Termite mud line and mud quilt identification method, device and equipment based on hyperspectrum and medium

Abstract

The invention discloses a termite mud line and mud quilt identification method, device and equipment based on hyperspectrum and a medium, and belongs to the technical field of termite prevention and control. The method comprises the steps of collecting full-band hyperspectral images of embankment termite mud lines and mud covers, realizing hyperspectral abnormal target identification based on a global or local background deviation distance, and then clustering abnormal targets and carrying out statistical analysis on a clustering result. Large-range, full-coverage and high-efficiency cruise detection positioning of termite mud lines and mud covers and other ground surface activity traces is realized, the identification and detection rate is high, and the clutter interference resistance is high.

Description

Technical Field

[0001] This application belongs to the field of termite control technology, and in particular relates to a method, device, equipment and medium for identifying termite mud lines based on hyperspectral imaging. Background Technology

[0002] When soil-dwelling termites venture out to forage, they construct exposed protective structures such as mud lines and mud blankets on dikes or food surfaces. These structures serve as pathways, provide insulation and moisture retention, and defend against predators. These traces of activity on the ground are clear indicators for discovering termite activity areas and determining the species and level of activity of termites.

[0003] The mud lines and mud blankets are hollow, thin layers of mud primarily composed of soil particles, saliva, fecal secretions (containing wood fibers), and water. The saliva and fecal secretions contain organic matter such as mucin, unique to termites. The other components are identical to the surrounding soil and their preferred plant matter. Generally, the mud lines are only 0.5-2 cm wide, with most less than 1 cm; the mud blankets are roughly the size of the food they cover. Due to their hollow structure, the moisture content of the mud lines and mud blankets differs from the background soil; the greater the moisture difference, the greater the color difference.

[0004] Currently, termite control mainly relies on manual inspections to detect termite lineages and soil deposits. This method is inefficient, labor-intensive, and severely affected by terrain and vegetation obstruction, making it prone to missed detections. Drones equipped with high-definition cameras to collect images and AI-based target morphology recognition are also directions for technological development. However, this method is easily interfered with by background clutter such as dry grass, dead branches, and other animal tracks, resulting in numerous false alarms or missed detections. Summary of the Invention

[0005] This application provides a method, apparatus, and device for identifying termite mud lines and mud deposits based on hyperspectral imaging, which at least solves the problem of false alarms or missed alarms for termite mud lines and mud deposits in related technologies.

[0006] In a first aspect, embodiments of this application provide a method for identifying termite mud lines based on hyperspectral imaging, including: Hyperspectral images of the dam surface were acquired, and the hyperspectral image data covered the visible-near infrared (VISNIR) band and the short-wave infrared (SWIR) band. The hyperspectral image is preprocessed, and the preprocessed hyperspectral image data is converted into multiple spatial pixel spectral vectors; For the VISNIR and SWIR bands, calculate the mean vector and covariance matrix of the background distribution of the spatial pixel spectral vectors; Based on the mean vector and covariance matrix, the squared Mahalanobis distance of each pixel is calculated as the anomaly deviation score. Based on the abnormal deviation score, multiple abnormal target pixels that meet the preset abnormal target threshold conditions are selected, and an abnormal target pixel space and its corresponding abnormal deviation score set are formed. Divide the multiple abnormal target pixels in the abnormal target pixel space into multiple clusters; Based on the average abnormal deviation score of each cluster after division, the clusters corresponding to termite mud lines and / or mud cover are identified, thereby realizing the identification of termite hazards in dams.

[0007] Secondly, embodiments of this application provide a device for identifying termite mud lines based on hyperspectral imaging, comprising: The acquisition module is used to acquire hyperspectral image data of the dam surface, the hyperspectral image data covering the visible-near infrared band (VISNIR) and the short-wave infrared band (SWIR); The preprocessing module is used to preprocess the hyperspectral image data and convert the preprocessed hyperspectral image data into multiple spatial pixel spectral vectors. The first calculation module is used to calculate the mean vector and covariance matrix of the background distribution of the spatial pixel spectral vector for the VISNIR and SWIR bands. The second calculation module is used to calculate the squared Mahalanobis distance of each pixel as an anomaly deviation score based on the mean vector and the covariance matrix. The filtering module is used to filter out multiple abnormal target pixels that meet the preset abnormal target threshold conditions based on the abnormal deviation score, and to form an abnormal target pixel space and its corresponding abnormal deviation score set. A partitioning module is used to divide multiple abnormal target pixels in the abnormal target pixel space into multiple clusters; The identification module is used to identify clusters corresponding to termite mud lines and / or mud cover based on the average abnormal deviation score of each cluster after division, thereby realizing the identification of termite hazards in dams.

[0008] Thirdly, embodiments of this application provide an electronic device, which includes: a processor and a memory storing computer program instructions; when the processor executes the computer program instructions, it implements the steps of the hyperspectral-based termite mud line identification method as described in any embodiment of the first aspect.

[0009] Fourthly, embodiments of this application provide a computer-readable storage medium storing computer program instructions, which, when executed by a processor, implement the steps of the hyperspectral-based termite mud line identification method as described in any embodiment of the first aspect.

[0010] Fifthly, embodiments of this application provide a computer program product, which is stored in a storage medium and executed by at least one processor to implement the steps of the hyperspectral-based termite mud line identification method provided in the first aspect of embodiments of this application.

[0011] The method, apparatus, and equipment for identifying termite mud lines and mud deposits based on hyperspectral imaging in this application acquire full-band hyperspectral images of termite mud lines and mud deposits on dikes, identify hyperspectral abnormal targets based on global or local background deviation distance, and then cluster the abnormal targets and perform statistical analysis on the clustering results to achieve large-scale, full-coverage, and high-efficiency cruise detection and positioning of surface activity traces such as termite mud lines and mud deposits. It also has a high identification and detection rate and strong resistance to clutter interference. Attached Figure Description

[0012] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments of this application will be briefly introduced below. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0013] Figure 1 This is a schematic flowchart of a method for identifying termite mud lines based on hyperspectral imaging, provided in an embodiment of this application. Figure 2 This is a schematic diagram of the structure of a termite mud line identification device based on hyperspectral imaging provided in an embodiment of this application; Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application.

[0014] Figure label: A hyperspectral termite mud line identification device 200 comprises a data acquisition module 201, a preprocessing module 202, a first calculation module 203, a second calculation module 204, a screening module 205, a segmentation module 206, and an identification module 207. Electronic device 300, processor 301, memory 302, communication interface 303, bus 310. Detailed Implementation

[0015] The features and exemplary embodiments of various aspects of this application will be described in detail below. To make the objectives, technical solutions, and advantages of this application clearer, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only intended to explain this application and not to limit it. For those skilled in the art, this application can be implemented without some of these specific details. The following description of the embodiments is merely to provide a better understanding of this application by illustrating examples.

[0016] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising..." does not exclude the presence of additional identical elements in the process, method, article, or apparatus that includes said element.

[0017] When soil-dwelling termites venture out to forage, they construct exposed protective structures such as mud lines and mud blankets on dikes or food surfaces. These structures serve as pathways, provide insulation and moisture retention, and defend against predators. These traces of surface activity are the most direct evidence for discovering termite activity areas and determining the species and level of activity of termites.

[0018] The mud lines and mud blankets are hollow, thin layers of mud primarily composed of soil particles, saliva and fecal secretions, wood fibers, and water. The saliva and fecal secretions contain organic matter such as mucin, unique to termites. Other components are identical to those found in the local soil and their preferred plant matter. Generally, the width of the mud lines ranges from 0.5 to 2 cm, with the vast majority being less than 1 cm. The size of the mud blanket is roughly equivalent to the size of the food it covers. Due to their hollow structure, the mud lines and mud blankets maintain a different level of moisture compared to the background soil; the greater the difference in moisture, the greater the color difference.

[0019] Currently, termite control mainly relies on manual inspections to detect termite lineages and soil deposits. This method is inefficient, labor-intensive, and severely affected by terrain and vegetation obstruction, making it prone to missed detections. Drones equipped with high-definition cameras to collect images and AI-based target morphology recognition are among the directions of current technological development. However, this method is easily interfered with by background clutter such as dry grass, dead branches, and other animal tracks, resulting in numerous false alarms or missed detections.

[0020] To address the problems in the related technologies, embodiments of this application provide a method, apparatus, device, and medium for identifying termite mud lines based on hyperspectral imaging.

[0021] The method for identifying termite mud lines based on hyperspectral imaging provided in this application will be described in detail below with reference to the accompanying drawings, through specific embodiments and application scenarios.

[0022] Figure 1A schematic flowchart of a hyperspectral-based method for identifying termite mud lines, according to an embodiment of this application, is shown. Figure 1 As shown, the method for identifying termite mud lines based on hyperspectral imaging may specifically include the following steps: S101. Collect hyperspectral image data of the dam surface, wherein the hyperspectral image data covers the visible-near infrared band (VISNIR) and the short-wave infrared band (SWIR). S102. The hyperspectral image data is preprocessed, and the preprocessed hyperspectral image data is converted into a spectral vector of multiple spatial pixels; S103. For the VISNIR and SWIR bands, calculate the mean vector and covariance matrix of the background distribution of the spatial pixel spectral vector; S104. Based on the mean vector and covariance matrix, calculate the squared Mahalanobis distance of each pixel as the anomaly deviation score. S105. Based on the abnormal deviation score, select multiple abnormal target pixels that meet the preset abnormal target threshold conditions, and form an abnormal target pixel space and its corresponding abnormal deviation score set. S106. Divide the multiple abnormal target pixels in the abnormal target pixel space into multiple clusters; S107. For the multiple clusters after division, based on the average abnormal deviation score of each cluster, identify the clusters corresponding to the termite mud line and / or mud cover, thereby realizing the identification of termite hazards in the dam.

[0023] Therefore, by collecting full-band hyperspectral images of termite mud lines and mud coverings on dikes, hyperspectral anomaly target identification is achieved based on global or local background deviation distance. Then, by clustering the anomaly targets and performing statistical analysis on the clustering results, a large-scale, full-coverage, and highly efficient cruise detection and positioning of surface activity traces such as termite mud lines and mud coverings can be achieved. Furthermore, the detection rate is high and the anti-clutter interference capability is strong.

[0024] The specific implementation methods for each of the above steps are described below.

[0025] In some embodiments, in S101, a drone equipped with a hyperspectral camera or hyperspectral sensor is used to conduct low-altitude aerial photography to collect hyperspectral image data of the dam surface.

[0026] Specifically, hyperspectral images are obtained by acquiring electromagnetic spectrum imaging of surface materials through hyperspectral sensors. They are 3D datasets including two-dimensional spatial pixel information and one-dimensional spectral information. Each pixel covers a continuous spectral range from visible light to infrared, enabling the differentiation of materials with different chemical compositions or physical states. Due to their composition, density, and hollow structure, the intensity-wavelength curve (spectral characteristics) of light reflected or scattered by mudline mud deposits is specific compared to the surrounding soil. Therefore, hyperspectral imaging technology exhibits high detection rate and low false alarm rate in the detection of termite mudline mud deposits.

[0027] Optionally, the spectral range of the hyperspectral camera can include 400-2500 nm. Iron oxides (such as hematite) in termite mud lines and soil cover cause subtle but detectable changes in the shape of the visible-near-infrared (VISNIR: 580-680 nm) reflectance spectrum. Water molecules in the mud lines and soil cover exhibit strong absorption peaks at 1450 nm and 1940 nm in the short-wave infrared (SWIR) band, distinguishing them from dry background soil. The area around 1730 nm is the fundamental frequency absorption region of the CH bonds of proteins and amino acids in termite saliva within the mud lines and soil cover, while the 2100-2180 nm band represents the combination frequency and overtone absorption region of the NH and OH bonds of proteins and amino acids. All of these factors contribute to the different spectral characteristic curves of the mud lines and soil cover compared to the surrounding soil vegetation.

[0028] Optionally, the VISNIR band spectral channels can be set to 600, with 1920 spatial channels; the SWIR band spectral channels can be set to 250, with 640 spatial channels.

[0029] Optionally, the drone can fly at an altitude of 3m. Then, the VISNIR ground resolution GSD ≤ 0.7mm, a 1cm wide mudline image has approximately 14 pixels, and the ground coverage width is 1.35m; the SWIR ground resolution GSD ≤ 1.8mm, a 1cm wide mudline image has approximately 6 pixels, and the ground coverage width is 1.1m.

[0030] In some embodiments, in S102, the acquired hyperspectral image data is subjected to radiometric calibration, atmospheric correction, and geometric correction to obtain preprocessed hyperspectral image data. The preprocessed hyperspectral image data is surface reflectance data to eliminate the influence of illumination and atmosphere and ensure the authenticity of spectral information.

[0031] Furthermore, the preprocessed hyperspectral image is represented as a spatial pixel spectral vector as follows: .in, , representing the spatial row index, M=1920@VISNIR, M=640@SWIR; This represents a spatial column index, which is related to the length of the collected data. , indicating the spectral band index, B=600@VISNIR, B=250@SWIR.

[0032] It's understandable that the surface of a typical dam is surrounded by neat lawns. Besides anomalies like termites, mud lines, and soil, there are also a few distracting elements such as stones, wood chips, and cow dung. Therefore, the hyperspectral image data of a dam generally conforms to a Gaussian distribution. Anomaly detection is essentially mean shift detection against a Gaussian background; that is, anomalies are pixels whose spectral features significantly differ from the background mean. If different varieties of lawn are planted in sections of the dam slope, the image needs to be divided into several local windows, and anomaly detection should be performed separately within each window. This involves executing steps S103 to S105.

[0033] In some embodiments, in S103, the mean vector of the spatial pixel spectral vector is calculated according to the following formula (1). This refers to the average spectrum of the entire image or a local window.

[0034] ; (1) The covariance matrix of the background distribution of the spatial pixel spectral vector is calculated according to the following formula (2). The diagonal elements represent the variance of each band, indicating the degree of dispersion of the data in that band; the off-diagonal elements represent the covariance between different bands, i.e., the correlation between spectral bands.

[0035] ; (2) Where T represents transpose.

[0036] Furthermore, in some embodiments, in S104, the squared Mahalanobis distance of each pixel is calculated according to the following formula (3) as the abnormal deviation score. Then, The larger the value, the further the pixel is from the background.

[0037] (3) As an optional embodiment, a method for determining pixel points is established. The binary hypothesis testing model for whether it is background or an abnormal target is as follows (4): (4) in, The mean vector representing the background (including outliers). The mean vector representing the anomalous targets. Represents the background covariance matrix; Indicates the background assumption, that is obey , The multivariate Gaussian distribution belongs to the background; Indicates the abnormal hypothesis, that is obey , The multivariate Gaussian distribution indicates an abnormal target.

[0038] Furthermore, in some embodiments, in S105, based on the binary hypothesis testing model, abnormal targets can be screened by setting an appropriate threshold τ, as shown in equation (5): (5) In other words, When the value is less than τ, the pixel It is then determined to be the background spectrum; in When the value is greater than or equal to τ, the pixel The spectrum was determined to be an abnormal target.

[0039] Optionally, the threshold τ can be calculated using the following formula (6): ; (6) in, This represents the mean of the scores that deviate abnormally. This represents the standard deviation of the score from the out-of-range criteria. It is a proportionality coefficient and It needs to be calibrated according to the actual situation.

[0040] Furthermore, pixels that meet the preset abnormal target threshold conditions are filtered out and identified as abnormal target pixels. And constitutes the abnormal target pixel space. As shown in equation (7): (7) Furthermore, the set of abnormal deviation scores corresponding to the pixel space of the abnormal target is represented as: ; (8) in, H represents the number of abnormal target pixels; the H values ​​may differ between the VISNIR and SWIR bands.

[0041] It should be noted that in the abnormal target pixel space Y, there may be an abnormal spectral set of targets such as termite mud lines, mud blankets, rotten wood blocks, cow dung, and stones. Therefore, in this embodiment, the abnormal targets are automatically distinguished and classified according to their abnormal deviation scores by S106, thereby eliminating the interference of false abnormal targets.

[0042] As an optional embodiment, a feature vector is constructed based on the abnormal deviation score set; the feature vector is preprocessed to obtain a preprocessed feature matrix; the target number of clusters is determined for the preprocessed feature matrix; and the K-means clustering algorithm is used to divide multiple abnormal target pixels in the abnormal target pixel space into clusters corresponding to the target number of clusters.

[0043] In practice, feature vectors are constructed based on the abnormal deviation score set. That is, the abnormal deviation scores of multiple abnormal target pixels are used as a one-dimensional feature vector. As shown in equation (9): (9) In practice, since the abnormal deviation scores usually follow a heavy-tailed distribution, in order to further improve the clustering effect, the feature vector can be adjusted according to the following formula (10). Perform a logarithmic transformation on each element in the array to obtain the transformed elements. : ; (10) in, Let H represent the i-th element in the feature vector, and let H represent the number of abnormal target pixels. Avoid taking the logarithm of 0.

[0044] In specific implementation, the following formulas (11) to (13) are used to... Standardize to obtain standardized elements , making mean =0 and variance 1: ; (11) ;(12) ; (13) Therefore, multiple standardized elements can be combined to form a feature vector, as shown in equation (14): ;(14) Where T represents transpose.

[0045] Furthermore, based on feature vectors Calculate the distance between samples using This represents the average distance from sample i to other samples in the same cluster. This represents the average distance from sample i to its nearest neighbor cluster. The silhouette coefficient (...) is used. The effect of different numbers of clusters is evaluated. That is, for the number of clusters k, the silhouette coefficient of each sample is calculated according to the following formula (15). : ; (15) in, The value range is [-1, 1]. The closer it is to 1, the better the clustering effect of sample i. The closer it is to -1, the more likely sample i is to be incorrectly clustered. The closer it is to 0, the more likely sample i is to be on the boundary between two clusters.

[0046] Then, the overall profile coefficient can be calculated according to the following formula (16). : (16) Therefore, we can iterate through the possible range of cluster numbers k = 2, 3, The k value corresponding to the maximum silhouette coefficient is selected as the target number of clusters. As shown in equation (17). Typically, one can take... To avoid too many clusters (where, Indicates to Round down).

[0047] (17) Furthermore, use Perform K-means clustering to minimize the sum of squared distances from all samples to the centroids of their respective clusters. The objective function is constructed as follows (18): ; (18) in, It is Divided into Clusters that are complete and do not overlap Represents a set of clusters. This represents the center point of the cluster. The objective function aims to find a partition that minimizes the sum of squared distances.

[0048] In specific implementation, from Random selection Initialize cluster centers Each sample is then assigned to the nearest cluster center.

[0049] Then, the updated cluster centers are recalculated according to the following formula (19): ; (19) in, , Let these represent the cluster center point and the number of samples updated in the t-th iteration, respectively. The center point of the j-th cluster updated in the (t+1)-th iteration is the cluster in the t-th iteration. The mean of all samples within the range.

[0050] Determine whether the convergence condition of the following formula (20) is met. If the convergence condition is met, it means that the center point of the iteration is stable, and the iteration is stopped.

[0051] ; (20) in, The tolerance threshold representing the change in the center point is set based on experience. .

[0052] Furthermore, in some embodiments, in S107, for the multiple clusters after partitioning, each cluster is calculated according to the following formulas (21) and (22). Number of similar abnormal targets and average abnormal deviation score : ; (twenty one) . (twenty two) Therefore, the average abnormal deviation score will be... The smallest cluster was identified as termite mud lines, which will affect the average abnormal deviation score. The smallest clusters were identified as termite mud sheets. It should be understood that termite mud lines are thin, linear mud sheets, small in area, and their composition is more similar to the background soil. Therefore, the distance between the hyperspectral pixels and the background... Minimum; generally, mud is a blocky layer of mud covering the surface of dead branches and dry grass; the distance between hyperspectral pixels and the background. The second smallest; interfering objects such as cow dung and stones differ significantly from the soil background in terms of material composition, and their hyperspectral pixels are far from the background. Larger.

[0053] It should be noted that the above description describes some embodiments of this application. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in a different order than that shown in the above embodiments and still achieve the desired result. Furthermore, the processes depicted in the drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0054] Based on the same technical concept, and corresponding to any of the above embodiments, this application also provides a hyperspectral termite mud line identification device 200.

[0055] like Figure 2As shown, the hyperspectral-based termite mud line identification device 200 may include: Acquisition module 201 is used to acquire hyperspectral image data of the dam surface, the hyperspectral image data covering the visible-near infrared band (VISNIR) and the short-wave infrared band (SWIR); The preprocessing module 202 is used to preprocess the hyperspectral image data and convert the preprocessed hyperspectral image data into multiple spatial pixel spectral vectors. The first calculation module 203 is used to calculate the mean vector and covariance matrix of the background distribution of the spatial pixel spectral vector for the VISNIR band and SWIR band. The second calculation module 204 is used to calculate the squared Mahalanobis distance of each pixel as an anomaly deviation score based on the mean vector and the covariance matrix. The filtering module 205 is used to filter out multiple abnormal target pixels that meet the preset abnormal target threshold conditions based on the abnormal deviation score, and to form an abnormal target pixel space and its corresponding abnormal deviation score set. The partitioning module 206 is used to divide multiple abnormal target pixels in the abnormal target pixel space into multiple clusters; The identification module 207 is used to identify clusters corresponding to termite mud lines and / or mud blankets based on the average abnormal deviation score of each cluster after division of multiple clusters, thereby realizing the identification of termite hazards in dams.

[0056] It should be noted that, for ease of description, the above devices are described in terms of function, divided into various modules. Of course, in implementing this application, the functions of each module can be implemented in one or more software and / or hardware.

[0057] The apparatus described above is used to implement the corresponding hyperspectral-based termite mud line identification method in any of the foregoing embodiments, and has the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0058] Based on the same technical concept, corresponding to any of the above embodiments, this application also provides an electronic device.

[0059] Figure 3 A schematic diagram of a more specific electronic device hardware structure provided in this embodiment is shown.

[0060] The electronic device 300 may include a processor 301 and a memory 302 storing computer program instructions.

[0061] Specifically, the processor 301 may include a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits that can be configured to implement the embodiments of this application.

[0062] Memory 302 may include mass storage for data or instructions. For example, and not limitingly, memory 302 may include a hard disk drive (HDD), floppy disk drive, flash memory, optical disk, magneto-optical disk, magnetic tape, or Universal Serial Bus (USB) drive, or a combination of two or more of these. Where appropriate, memory 302 may include removable or non-removable (or fixed) media. Where appropriate, memory 302 may be internal or external to the integrated gateway disaster recovery device. In a particular embodiment, memory 302 is non-volatile solid-state memory.

[0063] In certain embodiments, the memory may include read-only memory (ROM), random access memory (RAM), disk storage media devices, optical storage media devices, flash memory devices, and electrical, optical, or other physical / tangible memory storage devices. Thus, generally, memory includes one or more tangible (non-transitory) computer-readable storage media (e.g., memory devices) encoded with software including computer-executable instructions, and when the software is executed (e.g., by one or more processors), it is operable to perform the operations described with reference to the method according to one aspect of this application.

[0064] The processor 301 reads and executes computer program instructions stored in the memory 302 to implement any of the hyperspectral-based termite mud line identification methods in the above embodiments.

[0065] In some examples, the electronic device 300 may also include a communication interface 303 and a bus 310. For example, Figure 3 As shown, the processor 301, memory 302, and communication interface 303 are connected through bus 310 and complete communication with each other.

[0066] The communication interface 303 is mainly used to realize communication between various modules, devices, units and / or equipment in the embodiments of this application.

[0067] Bus 310 includes hardware, software, or both, that couples components of an online data traffic metering device together. For example, and not as a limitation, bus 310 may include an Accelerated Graphics Port (AGP) or other graphics bus, an Enhanced Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), HyperTransport (HT) interconnect, an Industry Standard Architecture (ISA) bus, an Infinite Bandwidth Interconnect, a Low Pin Count (LPC) bus, a memory bus, a Microchannel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association Local (VLB) bus, or other suitable buses, or combinations of two or more of these. Where appropriate, bus 310 may include one or more buses. Although specific buses are described and illustrated in embodiments of this application, any suitable bus or interconnect is contemplated herein.

[0068] For example, the electronic device 300 can be a mobile phone, tablet computer, laptop computer, handheld computer, in-vehicle electronic device, ultra-mobile personal computer (UMPC), netbook, or personal digital assistant (PDA), etc.

[0069] Based on the same technical concept, corresponding to any of the methods in the above embodiments, this application also provides a non-transitory computer-readable storage medium. This computer-readable storage medium stores computer program instructions; when executed by a processor, these computer program instructions implement any of the hyperspectral-based termite mud line identification methods in the above embodiments. Examples of computer-readable storage media include non-transitory computer-readable storage media such as portable disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, etc.

[0070] Based on the same technical concept, corresponding to any of the above-described embodiments, this application also provides a computer program product, which includes computer program instructions. In some embodiments, the computer program instructions can be executed by one or more processors of a computer to cause the computer and / or the processor to perform the hyperspectral-based termite mud line identification method. Corresponding to the execution entity for each step in each embodiment of the hyperspectral-based termite mud line identification method, the processor executing the corresponding step can belong to the corresponding execution entity.

[0071] It should be clarified that this application is not limited to the specific configurations and processes described above and shown in the figures. For the sake of brevity, detailed descriptions of known methods are omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of this application is not limited to the specific steps described and shown. Those skilled in the art can make various changes, modifications, and additions, or change the order of steps, after understanding the spirit of this application.

[0072] The functional blocks shown in the above-described structural diagram can be implemented as hardware, software, firmware, or a combination thereof. When implemented in hardware, they can be, for example, electronic circuits, application-specific integrated circuits (ASICs), appropriate firmware, plug-ins, function cards, etc. When implemented in software, the elements of this application are programs or code segments used to perform the required tasks. Programs or code segments can be stored on a machine-readable medium or transmitted over a transmission medium or communication link via data signals carried on a carrier wave. "Machine-readable medium" can include any medium capable of storing or transmitting information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disks, CD-ROMs, optical disks, hard disks, fiber optic media, radio frequency (RF) links, etc. Code segments can be downloaded via computer networks such as the Internet, intranets, etc.

[0073] It should also be noted that the exemplary embodiments mentioned in this application describe methods or systems based on a series of steps or apparatus. However, this application is not limited to the order of the above steps; that is, the steps can be performed in the order mentioned in the embodiments, or in a different order, or several steps can be performed simultaneously.

[0074] The aspects of this application have been described above with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It should be understood that each block in the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing apparatus to produce a machine such that these instructions, executable via the processor of the computer or other programmable data processing apparatus, enable the implementation of the functions / actions specified in one or more blocks of the flowchart illustrations and / or block diagrams. Such a processor can be, but is not limited to, a general-purpose processor, a special-purpose processor, a special application processor, or a field-programmable logic circuit. It is also understood that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can also be implemented by dedicated hardware performing the specified functions or actions, or can be implemented by a combination of dedicated hardware and computer instructions.

[0075] The above description is merely a specific implementation of this application. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, modules, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here. It should be understood that the protection scope of this application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and these modifications or substitutions should all be covered within the protection scope of this application.

Claims

1. A method for identifying termite mud lines based on hyperspectral imaging, characterized in that, include: Hyperspectral image data of the dam surface were collected, and the hyperspectral image data covered the visible-near infrared band (VISNIR) and the short-wave infrared band (SWIR). The hyperspectral image data is preprocessed, and the preprocessed hyperspectral image data is converted into multiple spatial pixel spectral vectors; For the VISNIR and SWIR bands, calculate the mean vector and covariance matrix of the background distribution of the spatial pixel spectral vectors; Based on the mean vector and covariance matrix, the squared Mahalanobis distance of each pixel is calculated as the anomaly deviation score. Based on the abnormal deviation scores, abnormal target pixels that meet the set threshold conditions are selected to form an abnormal target pixel space and its corresponding abnormal deviation score set. Divide the multiple abnormal target pixels in the abnormal target pixel space into multiple clusters; Based on the average abnormal deviation score of each cluster after division, the clusters corresponding to termite mud lines and / or mud cover are identified, thereby realizing the identification of termite hazards in dams.

2. The method according to claim 1, characterized in that, The multiple abnormal target pixels in the abnormal target pixel space are divided into multiple clusters, including: Based on the set of abnormal deviation scores, construct a feature vector; The feature vectors are preprocessed to obtain the preprocessed feature matrix; For the preprocessed feature matrix, determine the target number of clusters; The K-means clustering algorithm is used to divide multiple abnormal target pixels in the abnormal target pixel space into clusters corresponding to the target cluster number.

3. The method according to claim 2, characterized in that, The preprocessing of the feature vector to obtain the preprocessed feature matrix includes: The transformed elements are obtained by performing a logarithmic transformation on each element of the feature vector according to the following formula. : ； in, Let H represent the i-th element in the feature vector, and let H represent the number of abnormal target pixels. It represents a small positive number; According to the following formula Standardize to obtain standardized elements , making mean =0 and variance 1: ； ； ； The feature matrix is ​​composed of multiple standardized elements. , where T represents transpose.

4. The method according to claim 1, characterized in that, The process of identifying clusters corresponding to termite mud lines and / or mud blankets based on the average abnormal deviation score of each cluster after division includes: For the multiple clusters after division, the cluster with the smallest average abnormal deviation score is identified as termite mud line, and the cluster with the second smallest average abnormal deviation score is identified as termite mud blanket.

5. The method according to claim 1, characterized in that, The hyperspectral image data of the dam surface collected includes: Low-altitude aerial photography was conducted using drones equipped with hyperspectral cameras to collect hyperspectral image data of the dam surface.

6. The method according to claim 1, characterized in that, The preprocessing of the hyperspectral image data includes: performing radiometric calibration, atmospheric correction, and geometric correction on the hyperspectral image data to obtain preprocessed hyperspectral image data.

7. A device for identifying termite mud lines based on hyperspectral imaging, characterized in that, include: The acquisition module is used to acquire hyperspectral image data of the dam surface, the hyperspectral image data covering the visible-near infrared band (VISNIR) and the short-wave infrared band (SWIR); The preprocessing module is used to preprocess the hyperspectral image data and convert the preprocessed hyperspectral image data into multiple spatial pixel spectral vectors. The first calculation module is used to calculate the mean vector and covariance matrix of the background distribution of the spatial pixel spectral vector for the VISNIR and SWIR bands. The second calculation module is used to calculate the squared Mahalanobis distance of each pixel as an anomaly deviation score based on the mean vector and the covariance matrix. The filtering module is used to filter out multiple abnormal target pixels that meet the preset abnormal target threshold conditions based on the abnormal deviation score, and to form an abnormal target pixel space and its corresponding abnormal deviation score set. A partitioning module is used to divide multiple abnormal target pixels in the abnormal target pixel space into multiple clusters; The identification module is used to identify clusters corresponding to termite mud lines and / or mud blankets based on the average abnormal deviation score of each cluster after division, thereby realizing the identification of termite hazards in dams.

8. An electronic device, characterized in that, The device includes: a processor and a memory storing computer program instructions; when the processor invokes the computer program instructions, it implements the hyperspectral-based termite mud line identification method as described in any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer program instructions, which, when invoked by a processor, implement the hyperspectral-based termite mud line identification method as described in any one of claims 1-6.

10. A computer program product, characterized in that, When the instructions in the computer program product are executed by the processor of the electronic device, the electronic device performs the hyperspectral-based termite mud line identification method as described in any one of claims 1-6.

Images