can be transmitted through a contactless communication method, which is conducive to timely identification of harmful gas anomalies and ensuring the safety of workers entering the cable trench.

[0248] The cable surrounding mountain fire detection method based on the dual-channel detection model provided in the embodiment of the present application corresponds to the cable surrounding mountain fire detection device based on the dual-channel detection model provided in the above embodiment, has the same functional modules and beneficial effects, and will not be repeated here to avoid repetition.

[0249] Embodiment 9

[0250] like Fig. 9 As shown, an embodiment of the present application also provides an electronic device 900, including a processor 901, a memory 902, and a program or instruction stored in the memory 902 and executable on the processor 901. When the program or instruction is executed by the processor 901, each process of the embodiment of the cable surrounding mountain fire detection device based on the dual-channel detection model is implemented, and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.

[0251] It should be noted that the electronic devices in the embodiments of the present application include the mobile electronic devices and non-mobile electronic devices mentioned above.

[0252] Embodiment 10

[0253] An embodiment of the present application also provides a readable storage medium, on which a program or instruction is stored. When the program or instruction is executed by a processor, each process of the above-mentioned cable surrounding mountain fire detection device embodiment based on a dual-channel detection model is implemented, and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.

[0254] The processor is a processor in the electronic device described in the above embodiment. The readable storage medium includes a computer readable storage medium, such as a computer read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0255] Embodiment 11

[0256] An embodiment of the present application further provides a chip, which includes a processor and a communication interface, wherein the communication interface is coupled to the processor, and the processor is used to run programs or instructions to implement the various processes of the above-mentioned cable surrounding mountain fire detection device embodiment based on the dual-channel detection model, and can achieve the same technical effect. To avoid repetition, it will not be repeated here.

[0257] It should be understood that the chip mentioned in the embodiments of the present application can also be called a system-level chip, a system chip, a chip system or a system-on-chip chip, etc.

[0258] It should be noted that, in this article, the terms "comprise", "include" or any other variant thereof are intended to cover non-exclusive inclusion, so that the process, method, article or device including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise one..." do not exclude the presence of other identical elements in the process, method, article or device including the element. In addition, it should be pointed out that the scope of the methods and devices in the embodiments of the present application is not limited to performing functions in the order shown or discussed, and may also include performing functions in a substantially simultaneous manner or in reverse order according to the functions involved, for example, the described method may be performed in an order different from that described, and various steps may also be added, omitted, or combined. In addition, the features described with reference to certain examples may be combined in other examples.

[0259] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a computer software product, which is stored in a storage medium (such as ROM / RAM, a disk, or an optical disk), and includes a number of instructions for a terminal (which can be a mobile phone, a computer, a server, or a network device, etc.) to execute the methods described in each embodiment of the present application.

[0260] The embodiments of the present application are described above in conjunction with the accompanying drawings, but the present application is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of the present application, ordinary technicians in this field can also make many forms without departing from the purpose of the present application and the scope of protection of the claims, all of which are within the protection of the present application.

[0261] The above are only preferred embodiments of the present application and the technical principles used. The present application is not limited to the specific embodiments described herein, and various obvious changes, readjustments and substitutions that can be made by those skilled in the art will not deviate from the scope of protection of the present application. Therefore, although the present application is described in more detail through the above embodiments, the present application is not limited to the above embodiments, and may include more other equivalent embodiments without departing from the concept of the present application, and the scope of the present application is determined by the scope of the claims.

Claims

1. A cable surrounding mountain fire detection device based on a dual-channel detection model, characterized in that: The device comprises: An image acquisition module, used to acquire visible light images and infrared images within a preset range of the cable; A detection result acquisition module, used for inputting the visible light image into the first channel of the dual-channel detection model to obtain a first detection result; and inputting the infrared image into the second channel of the dual-channel detection model to obtain a second detection result; A fire source verification module, for determining an abnormal range of the fire source in the visible light image when the first detection result indicates that a fire source exists, and verifying the abnormal range of the fire source in the visible light image using the second detection result; or, for determining an abnormal range of the fire source in the infrared image when the second detection result indicates that a fire source exists, and verifying the abnormal range of the fire source in the infrared image using the first detection result; The verification result output module is used to determine the abnormal range of the successful verification as the inferred location of the fire source when the verification is successful, and generate wildfire alarm information with the inferred location of the fire source.

2. The cable surrounding mountain fire detection device based on the dual-channel detection model according to claim 1 is characterized in that: The detection result acquisition module is specifically used for: Extracting image features from the visible light image, matching the image features with pre-acquired fire source features, and if the match is successful, the first detection result is that there is a fire source; as well as, Identify the temperature distribution data of each area in the infrared image, and compare the temperature distribution data of each area with a preset temperature threshold. If the temperature value of at least one area exceeds the preset temperature threshold, determine that the second detection result is the presence of a fire source.

3. The cable surrounding mountain fire detection device based on the dual-channel detection model according to claim 1 is characterized in that: The fire source verification module is specifically used for: When the first detection result indicates that there is a fire source, fire source feature detection is performed on the visible light image within a preset range of the fire source location, and the area with fire source features within the preset range is used as the fire source abnormal range of the visible light image; or, When the second detection result indicates that there is a fire source, temperature data within a preset range of the fire source location is obtained, and an area where the temperature data is within a preset temperature threshold is used as an abnormal fire source range of the infrared image.

4. The cable surrounding mountain fire detection device based on the dual-channel detection model according to claim 1 is characterized in that: The fire source verification module is specifically used for: When the first detection result indicates that there is a fire source, determining the abnormal range of the fire source in the visible light image, obtaining the fire source confidence in the area corresponding to the abnormal range of the fire source in the visible light image in the second detection result, and if the fire source confidence is greater than the confidence threshold, the verification is successful; or, When the second detection result indicates that there is a fire source, determine the abnormal range of the fire source in the infrared image, and obtain the fire source confidence in the area corresponding to the abnormal range of the fire source in the infrared image in the first detection result. If the fire source confidence is greater than the confidence threshold, the verification is successful.

5. The cable surrounding mountain fire detection device based on the dual-channel detection model according to claim 4 is characterized in that: The verification result output module is specifically used for: The area where the fire source confidence is greater than the confidence threshold in the first detection result or the second detection result is obtained as the inferred location of the fire source.

6. The cable surrounding mountain fire detection device based on the dual-channel detection model according to claim 1 is characterized in that: The detection result acquisition module is further used for: Inputting the visible light image into a first channel of a dual-channel detection model to extract a first feature; as well as, Inputting the infrared image into the second channel of the dual-channel detection model to extract the second feature; Fusing the first feature with the second feature to obtain a result feature; Matching the result features with the fire source features to determine the candidate area; Accordingly, the fire source verification module is also used for: Identify whether there are multiple candidate areas overlapping and pointing to the same fire source in the candidate area. If so, filter out the redundant areas and retain the remaining candidate areas as the abnormal range of successful verification.

7. The cable surrounding mountain fire detection device based on the dual-channel detection model according to claim 6 is characterized in that: The detection result acquisition module is specifically used for: Obtaining a preset first weight of a first channel and a second weight of a second channel; The first feature and the second feature are fused according to a ratio of the first weight to the second weight to obtain a result feature.

8. A method for detecting mountain fires around cables based on a dual-channel detection model, characterized in that: The method comprises: Obtain visible light images and infrared images within the preset range of the cable; Inputting the visible light image into a first channel in a dual-channel detection model to obtain a first detection result; and inputting the infrared image into a second channel in the dual-channel detection model to obtain a second detection result; When the first detection result indicates that a fire source exists, determining the abnormal range of the fire source in the visible light image, and verifying the abnormal range of the fire source in the visible light image using the second detection result; or, when the second detection result indicates that a fire source exists, determining the abnormal range of the fire source in the infrared image, and verifying the abnormal range of the fire source in the infrared image using the first detection result; When the verification is successful, the abnormal range of the successful verification is determined as the inferred location of the fire source, and wildfire alarm information with the inferred location of the fire source is generated.

9. The cable surrounding wildfire detection method based on the dual-channel detection model according to claim 8, characterized in that: Inputting the visible light image into a first channel of a dual-channel detection model to obtain a first detection result; And, inputting the infrared image into the second channel of the dual-channel detection model to obtain a second detection result, including: Extracting image features from the visible light image, matching the image features with pre-acquired fire source features, and if the match is successful, the first detection result is that there is a fire source; as well as, Identify the temperature distribution data of each area in the infrared image, and compare the temperature distribution data of each area with a preset temperature threshold. If the temperature value of at least one area exceeds the preset temperature threshold, determine that the second detection result is the presence of a fire source.

10. An electronic device, characterized in that: It includes a processor, a memory, and a program or instruction stored in the memory and executable on the processor. When the program or instruction is executed by the processor, the steps of the cable perimeter mountain fire detection method based on a dual-channel detection model as described in any one of claims 8 to 9 are implemented.

Images