Method and device for monitoring state of train roof device and train

Abstract

The invention provides a train roof device state monitoring method and device and a train, and the method comprises the steps that in response to a state monitoring instruction, a target thermal imager is used for collecting a detection data set of a train roof, the detection data set comprises a thermal imaging sequence containing various roof devices, and the thermal imaging sequence comprises initial infrared images on the time sequence; for each initial infrared image, inputting the initial infrared image into a target detection model, and outputting target infrared images corresponding to different roof devices, the target detection model being obtained by training a training channel image collected by a target thermal imager; and for each roof device, abnormal temperature outlier detection processing is carried out on the plurality of target infrared images to obtain a first state monitoring result, and the first state monitoring result represents whether the roof device is in a temperature abnormal state or not.

Description

Technical Field

[0001] The present application relates to the field of electrified rail transit, and more specifically, to a method for monitoring the status of a train roof component, a status monitoring device, a train, an electronic device, a computer-readable storage medium, and a computer program product. Background Art

[0002] EMU, or power vehicle, refers to vehicles equipped with power devices in rail transit systems, including locomotives and power cars. EMUs are equipped with driving wheels, while the corresponding vehicles without driving devices are trailers. For a train to run normally on the track, it must have an EMU to provide sufficient traction for the entire train, but it does not need to be attached to a trailer without power.

[0003] The operation of high-speed trains has changed the way people travel. High-speed trains allow people to travel from one city to another quickly, safely and comfortably, greatly shortening travel time and bringing great convenience to people's lives and work. At the same time, high-speed trains have promoted social and economic development. The operation of high-speed trains has driven the economic development of the areas along the route and promoted the flow of people and goods.

[0004] In the process of realizing the concept of this application, it was found that during the operation of trains such as EMUs, high-voltage devices in contact with the distribution network are more likely to have problems, thereby posing risks to the operation of the trains. Summary of the invention

[0005] In view of this, the present application provides a method for monitoring the status of a train roof component, a status monitoring device, a train, an electronic device, a computer-readable storage medium, and a computer program product.

[0006] One aspect of the present application provides a method for monitoring the status of a train roof device, comprising:

[0007] In response to the status monitoring instruction, using the target thermal imager to collect a detection data set of the train roof, wherein the detection data set includes a thermal imaging sequence including a plurality of roof components, and the thermal imaging sequence includes an initial infrared image in a time sequence;

[0008] For each of the above initial infrared images, the above initial infrared images are input into a target detection model, and target infrared images corresponding to different above roof components are output, wherein the above target detection model is trained using the training channel images collected by the above target thermal imager;

[0009] For each of the above-mentioned roof components, abnormal temperature outlier detection processing is performed on the multiple above-mentioned target infrared images to obtain a first state monitoring result, wherein the above-mentioned first state monitoring result represents whether the above-mentioned roof component is in a state of abnormal temperature.

[0010] According to an embodiment of the present application, the detection data set further includes a channel image sequence including a plurality of roof components, and the channel image sequence includes a plurality of initial channel images in a time sequence.

[0011] According to an embodiment of the present application, the above-mentioned state monitoring method further includes:

[0012] For each of the initial channel images, input the initial channel image into the target detection model, and output a target channel image corresponding to the different roof components;

[0013] For each of the above target channel images, the above target channel image is compared and analyzed with the reference channel image to generate a second state monitoring result, wherein the above second state monitoring result represents whether the external shape of the above roof device is in an abnormal state.

[0014] According to an embodiment of the present application, the target channel image is compared and analyzed with the reference channel image to generate a second state monitoring result, including:

[0015] Calculate the image similarity between the target channel image and the reference channel image;

[0016] When the image similarity is less than the similarity threshold, generating an appearance abnormality result, wherein the appearance abnormality result indicates that the appearance of the roof component is abnormal;

[0017] When the image similarity is not less than the similarity threshold, a normal appearance result is generated, wherein the second state monitoring result includes the abnormal appearance result or the normal appearance result.

[0018] According to an embodiment of the present application, abnormal temperature detection processing is performed on a plurality of the above-mentioned target infrared images to obtain a first state monitoring result, including:

[0019] Based on the outlier detection algorithm, abnormal temperature detection processing is performed on the plurality of target infrared images to obtain the first state monitoring result.

[0020] According to an embodiment of the present application, abnormal temperature outlier detection processing is performed on the plurality of target infrared images based on an outlier detection algorithm to obtain the first state monitoring result, including:

[0021] For each of the above target infrared images, feature extraction processing is performed on the above target infrared images to obtain a plurality of infrared temperature information, wherein each of the above infrared temperature information corresponds to a device area in the above roof device;

[0022] For each of the device regions, performing outlier detection processing on the plurality of infrared temperature information corresponding to the device region to obtain a temperature detection result corresponding to the device region;

[0023] The first state monitoring result is generated according to the temperature detection results of the plurality of device regions.

[0024] According to an embodiment of the present application, before performing feature extraction processing, it also includes:

[0025] The target infrared image is subjected to Gaussian filtering to obtain a filtered target infrared image, and feature extraction is performed on the filtered target infrared image.

[0026] According to an embodiment of the present application, outlier detection processing is performed on the plurality of infrared temperature information corresponding to the device area to obtain a temperature detection result corresponding to the device area, including:

[0027] In the case where any of the above infrared temperature information is greater than a preset temperature threshold, a temperature anomaly result is generated, wherein the above temperature anomaly result indicates that a temperature anomaly exists in the above device area;

[0028] In the case that any of the above infrared temperature information is not greater than the preset temperature threshold, a normal temperature result is generated, wherein the above temperature detection result includes the above abnormal temperature result or the above normal temperature result.

[0029] According to an embodiment of the present application, the above-mentioned temperature detection result includes an abnormal temperature result or a normal temperature result.

[0030] According to an embodiment of the present application, the first state monitoring result is generated according to the temperature detection results of the plurality of device regions, including:

[0031] In the case where any of the temperature detection results is a temperature abnormality result, generating a state abnormality result, wherein the state abnormality result indicates that the temperature of the roof component is abnormal;

[0032] In the case that the plurality of temperature detection results are all normal temperature results, a normal state result is generated, wherein the first state monitoring result includes the abnormal state result or the normal state result.

[0033] According to an embodiment of the present application, the state monitoring method further includes:

[0034] For each of the temperature detection results, when the temperature detection result is an abnormal temperature result, a device region corresponding to the temperature detection result in the target infrared image is marked to obtain a marked infrared image;

[0035] A target monitoring set is displayed, wherein the target monitoring set includes at least one of the following: the first state monitoring result and the marked infrared image.

[0036] According to an embodiment of the present application, the above target detection model is trained in the following manner:

[0037] Acquire a training data set, wherein the training data set includes a plurality of the training channel images acquired by the target thermal imager and a label image corresponding to a different roof component in each of the training channel images, wherein the label image represents a device segmentation range of the roof component;

[0038] For each of the training channel images, input the training channel images into the initial visual model, and output predicted segmented images of different roof components;

[0039] For each of the above-mentioned roof devices, a target loss value is calculated according to the above-mentioned predicted segmentation image and the above-mentioned label image, so as to iteratively adjust the model parameters of the above-mentioned initial visual model according to the above-mentioned target loss value to obtain the above-mentioned target detection model.

[0040] Another aspect of the present application provides a state monitoring device for a train roof device, comprising:

[0041] A data acquisition module, configured to collect a detection data set of the train roof using a target thermal imager in response to a state monitoring instruction, wherein the detection data set includes a thermal imaging sequence including a plurality of roof components, and the thermal imaging sequence includes an initial infrared image in a time sequence;

[0042] A data processing module, for each of the initial infrared images, inputting the initial infrared image into a target detection model, and outputting target infrared images corresponding to different roof components, wherein the target detection model is trained using training channel images acquired by the target thermal imager;

[0043] The state monitoring module is used to perform abnormal temperature outlier detection processing on the plurality of target infrared images for each of the above-mentioned roof components to obtain a first state monitoring result, wherein the above-mentioned first state monitoring result indicates whether the above-mentioned roof component is in a state of abnormal temperature.

[0044] Another aspect of the present application provides a train, including: a status monitoring device for a train body and a train roof device.

[0045] Another aspect of the present application provides an electronic device, comprising:

[0046] one or more processors;

[0047] a memory for storing one or more programs,

[0048] When the one or more programs are executed by the one or more processors, the one or more processors implement the method as described above.

[0049] Another aspect of the present application provides a computer-readable storage medium storing computer-executable instructions, which are used to implement the method described above when executed.

[0050] Another aspect of the present application provides a computer program product, which includes computer executable instructions, and the instructions are used to implement the method as described above when executed.

[0051] According to the embodiment of the present application, by inputting the thermal imaging sequence of the roof device collected by the target thermal imager into the trained target detection model, the initial infrared image is segmented into corresponding target infrared images according to different roof devices, and then the target infrared image sequence of the same type of roof device is subjected to abnormal temperature outlier detection processing to determine whether the roof device has temperature abnormality. By accurately dividing different roof devices and performing abnormal temperature outlier detection on the divided images, it is possible to accurately identify whether the roof device is abnormal even in a complex train background environment, thereby improving the accuracy of abnormal roof device occurrence and reducing the accident rate of the train. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] The above and other objects, features and advantages of the present application will become more apparent through the following description of the embodiments of the present application with reference to the accompanying drawings, in which:

[0053] Figure 1 An exemplary system architecture to which a method for monitoring the condition of a train roof device according to an embodiment of the present application can be applied is shown;

[0054] Figure 2 A flow chart showing a method for monitoring the status of a train roof device according to an embodiment of the present application is shown;

[0055] Figure 3 A flow chart showing a method for monitoring the status of a train roof device according to another embodiment of the present application is shown;

[0056] Figure 4 A block diagram of a train roof device status monitoring device according to an embodiment of the present application is shown;

[0057] Figure 5 A block diagram of a training device for a target detection model according to an embodiment of the present application is shown;

[0058] Figure 6 A schematic diagram of a train according to an embodiment of the present application is shown;

[0059] Figure 7 A block diagram of an electronic device suitable for implementing the method described above according to an embodiment of the present application is shown. DETAILED DESCRIPTION

[0060] Below, embodiments of the present application will be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of the present application. In the following detailed description, for ease of explanation, many specific details are set forth to provide a comprehensive understanding of the embodiments of the present application. However, it is apparent that one or more embodiments may also be implemented without these specific details. In addition, in the following description, descriptions of known structures and technologies are omitted to avoid unnecessary confusion of the concepts of the present application.

[0061] The terms used herein are only for describing specific embodiments and are not intended to limit the present application. The terms "include", "comprising", etc. used herein indicate the existence of the features, steps, operations and / or components, but do not exclude the existence or addition of one or more other features, steps, operations or components.

[0062] All terms (including technical and scientific terms) used herein have the meanings commonly understood by those skilled in the art unless otherwise defined. It should be noted that the terms used herein should be interpreted as having a meaning consistent with the context of this specification and should not be interpreted in an idealized or overly rigid manner.

[0063] When using expressions such as "at least one of A, B, and C, etc.", they should generally be interpreted according to the meaning of the expression commonly understood by those skilled in the art (for example, "a system having at least one of A, B, and C" should include but is not limited to a system having A alone, B alone, C alone, A and B, A and C, B and C, and / or A, B, C, etc.).

[0064] Pantographs and insulators are key components for trains to obtain electrical energy from the overhead contact network. If these components fail, it may cause a power outage, affect the normal operation of the train, and even cause fires or electrical accidents in serious cases, endangering the safety of passengers and staff. In order to ensure the safety of staff and passengers and minimize losses, a system that can monitor such high-voltage devices in real time is needed. In the existing train roof high-voltage device status monitoring technology, it is a common practice to use a camera to shoot the train roof, obtain video data, and monitor the status of high-voltage devices such as pantographs and insulators in real time. However, there are the following problems in the monitoring process:

[0065] Existing monitoring methods mainly monitor high-voltage devices based on RGB video data. During the monitoring process, there is a problem that when the ambient light changes dramatically or the background is complex, the recognition accuracy of the visual algorithm will be affected, resulting in a high bit error rate. The current solution is to improve the image quality by adding image preprocessing steps, but there are still the following disadvantages: the preprocessing process increases the computational complexity, resulting in a slow response speed, and it is impossible to completely eliminate the impact of environmental factors on recognition accuracy.

[0066] In addition, although the use of thermal imaging data for outlier analysis and abnormal mutation detection of local temperature measurement points can determine the status of high-voltage devices, there are still the following inappropriate aspects: pure thermal imaging data weakens the characteristics of different devices considerably. Faced with a complex system such as the train roof, the edge features of thermal imaging data are also difficult to utilize, so the type of faulty device cannot be determined. Traditional thermal imaging analysis methods have low differentiation of device states under different working conditions and are prone to false alarms for high-voltage devices whose types cannot be determined.

[0067] In view of this, an embodiment of the present application provides a method, apparatus, train, equipment and medium for monitoring the status of a train roof device, the method comprising: responding to a status monitoring instruction, using a target thermal imager to collect a detection data set of the train roof, wherein the detection data set comprises a thermal imaging sequence including a plurality of roof devices, and the thermal imaging sequence comprises an initial infrared image in a time sequence; for each initial infrared image, the initial infrared image is input into a target detection model, and target infrared images corresponding to different roof devices are output, wherein the target detection model is trained using training channel images collected by a target thermal imager; for each roof device, abnormal temperature outlier detection processing is performed on multiple target infrared images to obtain a first status monitoring result, wherein the first status monitoring result characterizes whether the roof device is in an abnormal temperature state.

[0068] In the embodiments of this application, the collection, updating, analysis, processing, use, transmission, provision, disclosure, storage, etc. of the data involved (for example, including but not limited to user personal information) are in compliance with the provisions of relevant laws and regulations, are used for legitimate purposes, and do not violate public order and good morals. In particular, necessary measures are taken for user personal information to prevent illegal access to user personal information data and maintain the security of user personal information and network security.

[0069] Figure 1 An exemplary system architecture 100 to which a method for monitoring the condition of a train roof device according to an embodiment of the present application can be applied is shown. It should be noted that: Figure 1What is shown is merely an example of a system architecture to which the embodiments of the present application can be applied, in order to help those skilled in the art understand the technical content of the present application, but it does not mean that the embodiments of the present application cannot be used in other devices, systems, environments or scenarios.

[0070] like Figure 1 As shown, the system architecture 100 according to this embodiment may include a first terminal device 101, a second terminal device 102, a third terminal device 103, a network 104, a server 105, and a train 106. The network 104 is used to provide a medium for communication links between the first terminal device 101, the second terminal device 102, the third terminal device 103, and the server 105. The network 104 may include various connection types, such as wired and / or wireless communication links, etc.

[0071] The user may use the first terminal device 101, the second terminal device 102, and the third terminal device 103 to interact with the server 105 through the network 104 to receive or send messages, etc. Various communication client applications may be installed on the first terminal device 101, the second terminal device 102, and the third terminal device 103, such as shopping applications, web browser applications, search applications, instant messaging tools, email clients, and / or social platform software, etc. (only for example).

[0072] The first terminal device 101, the second terminal device 102, and the third terminal device 103 may be various electronic devices having display screens and supporting web browsing, including but not limited to smart phones, tablet computers, laptop computers, desktop computers, and the like.

[0073] The server 105 may be a server that provides various services, such as a background management server (only as an example) that provides support for websites browsed by users using the first terminal device 101, the second terminal device 102, and the third terminal device 103. The background management server may analyze and process the received data such as user requests, and feed back the processing results (such as web pages, information, or data obtained or generated according to user requests) to the terminal device.

[0074] The train 106 may be any train that needs to obtain electric energy from a high-voltage power grid, such as a high-speed train, a motor train, a subway, etc.

[0075] It should be noted that the state monitoring method of the train roof device provided in the embodiment of the present application can generally be executed by the server 105. Accordingly, the state monitoring device of the train roof device provided in the embodiment of the present application can generally be set in the server 105. The state monitoring method of the train roof device provided in the embodiment of the present application can also be executed by a server or server cluster that is different from the server 105 and can communicate with the first terminal device 101, the second terminal device 102, the third terminal device 103 and / or the server 105. Correspondingly, the state monitoring device of the train roof device provided in the embodiment of the present application can also be set in a server or server cluster that is different from the server 105 and can communicate with the first terminal device 101, the second terminal device 102, the third terminal device 103 and / or the server 105. Alternatively, the state monitoring method of the train roof device provided in the embodiment of the present application can also be performed by the first terminal device 101, the second terminal device 102 or the third terminal device 103, or can also be performed by other terminal devices different from the first terminal device 101, the second terminal device 102 or the third terminal device 103. Correspondingly, the status monitoring device of the train roof device provided in the embodiment of the present application can also be set in the first terminal device 101, the second terminal device 102 or the third terminal device 103, or in other terminal devices different from the first terminal device 101, the second terminal device 102 or the third terminal device 103.

[0076] It should be understood that Figure 1 The number of terminal devices, networks and servers in the embodiment is only . According to the implementation requirements, there can be any number of terminal devices, networks and servers.

[0077] Figure 2 A flow chart of a method for monitoring the status of a train roof device according to an embodiment of the present application is shown.

[0078] like Figure 2 As shown, the method for monitoring the status of a train roof component includes operations S201 to S203.

[0079] In operation S201, in response to a state monitoring instruction, a detection data set of a train roof is collected using a target thermal imager, wherein the detection data set includes a thermal imaging sequence including a plurality of roof components, and the thermal imaging sequence includes an initial infrared image in a time sequence;

[0080] In operation S202, for each initial infrared image, the initial infrared image is input into a target detection model, and target infrared images corresponding to different roof components are output, wherein the target detection model is trained using training channel images collected by a target thermal imager;

[0081] In operation S203, for each roof component, abnormal temperature outlier detection processing is performed on multiple target infrared images to obtain a first state monitoring result, wherein the first state monitoring result indicates whether the roof component is in a temperature abnormality state.

[0082] According to an embodiment of the present application, the status monitoring instruction may be a corresponding monitoring operation input by a staff in a train main control room or a train dispatching room on an electronic device such as a computer, and the electronic device may automatically generate the instruction in response to the operation.

[0083] According to an embodiment of the present application, the target thermal imager may be an online thermal imager, which may synchronously capture thermal imaging video and RGB channel video.

[0084] According to an embodiment of the present application, during the travel of the train, a thermal imaging sequence at the roof of the train is collected in real time by a target thermal imager, and each frame of the initial infrared image in the thermal imaging sequence is input into a trained target detection model to obtain a target infrared image corresponding to each roof device, wherein the roof device may be a high-voltage device such as a pantograph, an insulator, and a suspension device.

[0085] According to an embodiment of the present application, for each type of roof device, multiple consecutive target infrared images of the roof device are processed for abnormal temperature outlier detection, so that it can be determined in time whether the temperature of the roof device is abnormal, that is, the first state monitoring result.

[0086] According to the embodiment of the present application, by inputting the thermal imaging sequence of the roof device collected by the target thermal imager into the trained target detection model, the initial infrared image is segmented into corresponding target infrared images according to different roof devices, and then the target infrared image sequence of the same type of roof device is subjected to abnormal temperature outlier detection processing to determine whether the roof device has temperature abnormality. By accurately dividing different roof devices and performing abnormal temperature outlier detection on the divided images, it is possible to accurately identify whether the roof device is abnormal even in a complex train background environment, thereby improving the accuracy of abnormal roof device occurrence and reducing the accident rate of the train.

[0087] Figure 3 A flow chart of a method for monitoring the status of a train roof device according to another embodiment of the present application is shown.

[0088] According to an embodiment of the present application, the detection data set further includes a channel image sequence including a variety of roof components, and the channel image sequence includes a plurality of initial channel images in a time sequence.

[0089] According to the embodiments of the present application, Figure 3 As shown, the condition monitoring method further includes:

[0090] For each initial channel image, the initial channel image is input into the target detection model, and the target channel image corresponding to different roof components is output;

[0091] For each target channel image, the target channel image is compared and analyzed with the reference channel image to generate a second state monitoring result, wherein the second state monitoring result represents whether the external form of the roof device is in an abnormal state.

[0092] According to an embodiment of the present application, the channel image sequence is acquired at the same time as the target thermal imager acquires the thermal imaging sequence.

[0093] According to an embodiment of the present application, after the target thermal imager captures the RGB channel image sequence, each frame of the initial channel image in the channel image sequence is input into the target detection model to perform segmentation of different roof device images, thereby obtaining multiple target channel images corresponding to different roof devices.

[0094] According to an embodiment of the present application, for each type of roof device, each target channel image is compared and analyzed with a reference channel image of a roof device in normal form to determine whether the roof device is broken, damaged, or fractured, thereby obtaining a second state monitoring result.

[0095] According to the embodiments of the present application, by monitoring the temperature of the roof device and using the channel image collected by the target thermal imager to perform segmentation and comparative analysis to determine whether there are morphological abnormalities in the roof device, it is helpful to ensure the normal operation of the train from different aspects of the roof device and further improve the driving safety of the train.

[0096] According to an embodiment of the present application, the target channel image is compared and analyzed with the reference channel image to generate a second state monitoring result, including:

[0097] Calculate the image similarity between the target channel image and the reference channel image;

[0098] When the image similarity is less than the similarity threshold, an appearance abnormality result is generated, wherein the appearance abnormality result indicates that the appearance of the roof component is abnormal;

[0099] When the image similarity is not less than the similarity threshold, a normal appearance result is generated, wherein the second state monitoring result includes an abnormal appearance result or a normal appearance result.

[0100] According to an embodiment of the present application, in the process of determining whether the external shape of the roof device is in an abnormal state, it is necessary to calculate the image similarity between each target channel image and the reference channel image. If the image similarity is less than the similarity threshold, it means that there is an abnormality in the external shape of the roof device. If it is greater than or equal to the similarity threshold, it means that there is no abnormality in the external shape of the roof device. A corresponding second state monitoring result is generated based on whether there is an abnormality, such as an abnormal appearance result or a normal appearance result.

[0101] It should be noted that the similarity threshold can be set according to actual conditions, for example, it can be set to 0.9.

[0102] According to an embodiment of the present application, abnormal temperature detection processing is performed on multiple target infrared images to obtain a first state monitoring result, including:

[0103] Based on the outlier detection algorithm, abnormal temperature detection processing is performed on multiple target infrared images to obtain a first state monitoring result.

[0104] According to an embodiment of the present application, an outlier detection algorithm is a technique for identifying abnormal or distinctive data points in a data set. Common outlier detection algorithms include methods based on statistics, clustering, classification, information theory, distance, and density.

[0105] According to an embodiment of the present application, corresponding temperature values ​​are identified from a target infrared image, and outlier detection is performed on multiple temperature values ​​using an outlier detection algorithm to generate a first state monitoring result of whether there are outliers (ie, temperature anomalies).

[0106] According to an embodiment of the present application, abnormal temperature outlier detection processing is performed on multiple target infrared images based on an outlier detection algorithm to obtain a first state monitoring result, including:

[0107] For each target infrared image, feature extraction processing is performed on the target infrared image to obtain a plurality of infrared temperature information, wherein each infrared temperature information corresponds to a device area in the roof device;

[0108] For each device area, outlier detection processing is performed on multiple infrared temperature information corresponding to the device area to obtain a temperature detection result corresponding to the device area;

[0109] A first state monitoring result is generated according to the temperature detection results of the plurality of device regions.

[0110] According to an embodiment of the present application, since a roof device has many structures, in order to determine which structure of the roof device has a problem, feature extraction may be performed on the target infrared image of the roof device to determine the infrared temperature information of each device area in the roof device.

[0111] According to an embodiment of the present application, for each device area, outlier anomaly detection is performed on multiple infrared temperature information of the device area to generate a temperature detection result of whether the device area has abnormal temperature, and the temperature detection results of multiple device areas of the roof device are combined to obtain a first state monitoring result of whether the roof device has temperature abnormality.

[0112] According to an embodiment of the present application, by refining the temperature of the roof device to different device components and performing outlier detection on the device area temperature of each device component, it is possible to more accurately determine which device component of the roof device has an abnormality, thereby improving the safety of train travel.

[0113] According to an embodiment of the present application, before performing feature extraction processing, it also includes:

[0114] The target infrared image is subjected to Gaussian filtering to obtain a filtered target infrared image, and feature extraction is performed on the filtered target infrared image.

[0115] According to an embodiment of the present application, since there are other interfering noises in the captured infrared image, if the infrared image is directly used for outlier anomaly detection, it may affect the detection accuracy of temperature anomalies. Therefore, Gaussian filtering can be performed on it to reduce the noise in the imaging data, and then the filtered target infrared image can be used for feature extraction processing.

[0116] According to an embodiment of the present application, outlier detection processing is performed on multiple infrared temperature information corresponding to the device area to obtain a temperature detection result corresponding to the device area, including:

[0117] When any infrared temperature information is greater than a preset temperature threshold, a temperature anomaly result is generated, wherein the temperature anomaly result indicates that a temperature anomaly exists in the device area;

[0118] When any infrared temperature information is not greater than a preset temperature threshold, a normal temperature result is generated, wherein the temperature detection result includes an abnormal temperature result or a normal temperature result.

[0119] According to the embodiment of the present application, for the device area corresponding to each device component in the roof device, if the infrared temperature information of the device area exceeds the preset temperature threshold, it indicates that the device component is abnormal, thereby generating a corresponding temperature abnormality result. If the infrared temperature information is always not greater than the preset temperature threshold, it indicates that the device component corresponding to the device area is working normally, so a temperature normal result can be generated.

[0120] According to the embodiments of the present application, the preset temperature thresholds of different device regions need to be set by the staff according to the actual operating temperature of the device region.

[0121] According to an embodiment of the present application, the temperature detection result includes an abnormal temperature result or a normal temperature result.

[0122] According to an embodiment of the present application, a first state monitoring result is generated according to temperature detection results of multiple device regions, including:

[0123] In the case where any temperature detection result is a temperature abnormality result, a state abnormality result is generated, wherein the state abnormality result indicates that the temperature of the roof component is abnormal;

[0124] When the plurality of temperature detection results are all normal temperature results, a normal state result is generated, wherein the first state monitoring result includes an abnormal state result or a normal state result.

[0125] According to the embodiments of the present application, since the roof device is composed of different device components, when a problem occurs in any device component, it will affect the operation of the entire roof device. Therefore, as long as the temperature of one device area is abnormal, it is necessary to generate a first state monitoring result characterizing that the temperature of the roof device is abnormal.

[0126] According to an embodiment of the present application, only when all temperature detection results are normal temperature results, it is indicated that the operation of the roof device is in a normal state, thereby generating a first state monitoring result indicating that there is no abnormality in the temperature of the roof device.

[0127] According to an embodiment of the present application, the state monitoring method further includes:

[0128] For each temperature detection result, when the temperature detection result is an abnormal temperature result, a device area corresponding to the temperature detection result in the target infrared image is marked to obtain a marked infrared image;

[0129] A target monitoring set is displayed, wherein the target monitoring set includes at least one of the following: a first state monitoring result and a marked infrared image.

[0130] According to an embodiment of the present application, in order to more clearly inform the staff which roof component has a malfunction, the device component with abnormal temperature can be marked in the target infrared image with temperature abnormality, that is, the infrared image is marked, and the marked infrared image and the corresponding first state result are displayed to the relevant staff, so that the staff can promptly conduct a safety inspection on the problematic roof component, thereby further improving the safety of train operation.

[0131] According to an embodiment of the present application, the target detection model is trained in the following manner:

[0132] Acquire a training data set, wherein the training data set includes a plurality of training channel images acquired by a target thermal imager and a label image corresponding to a different roof component in each training channel image, and the label image represents a component segmentation range of the roof component;

[0133] For each training channel image, the training channel image is input into the initial visual model, and the predicted segmented images of different roof components are output;

[0134] For each roof device, a target loss value is calculated according to the predicted segmentation image and the label image, so as to iteratively adjust the model parameters of the initial visual model according to the target loss value to obtain the target detection model.

[0135] According to an embodiment of the present application, the initial visual model may be a visual model of the YOLO series or another type of target segmentation model.

[0136] According to an embodiment of the present application, the target thermal imager used for training and use can be the same, that is, "one model (target detection model) for one vehicle (train)" as much as possible. Of course, if trains of the same type can use the same model.

[0137] According to an embodiment of the present application, the target loss value may be calculated using mean square error loss (MSE), cross-entropy loss, Dice loss, contrastive loss, cosine similarity loss, and the like.

[0138] According to an embodiment of the present application, in the process of training a target detection model, a target thermal imager is used to collect training channel images of roof devices, and different roof devices in the training channel images are ranged to obtain label images corresponding to different roof devices.

[0139] According to an embodiment of the present application, each training channel image is input into the initial visual model, and predicted segmentation images corresponding to different roof devices are output. For each roof device, the target loss value is calculated based on its predicted segmentation image and label image, and the model parameters of the initial visual model are iteratively adjusted based on the target loss value to obtain a target detection model.

[0140] Figure 4 A block diagram of a train roof device status monitoring device according to an embodiment of the present application is shown.

[0141] like Figure 4 As shown, the train roof device status monitoring device 400 includes a data acquisition module 410 , a data processing module 420 , and a status monitoring module 430 .

[0142] A data acquisition module 410 is used to collect a detection data set of the train roof using a target thermal imager in response to a state monitoring instruction, wherein the detection data set includes a thermal imaging sequence including a plurality of roof components, and the thermal imaging sequence includes an initial infrared image in a time sequence;

[0143] The data processing module 420 is used to input the initial infrared image into the target detection model for each initial infrared image, and output the target infrared image corresponding to the different roof components, wherein the target detection model is trained using the training channel image collected by the target thermal imager;

[0144] The state monitoring module 430 is used to perform abnormal temperature outlier detection processing on multiple target infrared images for each roof component to obtain a first state monitoring result, wherein the first state monitoring result indicates whether the roof component is in an abnormal temperature state.

[0145] According to the embodiment of the present application, by inputting the thermal imaging sequence of the roof device collected by the target thermal imager into the trained target detection model, the initial infrared image is segmented into corresponding target infrared images according to different roof devices, and then the target infrared image sequence of the same type of roof device is subjected to abnormal temperature outlier detection processing to determine whether the roof device has temperature abnormality. By accurately dividing different roof devices and performing abnormal temperature outlier detection on the divided images, it is possible to accurately identify whether the roof device is abnormal even in a complex train background environment, thereby improving the accuracy of abnormal roof device occurrence and reducing the accident rate of the train.

[0146] According to an embodiment of the present application, the detection data set further includes a channel image sequence including a variety of roof components, and the channel image sequence includes a plurality of initial channel images in a time sequence.

[0147] According to an embodiment of the present application, the state monitoring device 400 further includes:

[0148] A second data processing module is used for inputting the initial channel image into the target detection model for each initial channel image, and outputting target channel images corresponding to different roof components;

[0149] The second state monitoring module is used to compare and analyze each target channel image with the reference channel image to generate a second state monitoring result, wherein the second state monitoring result represents whether the external shape of the roof device is in an abnormal state.

[0150] According to an embodiment of the present application, the second status monitoring module includes:

[0151] A calculation unit, used for calculating the image similarity between the target channel image and the reference channel image;

[0152] A first generating unit is used to generate an appearance abnormality result when the image similarity is less than a similarity threshold, wherein the appearance abnormality result indicates that the appearance of the roof component is abnormal;

[0153] The second generating unit is used to generate a normal appearance result when the image similarity is not less than a similarity threshold, wherein the second state monitoring result includes an abnormal appearance result or a normal appearance result.

[0154] According to an embodiment of the present application, the status monitoring module 430 includes:

[0155] The outlier detection submodule is used to perform abnormal temperature detection processing on multiple target infrared images based on the outlier detection algorithm to obtain a first state monitoring result.

[0156] According to an embodiment of the present application, the outlier detection submodule includes:

[0157] A first obtaining unit is used to perform feature extraction processing on each target infrared image to obtain a plurality of infrared temperature information, wherein each infrared temperature information corresponds to a device area in the roof device;

[0158] A second obtaining unit is used to perform outlier detection processing on multiple infrared temperature information corresponding to each device area to obtain a temperature detection result corresponding to the device area;

[0159] The third generating unit is used to generate a first state monitoring result according to the temperature detection results of the plurality of device regions.

[0160] According to an embodiment of the present application, the outlier detection submodule further includes:

[0161] The filtering unit is used to perform Gaussian filtering on the target infrared image to obtain the filtered target infrared image, so as to perform feature extraction on the filtered target infrared image.

[0162] According to an embodiment of the present application, the second obtaining unit includes:

[0163] A first generating subunit is used to generate a temperature abnormality result when any infrared temperature information is greater than a preset temperature threshold, wherein the temperature abnormality result indicates that a temperature abnormality exists in a device area;

[0164] The second generating subunit is used to generate a normal temperature result when any infrared temperature information is not greater than a preset temperature threshold, wherein the temperature detection result includes an abnormal temperature result or a normal temperature result.

[0165] According to an embodiment of the present application, the temperature detection result includes an abnormal temperature result or a normal temperature result.

[0166] According to an embodiment of the present application, the third generating unit includes:

[0167] A third generating subunit is used to generate a state abnormality result when any temperature detection result is a temperature abnormality result, wherein the state abnormality result indicates that the temperature of the roof component is abnormal;

[0168] The fourth generating subunit is used to generate a normal state result when the multiple temperature detection results are all normal temperature results, wherein the first state monitoring result includes an abnormal state result or a normal state result.

[0169] According to an embodiment of the present application, the state monitoring device 400 further includes:

[0170] The module is used to mark the device area corresponding to the temperature detection result in the target infrared image for each temperature detection result, when the temperature detection result is an abnormal temperature result, to obtain a marked infrared image;

[0171] The display module is used to display the target monitoring set, wherein the target monitoring set includes at least one of the following: a first state monitoring result and a marked infrared image.

[0172] Figure 5 A block diagram of a training device for a target detection model according to an embodiment of the present application is shown.

[0173] like Figure 5 As shown, the training device 500 of the target detection model includes an acquisition module 510, a segmentation module 520, and a training module 530.

[0174] An acquisition module 510 is used to acquire a training data set, wherein the training data set includes a plurality of training channel images acquired by a target thermal imager and a label image corresponding to a different roof component in each training channel image, and the label image represents a device segmentation range of the roof component;

[0175] A segmentation module 520 is used for inputting the training channel image into the initial visual model for each training channel image, and outputting predicted segmented images of different roof components;

[0176] The training module 530 is used to calculate the target loss value for each roof device according to the predicted segmentation image and the label image, so as to iteratively adjust the model parameters of the initial visual model according to the target loss value to obtain the target detection model.

[0177] According to the embodiments of the present application, any one or more of the modules, submodules, units, and subunits, or at least part of the functions of any one of them can be implemented in one module. According to the embodiments of the present application, any one or more of the modules, submodules, units, and subunits can be split into multiple modules for implementation. According to the embodiments of the present application, any one or more of the modules, submodules, units, and subunits can be at least partially implemented as hardware circuits, such as field programmable gate arrays (FPGAs), programmable logic arrays (PLAs), systems on chips, systems on substrates, systems on packages, application specific integrated circuits (ASICs), or can be implemented by hardware or firmware in any other reasonable way of integrating or packaging the circuit, or in any one of the three implementation methods of software, hardware, and firmware, or in any appropriate combination of any of them. Alternatively, according to the embodiments of the present application, one or more of the modules, submodules, units, and subunits can be at least partially implemented as computer program modules, and when the computer program modules are run, the corresponding functions can be executed.

[0178] For example, any of the data acquisition module 410, the data processing module 420, the state monitoring module 430, or the acquisition module 510, the segmentation module 520, and the training module 530 can be combined in one module / unit / subunit for implementation, or any of the modules / units / subunits can be split into multiple modules / units / subunits. Alternatively, at least part of the functions of one or more of these modules / units / subunits can be combined with at least part of the functions of other modules / units / subunits and implemented in one module / unit / subunit. According to an embodiment of the present application, at least one of the data acquisition module 410, the data processing module 420, the state monitoring module 430, or the acquisition module 510, the segmentation module 520, and the training module 530 can be at least partially implemented as a hardware circuit, such as a field programmable gate array (FPGA), a programmable logic array (PLA), a system on a chip, a system on a substrate, a system on a package, an application-specific integrated circuit (ASIC), or can be implemented by hardware or firmware such as any other reasonable way of integrating or packaging the circuit, or by any one of the three implementation methods of software, hardware, and firmware, or by an appropriate combination of any of them. Alternatively, at least one of the data acquisition module 410, the data processing module 420, the state monitoring module 430, or the acquisition module 510, the segmentation module 520, and the training module 530 can be at least partially implemented as a computer program module, and when the computer program module is run, the corresponding function can be performed.

[0179] It should be noted that the state monitoring device part of the train roof device in the embodiment of the present application corresponds to the state monitoring method part of the train roof device in the embodiment of the present application. The description of the state monitoring device part of the train roof device specifically refers to the state monitoring method part of the train roof device, which will not be repeated here.

[0180] Figure 6 A schematic diagram of a train according to an embodiment of the present application is shown.

[0181] like Figure 6 As shown, the train 600 includes a train body 610 and a condition monitoring device 400 for a train roof component.

[0182] According to the embodiment of the present application, by inputting the thermal imaging sequence of the roof device collected by the target thermal imager into the trained target detection model, the initial infrared image is segmented into corresponding target infrared images according to different roof devices, and then the target infrared image sequence of the same type of roof device is subjected to abnormal temperature outlier detection processing to determine whether the roof device has temperature abnormality. By accurately dividing different roof devices and performing abnormal temperature outlier detection on the divided images, it is possible to accurately identify whether the roof device is abnormal even in a complex train background environment, thereby improving the accuracy of abnormal roof device occurrence and reducing the accident rate of the train.

[0183] Figure 7 A block diagram of an electronic device suitable for implementing the method described above according to an embodiment of the present application is shown. Figure 7 The electronic device shown is merely an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

[0184] like Figure 7 As shown, the electronic device 700 according to an embodiment of the present application includes a processor 701, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 702 or a program loaded from a storage part 708 to a random access memory (RAM) 703. The processor 701 may include, for example, a general-purpose microprocessor (such as a CPU), an instruction set processor and / or a related chipset and / or a special-purpose microprocessor (for example, an application-specific integrated circuit (ASIC)), etc. The processor 701 may also include an onboard memory for caching purposes. The processor 701 may include a single processing unit or multiple processing units for performing different actions of the method flow according to an embodiment of the present application.

[0185] In RAM 703, various programs and data required for the operation of electronic device 700 are stored. Processor 701, ROM 702 and RAM 703 are connected to each other via bus 704. Processor 701 performs various operations of the method flow according to the embodiment of the present application by executing the program in ROM 702 and / or RAM 703. It should be noted that the program can also be stored in one or more memories other than ROM 702 and RAM 703. Processor 701 can also perform various operations of the method flow according to the embodiment of the present application by executing the program stored in the one or more memories.

[0186] According to an embodiment of the present application, the electronic device 700 may further include an input / output (I / O) interface 705, which is also connected to the bus 704. The electronic device 700 may further include one or more of the following components connected to the input / output (I / O) interface 705: an input portion 706 including a keyboard, a mouse, etc.; an output portion 707 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage portion 708 including a hard disk, etc.; and a communication portion 709 including a network interface card such as a LAN card, a modem, etc. The communication portion 709 performs communication processing via a network such as the Internet. A drive 710 is also connected to the input / output (I / O) interface 705 as needed. A removable medium 711, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 710 as needed, so that a computer program read therefrom is installed into the storage portion 708 as needed.

[0187] According to an embodiment of the present application, the method flow according to the embodiment of the present application can be implemented as a computer software program. For example, an embodiment of the present application includes a computer program product, which includes a computer program carried on a computer-readable storage medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through the communication part 709, and / or installed from the removable medium 711. When the computer program is executed by the processor 701, the above-mentioned functions defined in the system of the embodiment of the present application are executed. According to an embodiment of the present application, the system, equipment, device, module, unit, etc. described above can be implemented by a computer program module.

[0188] The present application also provides a computer-readable storage medium, which may be included in the device / apparatus / system described in the above embodiments; or may exist independently without being assembled into the device / apparatus / system. The above computer-readable storage medium carries one or more programs, and when the above one or more programs are executed, the method according to the embodiment of the present application is implemented.

[0189] According to an embodiment of the present application, the computer-readable storage medium may be a non-volatile computer-readable storage medium. For example, it may include, but is not limited to: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In the present application, a computer-readable storage medium may be any tangible medium containing or storing a program that may be used by or in combination with an instruction execution system, apparatus, or device.

[0190] For example, according to an embodiment of the present application, the computer-readable storage medium may include the ROM 702 and / or the RAM 703 described above and / or one or more memories other than the ROM 702 and the RAM 703 .

[0191] An embodiment of the present application also includes a computer program product, which includes a computer program, which contains program code for executing the method provided by the embodiment of the present application. When the computer program product runs on an electronic device, the program code is used to enable the electronic device to implement the method provided by the embodiment of the present application.

[0192] When the computer program is executed by the processor 701, the above functions defined in the system / device of the embodiment of the present application are executed. According to the embodiment of the present application, the system, device, module, unit, etc. described above can be implemented by a computer program module.

[0193] In one embodiment, the computer program may rely on tangible storage media such as optical storage devices, magnetic storage devices, etc. In another embodiment, the computer program may also be transmitted and distributed in the form of signals on a network medium, and downloaded and installed through the communication part 709, and / or installed from the removable medium 711. The program code contained in the computer program may be transmitted using any appropriate network medium, including but not limited to: wireless, wired, etc., or any suitable combination of the above.

[0194] According to an embodiment of the present application, the program code for executing the computer program provided by the embodiment of the present application can be written in any combination of one or more programming languages, specifically, these computing programs can be implemented using high-level process and / or object-oriented programming languages, and / or assembly / machine languages. Programming languages ​​include, but are not limited to, such as Java, C++, python, "C" language or similar programming languages. The program code can be executed entirely on the user computing device, partially on the user device, partially on the remote computing device, or completely on the remote computing device or server. In the case of a remote computing device, the remote computing device can be connected to the user computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computing device (e.g., using an Internet service provider to connect through the Internet).

[0195] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present application. In this regard, each box in the flowchart or block diagram may represent a module, a program segment, or a part of a code, and the above-mentioned module, program segment, or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box may also occur in a different order from the order marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flowchart, and the combination of the boxes in the block diagram or flowchart, can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions. It will be appreciated by those skilled in the art that the features recorded in the various embodiments of the present application can be combined and / or combined in a variety of ways, even if such a combination or combination is not clearly recorded in the present application. In particular, without departing from the spirit and teachings of the present application, the features described in the various embodiments of the present application may be combined and / or combined in a variety of ways. All of these combinations and / or combinations fall within the scope of the present application.

[0196] The embodiments of the present application are described above. However, these embodiments are only for the purpose of illustration, and are not intended to limit the scope of the present application. Although each embodiment is described above, this does not mean that the measures in each embodiment cannot be used in combination advantageously. The present application does not depart from the scope of the present application, and those skilled in the art may make various substitutions and modifications, which should all fall within the scope of the present application.

Claims

1. A method for monitoring the status of a train roof component, characterized in that: include: In response to the status monitoring instruction, using a target thermal imager to collect a detection data set of the train roof, wherein the detection data set includes a thermal imaging sequence including a plurality of roof components, and the thermal imaging sequence includes an initial infrared image in a time sequence; For each of the initial infrared images, the initial infrared image is input into a target detection model, and target infrared images corresponding to different roof components are output, wherein the target detection model is trained using training channel images collected by the target thermal imager; For each of the roof components, abnormal temperature outlier detection processing is performed on a plurality of the target infrared images to obtain a first state monitoring result, wherein the first state monitoring result indicates whether the roof component is in a state of abnormal temperature.

2. The method according to claim 1, characterized in that The detection data set also includes a channel image sequence containing a variety of roof components, and the channel image sequence includes a plurality of initial channel images in a time sequence; Wherein, the state monitoring method further comprises: For each of the initial channel images, input the initial channel image into the target detection model, and output a target channel image corresponding to different roof components; For each of the target channel images, the target channel image is compared and analyzed with the reference channel image to generate a second state monitoring result, wherein the second state monitoring result represents whether the external form of the roof device is in an abnormal state.

3. The method according to claim 2, characterized in that Comparing and analyzing the target channel image with the reference channel image to generate a second state monitoring result includes: Calculating the image similarity between the target channel image and the reference channel image; When the image similarity is less than a similarity threshold, generating an appearance abnormality result, wherein the appearance abnormality result indicates that an appearance of the roof component is abnormal; In a case where the image similarity is not less than a similarity threshold, a normal appearance result is generated, wherein the second state monitoring result includes the abnormal appearance result or the normal appearance result.

4. The method according to any one of claims 1 to 3, characterized in that Performing abnormal temperature detection processing on the plurality of target infrared images to obtain a first state monitoring result includes: Abnormal temperature detection processing is performed on the plurality of target infrared images based on an outlier detection algorithm to obtain the first state monitoring result.

5. The method according to claim 4, characterized in that Performing abnormal temperature outlier detection processing on the plurality of target infrared images based on an outlier detection algorithm to obtain the first state monitoring result includes: For each of the target infrared images, feature extraction processing is performed on the target infrared image to obtain a plurality of infrared temperature information, wherein each of the infrared temperature information corresponds to a device area in the roof device; For each of the device regions, performing outlier detection processing on a plurality of infrared temperature information corresponding to the device region to obtain a temperature detection result corresponding to the device region; The first state monitoring result is generated according to the temperature detection results of the plurality of device regions.

6. The method according to claim 5, characterized in that Before feature extraction, it also includes: The target infrared image is subjected to Gaussian filtering to obtain a filtered target infrared image, and feature extraction is performed on the filtered target infrared image.

7. The method according to claim 5, characterized in that Performing outlier detection processing on the plurality of infrared temperature information corresponding to the device area to obtain a temperature detection result corresponding to the device area includes: In the case where any of the infrared temperature information is greater than a preset temperature threshold, a temperature anomaly result is generated, wherein the temperature anomaly result indicates that a temperature anomaly exists in the device area; In the case that any of the infrared temperature information is not greater than a preset temperature threshold, a normal temperature result is generated, wherein the temperature detection result includes the abnormal temperature result or the normal temperature result.

8. The method according to claim 5 or 7, characterized in that: The temperature detection result includes an abnormal temperature result or a normal temperature result; The first state monitoring result is generated according to the temperature detection results of the plurality of device regions, including: In the case where any of the temperature detection results is a temperature abnormality result, generating a state abnormality result, wherein the state abnormality result indicates that the temperature of the roof component is abnormal; In the case that the plurality of temperature detection results are all normal temperature results, a normal state result is generated, wherein the first state monitoring result includes the abnormal state result or the normal state result.

9. The method according to claim 8, characterized in that Also includes: For each of the temperature detection results, if the temperature detection result is an abnormal temperature result, marking a device area corresponding to the temperature detection result in the target infrared image to obtain a marked infrared image; A target monitoring set is displayed, wherein the target monitoring set includes at least one of the following: the first state monitoring result and the marked infrared image.

10. The method according to claim 1, characterized in that The target detection model is trained as follows: Acquire a training data set, wherein the training data set includes a plurality of the training channel images acquired by the target thermal imager and a label image corresponding to a different roof component in each of the training channel images, wherein the label image represents a component segmentation range of the roof component; For each of the training channel images, input the training channel image into an initial visual model, and output predicted segmented images of different roof components; For each of the roof devices, a target loss value is calculated according to the predicted segmented image and the label image, so as to iteratively adjust the model parameters of the initial visual model according to the target loss value to obtain the target detection model.

11. A train roof device status monitoring device, characterized in that: include: A data acquisition module, configured to collect a detection data set of a train roof using a target thermal imager in response to a state monitoring instruction, wherein the detection data set includes a thermal imaging sequence including a plurality of roof components, and the thermal imaging sequence includes an initial infrared image in a time sequence; A data processing module, for each of the initial infrared images, inputting the initial infrared image into a target detection model, and outputting target infrared images corresponding to different roof components, wherein the target detection model is trained using training channel images acquired by the target thermal imager; The state monitoring module is used to perform abnormal temperature outlier detection processing on the multiple target infrared images for each of the roof components to obtain a first state monitoring result, wherein the first state monitoring result indicates whether the roof component is in an abnormal temperature state.

12. A train, characterized in that: include: The train body, The condition monitoring device for a train roof device as claimed in claim 11.

Images