Thermography measurement method and thermography measurement system for power distribution assets and power transmission assets

Abstract

Measuring the thermal emission of a power distribution asset and / or a power transmission asset 11 such as an overhead line is disclosed. The method comprises capturing a first image 1 of the asset usi

Description

The present invention refers to a method of measuring the thermal emission of an asset being a power distribution asset and / or power transmission asset, wherein the method utilizes multiple imaging devices to gain an improved insight in the asset. Furthermore, the present invention refers to a computer program product to execute such method. Additionally, the present invention refers to a thermography measurement system being adapted to realize such method. Power distribution assets and power transmission assets represent critical infrastructure and securing their operation on long-term is mandatory. Respective inspection of power distribution assets and power transmission assets requires the assessment of the current state of respective assets, which are critical for the reliable and safe generation and distribution of electrical power. An example of such asset are overhead lines containing overhead power lines and accompanying pylons that need to be inspected on a regular basis to prevent damages to occur that threaten the functionality or require longer downtimes. Traditional methods of monitoring power distribution and transmission assets often involve manual inspections or the use of specialized equipment to detect defects and assess the condition of the assets. However, these methods can be timeconsuming, are limited to available resources of respective equipment, and may not provide comprehensive data on the condition of the assets. For example, thermal imaging utilizing infrared radiation is utilized to monitor the thermal emission of respective assets. Taking into account the heat originating from the function of the respective power distribution assets and power transmission assets utilizing thermal emission data of such assets allows to gain a valuable insight into the operation of such assets and their current state. In this context, thermal imaging has been widely used for detecting anomalies in electrical systems and power distribution assets. It allows for the visualization of temperature variations, which can indicate potential issues such as overheating, loose connections, or insulation degradation. However, traditional thermal imaging has many limitations. While IR cameras are commonly available their function is limited to a great extent. For example, the resolution of available thermal imaging devices is very limited compared to other imaging devices . Increasing the resolution of a thermal infrared camera system is not easily possible. Thermal infrared sensors can't be made from silicone and sensor manufacturing technology is much more costly compared to regular imaging devices, while delivering much lower spatial resolution sensor. Assets should typically be monitored from a distance for practical and safety reasons, which requires longer focal length lens systems to provide a working imaging device. The comparatively large pixels, the use of specialized non-glass lenses, and the required coverage limits the available options. All combined with the requirement of mobile and airborne applications significantly limits the practical possibility to further increase the resolution for a thermal imaging device for monitoring respective assets. Therefore, there is a need to improve the available possibilities to monitor power distribution assets and power transmission assets to effectively identify respective problems and address respective needs by alternative means as soon as possible. Requiring improved inspection tools and improved inspection methods. This and further problems are solved by the products and methods as disclosed hereafter and in the claims. Further beneficial embodiments are disclosed in the dependent claims and the further description and figures. These benefits can be used to adapt the corresponding solution to specific needs or to solve additional problems. According to one aspect the present invention refers to a method of measuring the thermal emission of an asset being a power distribution asset and / or power transmission asset, the method containing capturing a first image of the asset utilizing a first imaging device, wherein the first imaging device is a thermal imaging device, capturing a second image of the asset utilizing a second imaging device, wherein the second imaging device is not a thermal imaging device, wherein the second image has a higher spatial resolution than the first image, processing the second image to identify the asset contained in the second image, processing the first image utilizing a computation module containing a processing unit, wherein data regarding the asset in the second image is utilized to determine a thermal emission of the asset in the first image. The method disclosed offers several significant benefits in the measurement of thermal emission for power distribution and transmission assets. By capturing a first image of the asset using a thermal imaging device and a second image using a non-thermal imaging device with higher spatial resolution, the method enables to make use of the higher spatial resolution for the evaluation of the asset. This improved spatial resolution allows for more detailed and precise identification and analysis of specific components and areas of the as set, leading to a more comprehensive understanding of its thermal emissions and condition. It is noted that using information about the respective assets and processing power to make use of the data of the second image allows to refine the data and further analyze the data of the first image. Overall providing significantly more insight than simply acquiring a higher resolution thermal imaging device. The respective analysis steps and processing of the data in this context is easily compensated and can be far simpler than expected. Allowing to enable a broad application of such solution. Furthermore, the combination of thermal imaging with higher-resolution non-thermal imaging facilitates comprehensive data fusion. By processing the second image to identify the asset and utilizing this data to determine the thermal emission of the asset in the first image, the method enables a holistic approach to asset assessment, incorporating both thermal and visual information for a more complete analysis. It was noted that the mere increase of resolution that can be provided herewith was only a part of the benefit. Making use of the additional data allows to combine different types of data directly with the thermal data. Like making use of a RGB camera noting some contamination of the asset like bird dropping resulting in a change of the thermal emission of the respective location of the asset. Being able to be immediately taken into account and detected by the inventive method. Surprisingly, such improved method allows to significantly improve the automatic determination of defects and the overall current state of the respective asset. Investing the processing power for such improvement of the first image resulted in a tremendously decreased processing power or required resources like drone run time required to reliable determine the state of the asset by utilize multiple fly overs to acquire the data. Thus, the additional effort is at least partially or even overcompensated by the processing power saved for an automatic evaluation or the time saved by an expert reviewing respective images. This capability supports proac tive maintenance and early detection of potential issues, contributing to improved asset monitoring and reliability. Moreover, the method’s utilization of data from the second image to support the measurement of the thermal emission of the asset in the first image enhances the accuracy of thermal emission measurements. This approach leverages the higher spatial resolution of the second image to refine the assessment of thermal emissions in the first image, leading to more precise and reliable results. In summary, the described invention offers benefits such as enhanced spatial resolution, comprehensive data fusion, accurate thermal emission determination, and advanced defect detection, thereby providing a more effective and detailed method for measuring the thermal emission of power distribution and transmission assets using commonly available means allowing its application on a broad scale. Compared to a slower close-up flight path or improved thermal imaging hardware, a lower amount of resources like fly time of a respective drone is achieved. Resulting in a significantly increased applicability of such solution to address the current and development needs of modern and future power grids. According to a further aspect the presented invention refers to a computer program product, tangibly embodied in a machine-readable storage medium, including instructions opera ble to cause a computing entity to execute an inventive method. The tangible embodiment of a computer program product in a machine-readable storage medium provides a convenient and efficient way to implement the monitoring method. It allows for easy installation and execution of the method on a computing entity. According to a further aspect the presented invention refers to a storage device for providing an inventive computer program product, wherein the device stores the computer program product and / or provides the computer program product for further use. According to a further aspect the present invention refers to a thermography measurement system for measuring the thermal emission of an asset being an power distribution asset and / or power transmission asset comprising: a first imaging device being a thermal imaging device for capturing thermal radiation from the asset; a second imaging device providing a higher resolution than the first imaging device, wherein the second imaging device is no thermal imaging device; a computation module containing a processing unit, wherein the computation module is adapted to execute an inventive method, a data storage containing a computer program product including instructions operable to cause the computation module to execute an inventive method. To simplify understanding of the presented invention it is referred to the detailed description hereafter and the figures attached as well as their description. Herein, the figures are to be understood being not limiting the scope of the present invention as they merely disclosing preferred embodiments explaining the invention further. Fig. 1 shows a scheme of a cutout of a first image. Fig. 2 shows a scheme of a cutout of a second image. Fig. 3 shows a scheme of the cutout of a first image further including a measurement window and two reference windows. Preferably, the embodiments hereafter contain, unless specified otherwise, at least one processor and / or data storage unit to implement the inventive method. Unless specified otherwise terms like "calculate", "process", "determine", "generate", "configure", "reconstruct" and comparable terms refer to actions and / or processes and / or steps modifying data and / or creating data and / or converting data, wherein the data are presented as physical variables or are available as such. The term "data storage" or comparable terms as used herein, for example, refer to a temporary data storage like RAM (Random Access Memory) or long-term data storage like hard drives or data storage units like CDs, DVDs, USB sticks and the like. Such data storage can additionally include or be connected to a processing unit to allow a processing of the data stored on the data storage. According to one aspect the presented invention refers to a method as described above. According to further embodiments it is preferred that the method contains the step of identifying incomplete pixels of the first image utilizing the data regarding the asset in the second image, wherein the incomplete pixels are pixels of the first image being partially covered by the asset, wherein a factor is determined for the incomplete pixels taking into account that the incomplete pixels only partially cover the asset in the first image, wherein the factor is utilized for the evaluation of the first image. By identifying incomplete pixels in the first image, which are partially covered by the asset, and determining a factor that accounts for this partial coverage, the method enhances the accuracy of thermal emission evaluation in a very easy way. While neural networks and deep learning can be utilized such simple approach allows to easily determine a respective factor whose effect can be easily understood and optimized manually for a specific application, if required. Also, such data can be utilized for creating training data for neural networks and the like. Such approach al lows for a more precise assessment of the thermal emissions from the asset, taking into account the incomplete pixel data and improving the overall reliability of the evaluation. Besides some approach utilizing neural networks and deep learning it is also, for example, possible to determine a factor by analyzing the raw image data directly. For example, an area of the high-resolution image is segmented into asset and background regions. From camera-to-camera pixel-wise registration, the segmentation is transferred to the low-resolution cameras. It is important that the sub-pixel accuracy is retained. For each partially covered pixel in the low-resolution image, a highly accurate asset ratio can thus be calculated. In a traditional implementation, one would pick fully covered pixels in the low-resolution image and determine object emissions from there. However, there might not be fully covered pixels as the object might be too small. In addition, we assume that the low-resolution thermal image can be slightly out of focus or the registration misaligned. In these cases, the traditional method would not work anymore. The method according to the present invention, however, can, for example, utilize all pixels from within the measurement window, even those that a traditional method would discard as background pixels. Utilizing the assumption that all thermal energy coming from the asset will be present in the measurement window, albeit with an unknown spread over nearby pixels due to blur, misalignment, and too small size. Thus, it can be advantageous to determine an asset factor for the measurement window in the thermal image over all pixels within that window. Herein, the measurement window is assumed to be a mix of background and foreground. Such ratio can be determined from the high-resolution segmentation, e.g. by counting the number of pixels . According to further embodiments it is preferred that the method contains acquiring at least three first images and a second image, more preferred at least five first images and a second image, wherein processing the at least three, more preferred at least five images is based on the second image. The acquisition of multiple first images and a second image provides a robust basis for processing and analysis. If these images are acquired from a moving airborne platform, the separation of the background from the foreground becomes more robust and outlier objects in the background can be avoided. By utilizing multiple first images and analyzing them along with a respective second image, the method thus enhances the robustness and consistency of the evaluation, reducing the impact of potential variations or anomalies in individual images and ensuring a more reliable assessment of the asset’s thermal emission. Wherein even a very simple thermal imaging device provides very good results. According to further embodiments it is preferred that the method contains acquiring a first image and at least two second images, more preferred at least three second images, even more preferred at least five second images, wherein the at least two second images, more preferred at least three second images, even more preferred at least five second images are processed to determine at least one characteristic of the asset, wherein the processing of the first image takes into account the at least one characteristic. The acquisition of a first image and multiple second images allows for comprehensive asset characterization. By processing the multiple second images to determine at least one characteristic of the asset and incorporating this characteristic into the processing of the first image, the method enables a more thorough and detailed evaluation of the asset’s thermal emission, considering various aspects of its condition and performance. Such method is, for example, very beneficial to analyze data of an overhead line being partially occluded by trees or the like. Based on multiple second images the three-dimensional location of the asset can be determined and background as well as foreground objects can be determined. For example, a branch of a tree partially covering the asset in a first image can be deter mined and the respective part of the asset in the first image can be either corrected, if possible, or it can be marked as less reliable data. Also, objects in the background emitting thermal energy can be identified and respective data can be utilized to optimize the first image. Based on the speed of the inventive method it becomes possible to directly determine such problem and combine multiple first images from different angles to, for example, correct one of the first images to remove the obstruction be objects in the background and foreground. Allowing, for example, a drone to identify such problem during a single flight and directly repeat and optimize the data acquisition. According to further embodiments it is preferred that the method contains acquiring a plurality of pairs of the first image and the second image, wherein the plurality of pairs of the first image and the second image are utilized to acquire 3D data. By acquiring a plurality of pairs of the first image and the second image, the method enables the utilization of these pairs to acquire 3D data. This 3D data provides a more comprehensive and detailed representation of the asset, allowing for advanced visualization and analysis of its thermal characteristics in three dimensions. For example, determinine the geometric shape of an asset and a curved area of its surface influences the thermal emissivity and emission at said location. A homogeneous thermal emission at such location that should provide an inhomogeneous emission based on its shape can be very easily identified using such additional 3D data. The term "plurality" as used herein refers to as least five units, more preferred at least seven units, even more preferred at least ten units. According to further embodiments it is preferred that the method contains acquiring a plurality of pairs of the first image and the second image, wherein the plurality of pairs of the first image and the second image are utilized to acquire additional data to be included in at least one first image. The acquisition of a plurality of pairs of the first image and the second image allows for the utilization of these pairs to acquire additional data to be included in at least one first image. By incorporating additional data obtained from the pairs of images, the method enriches the first image with supplementary information, enhancing the depth and breadth of the thermal assessment. For example, such data can be utilized to add missing thermal data based on a foreground object in a first image. Or remove unrelated thermal data from an object located in the background, wherein the amount of thermal emission in the background is identified in a second pair to correct the thermal emission in the first image of a first pair, his enriched data can provide a more comprehensive view of the asset’s thermal emissions and condition, supporting more thorough and accurate evaluations. For example, a conductor of an overhead line is a long object and each set of images will analyze only a section of the conductors. Processing the powerline in an overlapping sliding window can further be used to detect outliers and hot spots along a single conductor and between conductors of the same circuit. According to further embodiments it is preferred that the method utilizes a drone containing the first imaging device and the second imaging device. Typically, it is preferred that such drone also contains a third imaging device. For typical applications the second imaging device is a RGB camera and the third imaging device is a LIDAR device or the second imaging device is a LIDAR device and the third imaging device is a RGB camera, more preferred the second imaging device is a RGB camera and the third imaging device is a LIDAR device. Preferably, such drone is able to fly at least 20km at a time, more preferred at least 30km, even more preferred at least 50km. The utilization of a drone containing the first imaging device, the second imaging device, and optionally a third imaging device offers several advantages. By leveraging drone technology, the method enables versatile and flexible data acquisition, allowing for efficient and comprehensive imaging of power distribution and transmission assets from various angles and perspectives. The inclusion of different imaging devices, such as RGB cameras and LIDAR devices, further enhances the data collection capabilities, providing a multi-modal approach to asset assessment that encompasses visual, thermal, and spatial data. Herein, utilizing the inventive solution requiring only low-resolution thermal imaging devices with limited size and weight allows to realize drones being able to cover significant distances, such as at least 20km, 30km, or even 50km. Enhancing its utility for large-scale asset monitoring and inspection and offering a wide operational range for data acquisition. According to further embodiments it is preferred that the computation module is located in the drone. By locating the computation module within the drone, the method enables realtime data processing and analysis directly at the point of data acquisition. This onboard computation capability offers immediate insights into thermal and visual data, facilitating prompt decision-making and action in response to the asset’s condition. While it is also possible with a remote analysis the possibility to directly decide to deviate from an original monitoring strategy to, for example, include an additional round over an asset to acquire additional data even in case the communication with a base station is lost is highly beneficial. Easily compensating for the additional effort to provide the required processing power and the like. Additionally, by processing data onboard, the method can reduce the need for extensive data transmission and enable more efficient use of resources additionally compensating for such modification and related requirements to be fulfilled. According to further embodiments it is preferred that the computation module is located remotely from the drone. In scenarios where remote computation is preferred, the method allows for the processing and analysis of acquired data to be conducted remotely from the drone. This approach offers flexibility and scalability, as it enables centralized data processing, storage, and analysis, potentially leveraging cloud computing resources. Respective application benefitting from such solution, for example, include application cases with a reliable connection being available like a 5G network or applications requiring a high number of drones to be sent out simultaneously to, for example, determine the state of an overhead line over a huge area at the same time. Reducing the resources to be spent on a single drone. Also, application cases like monitoring the state of a power generation asset indoors utilizing small drones flying through a facility benefit significantly from such. For example, monitoring a respective fleet of gas turbines in a power generation facility like a power plant, wherein the size of respective drones is extremely limited, for example, to avoid accidents. Such solution benefits from a remote computation and utilizing a large fleet of cheap drones merely acquiring data as requested from the remote control and processing unit. According to further embodiments of this preferred that the method contains the step of processing the second image to determine characteristics of the asset, wherein the characteristics tic selected from the group distance, asset size, asset contain at least one characteris consisting of asset shape, asset orientation, conductor location, asset location, age, corrosion. By extracting these characteristics from the second image, the method enables comprehensive asset characterization, providing insights into the physical attributes, condition, and environmental factors related to the asset. Respective data can be utilized to more reliable determine the current state of the asset and analyze deviations observed in the thermal data to determine, for example, problems at an early state. This information can be instrumental in asset management, maintenance planning, and performance assessment. Determining, for example, the corrosion of a respective component of the asset like the color of an isolator allows to correspondingly interpret the thermal information accordingly. For example, in case such corrosion of a respective component decreases the thermal emission at this location the thermal emission in the first image needs to be corrected accordingly and the respectively higher thermal emission may show a problem to the expert that would not be available even with the highest resolution thermal imaging device . According to further embodiments of this preferred that the step of determining the thermal emission of the asset utilizes an asset shape, wherein data regarding the asset shape is acquired as characteristic from the second image and / or from a combination of constructional data and the second image and / or a combination of the second image and data from a third imaging device. The described embodiment offers several significant benefits in the determination of thermal emission of the assets. By utilizing the asset shape as a factor in determining the thermal emission of the asset, the method enhances the accuracy and reliability of thermal emission measurements. The incorporation of asset shape data allows for a more comprehensive and nuanced understanding of the asset's thermal behavior, taking into account its physical structure and geometry. By integrating constructional data or additional data from a third imaging device, the method ensures an even more robust and multi-faceted approach to asset shape characterization. According to further embodiments of this preferred that the method contains acquiring data regarding an age and / or a corrosion of a component of the asset, wherein the age and / or the corrosion are utilized to determine a theoretical emissivity or evaluate a thermal emission of the first image. For example, such data regarding an age and / or a corrosion of a component of the asset can be acquired from a database. Like an asset database containing data regarding a time a component of the asset was replaced or service. Or like a historic database containing data regarding a corrosion state of comparable component of comparable assets at a specified time. By incorporating data regarding the age and / or corrosion of a component of the asset, the method is enabled to even further enhances the evaluation of thermal emission. This approach takes into account the impact of aging and corrosion on the asset's thermal properties, providing a more realistic and nuanced evaluation. While the required processing power to correspondingly take respective features into account is significant a corresponding effort was still found to be highly interesting for cases requiring an even more detailed analysis. Like highly relevant components being very difficult to be serviced or replaced. According to further embodiments of this preferred that the method contains the step of processing the second image or the second image and construction data to determine 3D characteristics of the asset, wherein the 3D characteristics contain at least one characteristic selected from the group consisting of asset shape, asset distance, asset size, asset orientation, conductor location, asset location, wherein the at least one 3D characteristic is utilized to identify pixels of the first image at least partially containing the asset. By utilizing the 3D characteristics to identify pixels of the first image a significant reduction of the required processing power can be achieved for many application cases as, for example, an overhead line cable providing a known shape or orientation can be fitted quite easily allowing to more easily determine the border of the respective asset in the first image in turn allowing to simplify the determination of the respective incomplete pixels. According to further embodiments it is preferred that the second imaging device is a RGB camera or a LIDAR device, preferably a RGB camera. It was noted that such imaging devices are especially beneficial for typical application cas es . According to further embodiments it is preferred that the method contains the step of determining full pixels and incomplete pixels of the first image based on the second picture providing a higher resolution than the first image, wherein the full pixels are pixels of the first image being completely covered be the asset, wherein the incomplete pixels are pixels of the first image being partially covered by the asset, wherein the method contains the step of determining a measurement window, wherein the measurement window contains the thermal emission of the asset based on the full pixels and incomplete pixels, wherein the method contains the step of determining at least one reference window, wherein the at least one reference window does not include thermal emission from the asset, wherein preferably the reference window, more preferred pixels of the reference window, are near to the measurement window. This enables a more detailed and nuanced analysis of the asset’s thermal emission, distinguishing between full pixels completely covered by the asset and incomplete pixels partially covered by the asset. Utilizing such reference windows allows to surprisingly easily include corrections based on thermal emission originating in the background of the first image. Enabling further improvements to provide a comprehensive and comparative analysis of the asset’s thermal behavior. The phrase "near to the measurement window" as used herein refers to pixels of the first image being near to the pixels of the measurement window. For example, it refers to pixels being distanced at most 50 pixels, preferably at most 30 pixels, even more preferred at most 20 pixels, distanced from the pixels relating to the measurement window. Herein, the number of pixels is based on the number of pixels that need to be crossed to reach the respective pixel, wherein the first pixel neighboring a full pixel and / or neighboring pixel being outside the measurement window represents a distance of one pixel. This proximity-based approach ensures that the reference window is strategically positioned in relation to the measurement window, facilitating accurate and contextually relevant comparisons for thermal emission analysis. According to further embodiments of this preferred that at least two reference windows are determined. Herein, such multiple reference windows can be neighboring each other or partially overlap, preferable may only being neighboring each other. Typically, it is even preferred that at least some of the reference windows like at least two of the at reference windows are distanced by at least one pixel from each other. While it is possible to artificially additionally include slightly overlapping reference windows to formally include them, it is typically beneficial to at least include some completely independent reference windows. It was noted that such embodiment allows for more accurate and comprehensive analysis of the data. Making use of multiple reference windows is surprisingly beneficial, as it was noted that irregularities originating, for example, in the background at a very specific location can be easily identified in such manner. Additionally, by ensuring that at least some of the reference windows are distanced by at least one pixel from each other, the method can capture a wider range of information and improve the accuracy of the measurements. According to further embodiments it is preferred that the at least two reference window are compared, wherein a standard deviation of the thermal emission of the at least two reference windows is determined, wherein an action is triggered in case the standard deviation of the thermal emission of the at least two reference windows exceeds a predefined threshold value. Such action can be, for example, stopping the further processing of the first image and second image or creating an output like a failure notice or advice regarding the quality of the processing. By setting a predefined threshold value for the standard deviation, the method can automatically trigger an action when the deviation exceeds this threshold. Surprisingly, it was noted that specifying a respective threshold value is quite simple and easi- ly transferable to subsequent measurements of the same thermography measurement system. Such threshold value can be, for example, be specified based on historic data or can be instantly created during, for example, a flight of a drone car-5 rying the thermography measurement system. Like comparing the standard deviations of reference windows of a minimum number of images like at least 100, wherein the distance between taking the images fulfills a specified criteria like being distanced at least 100m. Then specifying the threshold value 10 by picking the highest standard deviation of the group containing the lowest 90% of the standard deviations and adding a safety margin. Further methods are available to the skilled person and can be easily utilized or such method. This approach provides a reliable and automated way to detect anoma-15 lies or variations in the thermal emission, allowing for timely intervention or further analysis when necessary. According to further embodiments it is preferred that the method contains the step of automatically identifying defects 20 of the asset based on the thermal emission of the asset in the first image. Such identification preferably utilizes historic data or simulation data of a thermal emission of the asset. For example, the thermal emission to the expected under the respective 25 conditions like the transmitted current at the given time is taken into account to simulate the thermal emission that should be observed. Based on the observed thermal emission of the asset respective deviations can be identified. Like hot spots indicating damages leading to a higher resistance of a 30 power transmission unit resulting in an increased generation of heat at the specific location. By utilizing historic or simulation data of the asset’s thermal emission under specific conditions, the method further simplifies to accurately identify deviations and anomalies in the observed thermal 35 emission. Historic data can refer to data seen on the same infrastructure in the past, or data from other similar infrastructure located entirely somewhere else. It is further beneficial to compare data of other conductors along the same powerline. Typically, other conductors on the same powerline have experienced the same about of aging and weathering. This allows for the automated detection of issues such as hot spots, which can indicate damages or increased resistance in power transmission units, enabling proactive maintenance and intervention to prevent potential failures. According to further embodiments of this preferred that the second imaging device is a RGB camera. This, for example, allows for capturing high-resolution color images, which can provide additional visual information for analysis and complement the thermal imaging data. The combination of thermal and RGB imaging enhances the overall understanding of the asset's condition and facilitates more comprehensive inspections and assessments. Additionally, such cameras as broadly available, are well established, reliable and provide a very high resolution to be utilized for the purpose as specified herein. According to further embodiments of this preferred that the second imaging device is a LIDAR imaging device. LIDAR technology can provide highly accurate and detailed 3D spatial information about the asset, allowing for precise measurements and analysis. By incorporating LIDAR data alongside thermal imaging, the method can, for example, offer a very comprehensive understanding of the asset's physical characteristics and aid in the accurate estimation of size, orientation, and potential defects rendering it especially beneficial in case such characteristics or the like are targeted to be utilized. According to further embodiments it is preferred that the computation module estimates the size of the object and orientation using LIDAR scanning, or construction data. This allows for accurate and precise measurements, aiding in the assessment of the object's dimensions and spatial orientation. By leveraging LIDAR technology or construction data, the method can provide detailed information for further analysis and decision-making, enhancing the overall understanding of the object’s physical characteristics. For example, it is surprisingly beneficial to utilize information whether the observed asset is a line-like object, round object, or a rectangular object. Even classifying certain parts of such asset accordingly can significantly simplify the analysis of such assets . According to further embodiments it is preferred that the system is capable of detecting defects in the object based on a determined thermal emission of the asset. Utilizing the determined thermal emission makes it surprisingly easy to identify defects. For example, a lack of thermal emission can be attributed to a damage of the respective asset resulting in a missing part of it being easily identified based on a respective irregularity of the thermal emission combined with a respective obseration in, for example, a RGB image. For example, an oxidation can be identified based on a reduced or increased thermal emission at a given location based on changed thermal transmission resulting from such oxidation layer compared to unoxidized metal. Wherein such observation can even be substantiated by, for example, RBG images based on the respective color at the corresponding location. This capability allows for the automated identification of anomalies or irregularities in the object’s thermal emission, enabling proactive detection of potential defects or malfunctions. By leveraging thermal data for defect detection, the system can facilitate timely maintenance and intervention, ultimately contributing to improved asset reliability and performance. According to further embodiments it is preferred that the method contains the step of creating a thermal image, wherein the thermal image shows the thermal radiation of the asset, wherein the thermal radiation of the asset in the thermal image is corrected by correcting the respective thermal radiation being visible in the first image taking into account thermal radiation not relating to the asset based on the first image, wherein the thermal radiation being visible in the thermal image is based on the first image taking into account the determined thermal emission of the asset. By correcting the thermal radiation visible in the thermal image based on the first image and taking into account the determined thermal emission of the asset, the method ensures that the thermal image accurately represents the asset's thermal radiation while accounting for any external factors. Being surprisingly simple using the method as described herein. This allows for more precise and reliable thermal analysis, leading to improved defect detection and asset assessment. According to further embodiments it is preferred that the method utilizes a thermography measurement system containing a first imaging device being a thermal imaging device for capturing thermal radiation from an object; a second imaging device providing a higher resolution than the first imaging device for estimating the size and orientation of the object, wherein the second imaging device is no thermal imaging device; a computation module for determining measurement windows, reference windows, and computing the mean emission of the foreground object using information from the higher resolution imaging device; wherein the method contains the step of capturing a first image utilizing the first imaging device, wherein the method contains the step of capturing a second image utilizing the second imaging device, wherein the resolution of the second image is higher than the resolution of the first image, wherein the method contains the step of generating a thermal inspection output, wherein the thermal inspection output provides a resolution being higher than the resolution of the first image. This allows, for example, for more accurate and detailed analysis, leading to improved defect detection and precise assessment of the object's condition. According to a further aspect the presented invention refers to a computer program product, tangibly embodied in a machine-readable storage medium, including instructions operable to cause a computing entity to execute an inventive method. The tangible embodiment of a computer program product in a machine-readable storage medium provides a convenient and efficient way to implement the monitoring method. It allows for easy installation and execution of the method on a computing entity. According to a further aspect the presented invention refers to a storage device for providing an inventive computer program product, wherein the device stores the computer program product and / or provides the computer program product for further use. According to a further aspect the present invention refers to a thermography measurement system for measuring the thermal emission of an asset being an power distribution asset and / or power transmission asset comprising: a first imaging device being a thermal imaging device for capturing thermal radiation from the asset; a second imaging device providing a higher resolution than the first imaging device, wherein the second imaging device is no thermal imaging device; a computation module containing a processing unit, wherein the computation module is adapted to execute an inventive method, a data storage containing a computer program product including instructions operable to cause the computation module to execute an inventive method. According to further embodiments it is preferred that the thermography measurement system contains a drone, wherein the drone contains the first imaging device being a thermal imaging device, the second imaging device being no thermal imaging device, and a third imaging device, wherein the third imaging device is no thermal imaging device and wherein the third imaging device utilizes a different measurement method than the second imaging device. For example, according to typical application cases the second imaging device can be a RGB camera and the third imaging device can be a LIDAR device. The following detailed description of the figure uses the figure to discuss illustrative embodiments, which are not to be construed as restrictive, along with the features and further advantages thereof. Figure 1 shows a scheme of a cutout of a first image 1. Herein, the cutout is separated into a large number of squares representing the pixels of the image. The first image 1 is a thermal image being captured by a drone carrying a thermal imaging device being an infrared camera flying along an overhead line to determine its status. The cutout contains a part of the asset 11 and the background 12. The asset 11 in this case is a power transmission asset 11 being an overhead line. Some areas of the first image will contain pixels that fully contain the asset 11, these pixels denoted as full pixels 3. Other pixels will be partially covered by the asset 11, for example, containing background and the asset 11 denoted as incomplete pixels. Further pixels are at a further distance from the asset and exclusively show information from the background 12. However, this information cannot be known from the first image alone and information from the second image is used so that the full pixels 3 and incomplete pixels 4 are determined. Only exemplarily a small number of full pixels 3 and incomplete pixels is labelled in figure 1. Taking the position of the asset 11 in the first image and second image it becomes very easy to identify the incomplete pixels only being partially covered by the asset 11. While the thermal image alone often has an unclear border of the respective objects shown therein, making it hard to clearly identify the limits of such asset 11 in a thermal image. As the first imaging device and the second imaging device in many applica tions cases take the pictures simultaneously and take the image from slightly different positions a respective difference needs to be corrected in such cases. Being, however, easily possible with commonly available methods. The embodiment as disclosed in figure 1 contains that a factor of the incomplete pixels is determined. Herein, the factor expresses the relative amount of the respective incomplete pixel being covered by the asset 11, wherein this data acquired by the second image is utilized to determine said factor. Said factor again can be utilized to evaluate the first image. Figure 2 shows a scheme of a cutout of a second image. Herein, the cutout is also separated into an even larger number of squares representing the pixels of the image. The number of pixels is higher than the number in the first image as the spatial resolution is higher. The second image is an RGB image being captured by a drone carrying an imaging device being no thermal imaging device. As the second image of figure 2 is the second image of a pair related to the first image the same asset 11 is shown. Namely, an overhead line power representing a transmission asset 11 and being monitored by a drone containing also the RGB camera and flying along the line to determine its status. The cutout also contains a part of the asset 11 and the background 12. From the second image, the average asset dimensions can be computed. For example, for a conductor, a sub-pixel fitting can be carried out to determine the width of the conductor which is assumed to be constant within the image. This information is then projected onto the first image. As described above the data of the second image is utilized by a computation module to process the first image and determine a thermal emission of the asset 11 in the first image. Allowing to acquire a significantly increased insight of the thermal emission of the asset 11 compared to using solely the first image. Especially, the level of details and resolution acquired of such thermal information is significantly increased allowing to differentiate different components of the asset 11 far better. Also, allowing to detect certain topics at all. As many observations like minor damages or a highly localized oxidation would not even be able to be detected without such improved thermal data. The example as shown in figure 1 contains such computation module in the drone. Allowing to directly process the acquired data. This is very beneficial for typical applications like overhead line inspection. Instantly identifying such problem that the asset 11 is partially covered in the first image and trying to correct the data accordingly allows to additionally include further loops at a given location to acquire even more pairs of first images and second images in case the available first images and second images are not providing enough data for such correction. Also, it is possible to specify that in case after a specified number of tries no corrected first image can be provided the process is started again from scratch from a more different angle. The first image as shown in the cutout of figure 1 and the second image as shown in the cutout of figure 2 are part of a plurality of pairs of the first image and the second image. The plurality of pairs of the first image and the second image are utilized to acquire 3D data of the asset 11. For example, it becomes possible to a highly detailed 3D thermal emission picture of the corresponding asset 11, which can be reviewed subsequently as desired. However, the plurality of pairs of the first image and the second image are not only used to create such 3D data of the asset 11. They are also used to check and correct the data acquired from the first image. For example, in case the asset 11 is partially covered by the environment in a facility or by plants outdoors it becomes possible to add data missing from a selected first image by using the data of a further first image and create a new version of the selected first image by calculating the missing part. Furthermore, the second image is utilized in the example as shown in figures 1 and 2 to determine characteristics of the asset 11. For example, the second image can be utilized to determine the exact asset 11 orientation, wherein this asset 11 characteristic is utilized to determine where the edges of the asset 11 should be in a cutout of the first image or second image. Also, the second image can be, for example, utilized to identify damages in a corrosion protection of the asset 11 or a corrosion that already exists. Allowing to correctly understand an impaired thermal emission at that location based on the corrosion influencing said thermal emission . Not shown is a third image. However, like described above additional data like a third image can be very beneficially utilized to gain even more detailed insight. Like using a LIDAR imaging device in addition to create the third image providing additional data. For example, LIDAR data can also be used to accurately estimate the width and location of the conductor that is visible in figure 1. Also not shown in the figures is the utilization of the construction data to improve the level of detail and reliability of the data acquired. For example, combined with characteristics of the asset 11 acquired by the second image it becomes possible to closely evaluate the thermal emission as shown in the first image and spot even deviations extending over a significant scale resulting in the deviation influencing the thermal emission over the whole first image taking into account theoretical data acquired from the construction data. Figure 3 shows a scheme of the cutout of a first image 1 further including a measurement window 5 and two reference windows 6. The measurement window 5 contains the full pixels 3 and incomplete pixels 4 containing the thermal emission of the asset 11. The reference windows 6 are located below and above the measurement window 5 and contain thermal emission of the background 12. The reference windows 6 are located near the measurement window 5. Herein, they are distanced one pixels to keep a safe distance while acquiring reference data representing the thermal emission of the background 12 of the asset 11. Such reference data can be beneficial for many reasons. For example, it can be utilized to identify irregularities located in the background 12 preventing the thermal emission of the asset 11 to be correctly identified. For example, some localized heat source in the background 12 overlapping with the thermal emission of the asset 11. Resulting, for example, in the shape of the asset 11 or the thermal emission to be incorrectly determined in the first image 1. It was also noted that the reliability of the data can be further improved by including a threshold value for a difference of the standard deviation within and / or between the two reference windows 6. In case the standard deviations differentiate two much a potential problem is indicated and an action is triggered. In the case as shown a second measurement is triggered and the first image 1 is blocked. While it might be used as reference data for further analysis a different first image 1 will be utilized for analysis. Based on the measurement and reference windows shown in figure 3, the following computation can be used to accurately determine the emission from the asset: The pixels of the reference window are projected back into the second image. In the second image, the high resolution pixels can be counted and an asset ratio is determined. The asset ratio gives the number of high resolution pixels that contain the asset in relation to the total number of pixels within the reference window. In the first image, the average emissions over the measurement and reference windows are computed. The asset ratio can then be used to solve the following relation: Measured^raditation = (1 - asset ratio) * background radiation + asset ratio * asset radiation The underlying assumption is that the radiation in the measurement window is composed of a certain amount originating from the asset, mixed with radiation from the background. The 5 asset ratio determined from the second high resolution image determines the linear mixing of these two source of radiation. As long as the background is homogeneous, the reference windows can be used to estimate the background radiation. 10 The present invention was only described in further detail for explanatory purposes. However, the invention is not to be understood being limited to these embodiments as they represent embodiments providing benefits to solve specific problems or fulfilling specific needs. The scope of the protec-15 tion should be understood to be only limited by the claims attached.

Claims

1. Method of measuring the thermal emission of an asset (11) being a power distribution asset (11) and / or power transmission asset (11), the method containing capturing a first image (1) of the asset (11) utilizing a first imaging device, wherein the first imaging device is a thermal imaging device, capturing a second image (2) of the asset (11) utilizing a second imaging device, wherein the second imaging device is not a thermal imaging device, wherein the second image (2) has a higher spatial resolution than the first image (1), processing the second image (2) to identify the asset (11) contained in the second image (2), processing the first image (1) utilizing a computation module containing a processing unit, wherein data regarding the asset (11) in the second image (2) is utilized to determine a thermal emission of the asset (11) in the first image (1).

2. Method according to claim 1, wherein the method contains the step of identify incomplete pixels (4) of the first image (1) utilizing the data regarding the asset (11) in the second image (2) , wherein the incomplete pixels (4) are pixels of the first image (1) being partially covered by the asset (11), wherein a factor is determined for the incomplete pixels (4) taking into account that the incomplete pixels (4) only partially cover the asset (11) in the first image (1), wherein the factor is utilized for the evaluation of the first image (1).

3. Method according to any of the preceding claims, wherein the method contains acquiring a plurality of pairs of the first image (1) and the second image (2), wherein the plurality of pairs of the first image (1) and the second image (2) are utilized to acquire 3D data.

4. Method according to any of the preceding claims, wherein the method contains acquiring at least three first images (1) and a second image (2), more preferred at least five first images (1) and a second image (2), wherein processing the at least three, more preferred at least five images is based on the second image (2) .

5. Method according to any of the preceding claims, wherein the method contains acquiring a first image (1) and at least two second images (2), more preferred at least three second images (2), even more preferred at least five second images (2) , wherein the at least two, more preferred at least three, even more preferred at least 5, second images (2) are processed to determine at least one characteristic of the asset (11), wherein the processing of the first image (1) takes into account the at least one characteristic..

6. Method according to any of the preceding claims, wherein the method contains acquiring a plurality of pairs of the first image (1) and the second image (2), wherein the plurality of pairs of the first image (1) and the second image (2) are utilized to acquire additional data to be included in at least one first image (1).

7. Method according to any of the preceding claims, wherein the method utilizes a drone containing the first imaging device and the second imaging device.

8. Method according to any of the preceding claims, wherein the method contains the step of processing the second image (2) to determine characteristics of the asset (11),wherein the characteristics contain at least one characteristic selected from the group consisting of asset shape, asset distance, asset size, asset orientation, conductor location, asset location, age, corrosion.

9. Method according to any of the preceding claims, wherein the second imaging device is a RGB camera or a LIDAR device.

10. Method according to any of the preceding claims, wherein the method contains the step of determining full pixels (3) and incomplete pixels (4) of the first image (1) based on the second picture providing a higher resolution than the first image (1) , wherein the full pixels (3) are pixels of the first image (1) being completely covered be the asset (11), wherein the incomplete pixels (4) are pixels of the first image (1) being partially covered by the asset (11), wherein the method contains the step of determining a measurement window (5), wherein the measurement window (5) contains the thermal emission of the asset (11) based on the full pixels (3) and incomplete pixels (4), wherein the method contains the step of determining at least one reference window (6), wherein the at least one reference window (6) does not include thermal emission from the asset (11), wherein preferably the reference window (6), more preferred pixels of the reference window (6), are near to the measurement window (5).

11. Method according to claim 10, wherein the at least two reference windows (6) are compared, wherein a standard deviation of the thermal emission of the at least two reference windows (6) is determined, wherein an action is triggered in case the standard deviation of the thermal emission of the at least two reference windows (6) exceeds a predefined threshold value.

12. Method according to any of the preceding claims, wherein the computation module estimates the size of the object and orientation using data acquired by a LIDAR device, or construction data.

13. Computer program product, tangibly embodied in a machine-readable storage medium, including instructions opera-bleto cause a computing entity to execute a method accordingto any of claims 1 to 12.

14. A thermography measurement system for measuring the thermal emission of an asset (11) being an power distribution asset (11) and / or power transmission asset (11) comprising: a first imaging device being a thermal imaging device for capturing thermal radiation from the asset (11);a second imaging device, wherein the second imaging device provides a higher resolution than the first imaging device, wherein the second imaging device is not a thermal imaging device, wherein the second imaging device provides a higher spatial resolution than the first imaging device, a computation module containing a processing unit, wherein the computation module is adapted to execute a method according to any of claims 1 to 12, a data storage containing a computer program product including instructions operable to cause the computation module to execute a method according to any of claims 1 to 12.

15. The thermography measurement system according to claim 14, wherein the thermography measurement system contains a drone, wherein the drone contains the first imaging device being a thermal imaging device, the second imaging device being no thermal imaging device, anda third imaging device, wherein the third imaging device is no thermal imaging device and wherein the third imaging device utilizes a different measurement method than the second imaging device.

Images