Satellite remote sensing data-based cloud tracking method and device, and electronic equipment

Abstract

The application provides a satellite remote sensing data-based tracer cloud tracking method and device and electronic equipment, which can efficiently and accurately track and detect tracer clouds. The method comprises the following steps: acquiring a first satellite image and a second satellite image; performing similarity analysis on the first satellite image and the second satellite image, and determining an original region and a changed region in the first satellite image and the second satellite image respectively; performing image enhancement transformation on the first satellite image and the second satellite image, and correspondingly generating a first enhanced image and a second enhanced image; and determining a cloud target corresponding to the tracer cloud in the original region in the changed region by performing image matching on the first enhanced image and the second enhanced image. The device comprises an image acquisition module, a similarity analysis module, an image enhancement module and a target tracking module. The electronic equipment is used for executing the tracer cloud tracking method.

Description

Technical Field

[0001] One or more embodiments of this application relate to the field of satellite remote sensing technology, and in particular to a method, apparatus and electronic device for tracking cloud based on satellite remote sensing data. Background Technology

[0002] In the fields of satellite remote sensing and weather forecasting, accurate and efficient tracking and monitoring of tracer clouds is a crucial foundation for aerosol prediction research. Some related technologies employ image matching based on satellite remote sensing images for tracer tracking. However, due to the rich content and massive data volume of satellite remote sensing images, directly using them for image matching requires high computational power, has low efficiency, and results in poor overall matching and tracking performance. Other related technologies utilize deep learning methods to track tracer clouds using neural networks. However, building mature deep neural network models often requires large amounts of sample image data as a training basis. In the fields of satellite remote sensing and weather forecasting, the complexity of actual weather conditions makes it difficult to obtain sufficient and suitable sample data. Therefore, the tracking results using deep learning methods are often unsatisfactory. Summary of the Invention

[0003] In view of this, the purpose of one or more embodiments of this application is to provide a method, apparatus and electronic device for tracking tracer clouds based on satellite remote sensing data, which can efficiently and accurately track and detect tracer clouds.

[0004] In view of the above objectives, in a first aspect, one or more embodiments of this application provide a tracking cloud method based on satellite remote sensing data, comprising:

[0005] First and second satellite images were acquired from high-resolution satellite remote sensing video data;

[0006] By performing similarity analysis on the first satellite image and the second satellite image, a changed region is determined in the second satellite image, and an original region corresponding to the changed region is determined in the first satellite image, wherein the original region contains a tracer cloud;

[0007] Image enhancement transformations are performed on the first satellite image and the second satellite image to generate a first enhanced image and a second enhanced image respectively;

[0008] By performing image matching between the first enhanced image and the second enhanced image, a cloud target corresponding to the tracer cloud in the original region is determined in the changed region.

[0009] Optionally, the step of determining the change region in the second satellite image by performing similarity analysis on the first satellite image and the second satellite image includes:

[0010] The grid feature data of the first satellite image and the second satellite image are extracted using a convolutional neural network;

[0011] The Mahalanobis distance between multiple image grids of the second satellite image and the first satellite image is calculated based on the grid feature data.

[0012] The Mahalanobis distance of multiple image grids is compared with a region segmentation threshold, and multiple image grids whose Mahalanobis distance is less than the region segmentation threshold are selected to form the change region.

[0013] Optionally, the method for determining the region segmentation threshold includes:

[0014] Obtain a baseline sample image and a corresponding labeled sample image;

[0015] Calculate the Mahalanobis distance between the changed and unchanged regions of the labeled sample image compared to the baseline sample image, and generate a Mahalanobis distance histogram.

[0016] The region segmentation threshold is determined based on the Mahalanobis distance histogram.

[0017] Optionally, the step of performing image enhancement transformation on the first satellite image and the second satellite image respectively includes:

[0018] Perform piecewise linear transformation on the image grayscale of the first satellite image and the second satellite image;

[0019] Gaussian low-pass filtering transformation is performed on the frequency domain data corresponding to the first satellite image and the second satellite image.

[0020] Optionally, the transformation formula for piecewise linear transformation of the image grayscale is:

[0021]

[0022] Where h represents the transformed pixel gray value, f(x,y) represents the original pixel (x,y) gray value, L represents the maximum gray value of the image, and a, b, c, d represent the gray value segments.

[0023] Optionally, performing Gaussian low-pass filtering on the frequency domain data corresponding to the first satellite image and the second satellite image includes:

[0024] A two-dimensional Gaussian low-pass filter is used to filter and transform the frequency domain data of the corresponding frequency domain data of the first satellite image and the second satellite image;

[0025] The transfer function of the two-dimensional Gaussian low-pass filter is:

[0026]

[0027] Where D(u,v) represents the distance between the point (u,v) in the spectrum image and the origin of the Fourier transform, and D0 represents the cutoff frequency.

[0028] Optionally, determining the cloud target corresponding to the tracer cloud in the original region in the changed region by performing image matching between the first enhanced image and the second enhanced image includes:

[0029] A target window is determined in the first enhanced image based on the original image, and a search window is determined in the second enhanced image based on the changed region. The target window is then matched with the search window to determine a set of positive matching points.

[0030] A target window is determined in the second enhanced image based on the changed region, and a search window is determined in the second enhanced image based on the original region. The target window is then matched with the search window to determine a set of inverse matching points.

[0031] The intersection of the forward matching point set and the reverse matching point set is taken as the successful matching point set;

[0032] Based on the set of successfully matched points, multiple target points corresponding to the tracer cloud in the original region are determined in the changed region, and the cloud target corresponding to the tracer cloud is determined based on the multiple target points.

[0033] Optionally, the step of matching the target window with the search window includes:

[0034] Calculate the correlation coefficient between multiple pixels in the target window and multiple pixels in the search window, and determine the matching point between the target window and the search window based on the correlation coefficient;

[0035] Wherein, the correlation coefficient is:

[0036]

[0037] ρ represents the correlation coefficient, g(c,r) represents the gray value of pixel (c,r), (c1,r1) represents the pixel coordinates of the target window, (c2,r2) represents the pixel coordinates of the search window, i and j are the pixel indices of the target window and the search window, respectively, and the values ​​of i and j are both in the range of 1 to n. Based on the above objective, in a second aspect, one or more embodiments of this application also provide a tracking cloud device based on satellite remote sensing data, comprising:

[0038] The image acquisition module is used to acquire first and second satellite images from high-resolution satellite remote sensing video data;

[0039] The similarity analysis module is used to perform similarity analysis between the first satellite image and the second satellite image, determine the changed region in the second satellite image, and determine the original region in the first satellite image corresponding to the changed region, wherein the original region contains a tracer cloud;

[0040] An image enhancement module is used to perform image enhancement transformations on the first satellite image and the second satellite image, generating a first enhanced image and a second enhanced image accordingly; and

[0041] The target tracking module is used to determine a cloud target in the changed region corresponding to the tracer cloud in the original region by performing image matching between the first enhanced image and the second enhanced image.

[0042] In view of the above objectives, in a third aspect, one or more embodiments of this application also provide an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the tracking cloud method based on satellite remote sensing data as described in the first aspect.

[0043] As can be seen from the above, the tracking cloud method, apparatus, and electronic device based on satellite remote sensing data provided in one or more embodiments of this application have the following beneficial technical effects:

[0044] (1) By performing similarity analysis on the first satellite image and the second satellite image, the original region and the corresponding change region are determined in the first satellite image and the second satellite image, respectively, thereby greatly reducing the data range that needs to be focused on and greatly reducing the amount of data calculation for subsequent image matching, thereby improving the calculation efficiency; by performing enhancement transformation on the first satellite image and the second satellite image, the information data of the satellite image is optimized and reduced in a targeted manner, which can improve the targeting and accuracy of subsequent image matching calculation; finally, image matching is performed based on the original region and the change region, thereby determining the cloud target corresponding to the tracer cloud in the change region according to the matching result, which can realize the tracking and detection of the tracer cloud.

[0045] (2) When performing image matching between the first enhanced image and the second enhanced image based on the original region and the changed region, the bidirectional correlation coefficient method is used for matching. Only when both the forward and reverse matching passes is the matching considered successful. This method verifies the single matching result and removes the influence of noise points, thereby improving the matching accuracy and ensuring the accuracy of the matching calculation result. Attached Figure Description

[0046] To more clearly illustrate the technical solutions in one or more embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only one or more embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0047] Figure 1 A schematic diagram of a tracking cloud method based on satellite remote sensing data provided in one or more embodiments of this application;

[0048] Figure 2 A schematic diagram of a similarity analysis method in a tracking cloud method based on satellite remote sensing data provided in one or more embodiments of this application;

[0049] Figure 3 A piecewise linear transformation function graph in a tracking cloud method based on satellite remote sensing data provided in one or more embodiments of this application;

[0050] Figure 4 A schematic diagram of a bidirectional correlation coefficient matching method in a satellite remote sensing data-based cloud tracking method provided in one or more embodiments of this application;

[0051] Figure 5A schematic diagram of a tracking cloud device based on satellite remote sensing data provided for one or more embodiments of this application;

[0052] Figure 6 This is a schematic diagram of the structure of a tracking cloud electronic device based on satellite remote sensing data, provided for one or more embodiments of this application. Detailed Implementation

[0053] To make the objectives, technical solutions, and advantages of this disclosure clearer, the following detailed description is provided in conjunction with specific embodiments and the accompanying drawings.

[0054] It should be noted that, unless otherwise defined, the technical or scientific terms used in one or more embodiments of this application should have the ordinary meaning understood by one of ordinary skill in the art to which this disclosure pertains. The terms "first," "second," and similar terms used in one or more embodiments of this application do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed following the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0055] The technical solution of this application will be described below with reference to specific embodiments.

[0056] In one aspect, embodiments of this application provide a method for tracking cloud tracer based on satellite remote sensing data.

[0057] like Figure 1 As shown, one or more optional embodiments of this application provide a tracking cloud method based on satellite remote sensing data, comprising:

[0058] S1: Obtain the first satellite image and the second satellite image from high-resolution satellite remote sensing video data.

[0059] A first satellite image and a second satellite image can be sequentially extracted from high-resolution satellite remote sensing video data. The first satellite image contains a tracer cloud, and the second satellite image contains a cloud target corresponding to the tracer cloud in the first satellite image. In some embodiments, the first satellite image and the second satellite image can be two consecutive frames. For example, the second satellite image is the current frame in the high-resolution satellite remote sensing video, and the first satellite image is the previous frame. The purpose of this application is to track the tracer cloud in the previous frame within the current frame. In some embodiments, there may also be a gap of several frames between the first satellite image and the second satellite image. The time interval or the number of frame intervals between the first satellite image and the second satellite image can be flexibly adjusted and set according to actual conditions.

[0060] S2: By performing similarity analysis on the first satellite image and the second satellite image, a changed region is determined in the second satellite image, and an original region corresponding to the changed region is determined in the first satellite image, wherein the original region contains a tracer cloud.

[0061] It is understandable that the movement and changes of the tracer cloud are reflected in the satellite remote sensing data as changes in the content of the satellite image. Therefore, in some optional embodiments, similarity analysis can be performed on the first satellite image and the second satellite image to determine the area where the second satellite image differs from the first satellite image, i.e., the changed area, thereby narrowing down the area of ​​interest to the original area and the changed area.

[0062] S3: Perform image enhancement transformation on the first satellite image and the second satellite image to generate a first enhanced image and a second enhanced image respectively.

[0063] Image enhancement involves adding information or transforming data to an original image using specific methods. This selectively highlights features of interest or suppresses unwanted features to meet specific analytical needs. Image enhancement transformations can be applied to both the first and second satellite images separately, allowing for targeted optimization and adjustment of the information content within the satellite images, facilitating subsequent image matching and analysis. Image enhancement transformations can be categorized into spatial domain enhancement and frequency domain enhancement. Spatial domain enhancement methods include linear transformations, nonlinear transformations, histogram equalization, histogram specification, and spatial domain filtering, while frequency domain enhancement methods include smoothing, sharpening, high-pass filtering, and low-pass filtering.

[0064] S4: By performing image matching between the first enhanced image and the second enhanced image, a cloud target corresponding to the tracer cloud in the original region is determined in the changed region.

[0065] By focusing on the original region in the first enhanced image and the changed region in the second enhanced image, image matching can be performed between the two. Multiple points corresponding to the original region in the first enhanced image can be identified in the changed region of the second enhanced image, thereby further identifying multiple points corresponding to the tracer cloud in the original region. Thus, cloud targets corresponding to the tracer cloud can be identified in the changed region, enabling the tracking and identification of the tracer cloud in the second satellite image.

[0066] The aforementioned satellite remote sensing data-based cloud tracking method analyzes the similarity between the first and second satellite images to determine the original region and corresponding changed regions in both images, thereby significantly narrowing the scope of data requiring attention and greatly reducing the computational load for subsequent image matching, thus improving computational efficiency. Furthermore, by performing enhancement transformations on the first and second satellite images, the information data of the satellite images is optimized and reduced in a targeted manner, improving the targeting and accuracy of subsequent image matching calculations. Finally, image matching is performed based on the original region and the changed region, and the cloud target corresponding to the tracer cloud in the changed region is determined according to the matching result, enabling the tracking and detection of the tracer cloud.

[0067] like Figure 2 As shown, in one or more optional embodiments of this application, a method for tracing cloud based on satellite remote sensing data is provided, wherein determining the change region in the second satellite image by performing similarity analysis on the first satellite image and the second satellite image includes:

[0068] S201: Use a convolutional neural network to extract grid feature data from the first satellite image and the second satellite image.

[0069] Convolutional neural network models can be used to extract features from high spatial resolution satellite remote sensing images, obtaining downsampled grid feature data. This grid feature data possesses a high degree of abstraction, can adapt to complex deformations and lighting variations, and can effectively describe the overall characteristics of the target.

[0070] S202: Calculate the Mahalanobis distance between multiple image grids of the second satellite image and the first satellite image based on the grid feature data.

[0071] Mahalanobis distance can be used to characterize the phase velocity of the first satellite image and the second satellite image. The Mahalanobis distance between the corresponding grids is calculated based on the grid feature data of multiple grids in the first satellite image and the grid feature data of multiple grids in the second satellite image.

[0072] S203: Compare the Mahalanobis distance of the multiple image grids with a region segmentation threshold, and select multiple image grids whose Mahalanobis distance is less than the region segmentation threshold to form the change region.

[0073] The Mahalanobis distance is used to characterize the similarity of image content. For image grids with a Mahalanobis distance less than the region segmentation threshold, it indicates that their image content has changed significantly and they belong to the changed region. The region segmentation threshold, as a standard for judging whether the image content has changed significantly, can be obtained based on historical image data.

[0074] In some optional embodiments of this application, the method for determining the region segmentation threshold includes:

[0075] A baseline sample image and a corresponding labeled sample image are obtained. The labeled sample image is compared to the baseline sample image, and unchanging and changed regions are marked in the labeled sample image.

[0076] The Mahalanobis distance between the changed and unchanged regions of the labeled sample image compared to the baseline sample image is calculated, and a Mahalanobis distance histogram is generated. There is a significant difference in the Mahalanobis distance between the unchanged and changed regions. Based on the corresponding Mahalanobis distance histogram, the difference in Mahalanobis distance between different regions can be intuitively determined, thereby allowing the determination of the region segmentation threshold for dividing the unchanged and changed regions based on the Mahalanobis distance histogram.

[0077] In one or more optional embodiments of this application, a tracking cloud method based on satellite remote sensing data is provided, wherein the image enhancement transformation of the first satellite image and the second satellite image includes:

[0078] A piecewise linear transformation is performed on the grayscale values ​​of the first satellite image and the second satellite image.

[0079] Gaussian low-pass filtering transformation is performed on the frequency domain data corresponding to the first satellite image and the second satellite image.

[0080] The piecewise linear transformation is a spatial enhancement, as shown in the figure. The piecewise linear transformation can expand the grayscale range of useful information on the image and increase the contrast, while the grayscale range of corresponding noise is compressed to a smaller area at the end.

[0081] The transformation formula for piecewise linear transformation of the image grayscale is:

[0082]

[0083] Where h represents the transformed pixel grayscale value, f(x,y) represents the original grayscale value of pixel (x,y), L represents the maximum grayscale value of the image, and a, b, c, d represent the grayscale segment values. The grayscale segment values ​​a, b, c, d can be flexibly adjusted according to actual conditions.

[0084] (0,0)

[0085] Among them, Gaussian low-pass filtering transformation belongs to frequency domain transformation. In some optional embodiments of this application, the step of performing Gaussian low-pass filtering transformation on the corresponding frequency domain data of the first satellite image and the second satellite image includes:

[0086] A two-dimensional Gaussian low-pass filter is used to filter and transform the frequency domain data of the corresponding frequency domain data of the first satellite image and the second satellite image;

[0087] The transfer function of the two-dimensional Gaussian low-pass filter is:

[0088]

[0089] Where D(u,v) represents the distance between the point (u,v) in the spectrum image and the origin of the Fourier transform, and D0 represents the cutoff frequency.

[0090] The Gaussian low-pass filter change retains low-frequency data information in the frequency domain, suppresses high frequencies that reflect the edge features of sudden grayscale changes, and filters out isolated point noise contained in high frequencies, resulting in a noise reduction and smoothing effect.

[0091] like Figure 4 As shown, in one or more optional embodiments of this application, a method for tracking tracer clouds based on satellite remote sensing data is provided. The step of determining a cloud target corresponding to the tracer cloud in the original region within the changed region by performing image matching between the first enhanced image and the second enhanced image includes:

[0092] S401: Determine a target window in the first enhanced image based on the original image, determine a search window in the second enhanced image based on the changed region, and match the target window with the search window to determine a positive matching point set.

[0093] Using the first enhanced image as the left image and the second enhanced image as the right image, forward matching is performed. Matching points corresponding to multiple pixels in the target window are determined within the search window. The matching result is multiple sets of matching points, which constitute the forward matching point set. For example, m1 in the search window is determined through image matching. ′ The point matches point m1 in the target window, and point m1 matches point m1. ′The points form a matching point group (m1, m1) ′ By matching, p groups of matching points are determined, forming the forward matching point group [(m1, m1...]. ′ ),(m2,m2 ′ ),…,(m p ,m p ′ )).

[0094] S402: Determine a target window in the second enhanced image based on the changed region, determine a search window in the second enhanced image based on the original region, call the target window and the search window to match, and determine the inverse matching point set.

[0095] In reverse matching, the left and right images are swapped, with the second enhanced image as the left image and the first enhanced image as the right image. Image matching is then performed using the target window and the search window. Reverse matching also determines multiple sets of matching points to form a reverse matching point set. For example, m2 in the search window can be determined through image matching. ′ The point matches point m2 in the target window, and point m2 matches point m2. ′ The points form a matching point group (m2, m2) ′ ). By matching, q sets of matching points are determined, forming the reverse matching point group [(m2, m2)]. ′ ),(m3,m3 ′ ),…,(m p ,m p ′ ),…,(m q-2 ,m q-2 ′ (m) q-1 ,m q-1 ′ )]

[0096] S403: The intersection of the forward matching point set and the reverse matching point set is taken as the successful matching point set. Taking the above two matching point sets as an example, the forward matching point group [(m1,m1...]...] ′ ),(m2,m2 ′ ),…,(m p ,m p ′ )] and the reverse matching point group [(m2,m2)] ′ ),(m3,m3 ′ ),…,(m p ,m p ′ ),…,(m q-2 ,m q-2 ′ (m) q-1 ,mq-1 ′ The intersection of [(m2,m2)] is [(m2,m2)] ′ ),(m3,m3 ′ ),…,(m p ,m p ′ )).

[0097] S404; Based on the successfully matched point set, determine multiple target points in the changed region corresponding to the tracer cloud in the original region, and based on the multiple target points, determine the cloud target corresponding to the tracer cloud.

[0098] The satellite remote sensing data-based cloud tracking method uses a two-way correlation coefficient method for matching. A successful match is considered achieved only if both the forward and reverse matching passes. This method verifies the single matching result and eliminates the influence of noise points, as shown in the example above where (m1,m1) are the forward matching points. ′ The fact that a match point was rejected and failed validation indicates that this set of matching points is likely noise. Using this bidirectional matching method can improve matching accuracy and ensure the accuracy of the matching calculation results.

[0099] In one or more optional embodiments of this application, a method for tracking cloud based on satellite remote sensing data is provided, wherein the step of matching the target window with the search window includes:

[0100] Calculate the correlation coefficient between multiple pixels in the target window and multiple pixels in the search window, and determine the matching point between the target window and the search window based on the correlation coefficient;

[0101] Wherein, the correlation coefficient is:

[0102]

[0103] ρ represents the correlation coefficient, g(c,r) represents the gray value of pixel (c,r), (c1,r1) represents the pixel coordinates of the target window, (c2,r2) represents the pixel coordinates of the search window, i and j are the pixel indices of the target window and the search window, respectively, and the values ​​of i and j are both in the range of 1 to n.

[0104] It should be noted that the method of this disclosure embodiment can be executed by a single device, such as a computer or server. The method of this embodiment can also be applied to a distributed scenario, where multiple devices cooperate to complete the task. In such a distributed scenario, one of these devices may execute only one or more steps of the method of this disclosure embodiment, and the multiple devices will interact with each other to complete the method described.

[0105] It should be noted that the above description describes some embodiments of this disclosure. 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.

[0106] Based on the same inventive concept, corresponding to any of the above embodiments, this disclosure also provides a tracking cloud device based on satellite remote sensing data.

[0107] refer to Figure 5 The tracking cloud device based on satellite remote sensing data includes:

[0108] Image acquisition module 501 is used to acquire a first satellite image and a second satellite image from high-resolution satellite remote sensing video data;

[0109] The similarity analysis module 502 is used to perform similarity analysis on the first satellite image and the second satellite image, determine the changed region in the second satellite image, and determine the original region in the first satellite image corresponding to the changed region, wherein the original region contains a tracer cloud;

[0110] Image enhancement module 503 is used to perform image enhancement transformation on the first satellite image and the second satellite image, and generate a first enhanced image and a second enhanced image accordingly; and

[0111] The target tracking module 504 is used to determine a cloud target in the changed region corresponding to the tracer cloud in the original region by performing image matching between the first enhanced image and the second enhanced image.

[0112] For ease of description, the above apparatus is described in terms of its functions, divided into various modules. Of course, in implementing this disclosure, the functions of each module can be implemented in one or more software and / or hardware.

[0113] The apparatus described above is used to implement the corresponding satellite remote sensing data-based tracking cloud method in any of the foregoing embodiments, and has the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0114] Based on the same inventive concept, corresponding to the methods of any of the above embodiments, this disclosure also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements the tracking cloud method based on satellite remote sensing data as described in any of the above embodiments.

[0115] Figure 6 This embodiment illustrates a more specific hardware structure of an electronic device, which may include a processor 1010, a memory 1020, an input / output interface 1030, a communication interface 1040, and a bus 1050. The processor 1010, memory 1020, input / output interface 1030, and communication interface 1040 are interconnected internally via the bus 1050.

[0116] The processor 1010 can be implemented using a general-purpose CPU (Central Processing Unit), microprocessor, application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this specification.

[0117] The memory 1020 can be implemented in the form of ROM (Read Only Memory), RAM (Random Access Memory), static storage device, dynamic storage device, etc. The memory 1020 can store the operating system and other applications. When the technical solutions provided in the embodiments of this specification are implemented by software or firmware, the relevant program code is stored in the memory 1020 and is called and executed by the processor 1010.

[0118] The input / output interface 1030 is used to connect input / output modules to realize information input and output. Input / output modules can be configured as components within the device (not shown in the figure) or externally connected to the device to provide corresponding functions. Input devices may include keyboards, mice, touchscreens, microphones, various sensors, etc., while output devices may include displays, speakers, vibrators, indicator lights, etc.

[0119] The communication interface 1040 is used to connect a communication module (not shown in the figure) to enable communication between this device and other devices. The communication module can communicate via wired means (such as USB, Ethernet cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.).

[0120] Bus 1050 includes a pathway for transmitting information between various components of the device, such as processor 1010, memory 1020, input / output interface 1030, and communication interface 1040.

[0121] It should be noted that although the above-described device only shows the processor 1010, memory 1020, input / output interface 1030, communication interface 1040, and bus 1050, in specific implementations, the device may also include other components necessary for normal operation. Furthermore, those skilled in the art will understand that the above-described device may only include the components necessary for implementing the embodiments of this specification, and not necessarily all the components shown in the figures.

[0122] The electronic devices described above are used to implement the corresponding satellite remote sensing data-based tracking cloud method in any of the foregoing embodiments, and have the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0123] Based on the same inventive concept, corresponding to the methods of any of the above embodiments, this disclosure also provides a non-transitory computer-readable storage medium storing computer instructions for causing the computer to execute the tracking cloud method based on satellite remote sensing data as described in any of the above embodiments.

[0124] The computer-readable medium of this embodiment includes permanent and non-permanent, removable and non-removable media, and information storage can be implemented by any method or technology. Information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transfer medium that can be used to store information accessible by a computing device.

[0125] The computer instructions stored in the storage medium of the above embodiments are used to cause the computer to execute the tracking cloud method based on satellite remote sensing data as described in any of the above embodiments, and have the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0126] Those skilled in the art should understand that the discussion of any of the above embodiments is merely exemplary and is not intended to imply that the scope of this disclosure (including the claims) is limited to these examples; within the framework of this disclosure, the technical features of the above embodiments or different embodiments can also be combined, the steps can be implemented in any order, and there are many other variations of different aspects of one or more embodiments of this application as described above, which are not provided in detail for the sake of brevity.

[0127] Additionally, to simplify the description and discussion, and to avoid obscuring one or more embodiments of this application, the well-known power / ground connections to integrated circuit (IC) chips and other components may or may not be shown in the provided drawings. Furthermore, the apparatus may be illustrated in block diagram form to avoid obscuring one or more embodiments of this application, and this also takes into account the fact that the details of implementation of these block diagram apparatuses are highly dependent on the platform on which one or more embodiments of this application will be implemented (i.e., these details should be fully understood by those skilled in the art). While specific details (e.g., circuits) have been set forth to describe exemplary embodiments of this disclosure, it will be apparent to those skilled in the art that one or more embodiments of this application may be implemented without these specific details or with variations thereof. Therefore, these descriptions should be considered illustrative rather than restrictive.

[0128] Although this disclosure has been described in conjunction with specific embodiments thereof, many substitutions, modifications, and variations of these embodiments will be apparent to those skilled in the art from the foregoing description. For example, other memory architectures (e.g., dynamic RAM (DRAM)) may be used with the embodiments discussed.

[0129] One or more embodiments of this application are intended to cover all such substitutions, modifications, and variations that fall within the broad scope of the appended claims. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of one or more embodiments of this application should be included within the protection scope of this disclosure.

Claims

1. A satellite remote sensing data-based cloud tracking method, characterized in that, The method comprises the following steps: obtaining a first satellite image and a second satellite image from high-resolution satellite remote sensing video data; determining a change region in the second satellite image and an original region corresponding to the change region in the first satellite image by performing similarity analysis on the first satellite image and the second satellite image, wherein the original region contains a tracer cloud; generating a first enhanced image and a second enhanced image by performing image enhancement transformation on the first satellite image and the second satellite image; determining a cloud target corresponding to the tracer cloud of the original region in the change region by performing image matching on the first enhanced image and the second enhanced image.

2. The method of claim 1, wherein, The method of determining a change region in the second satellite image by performing similarity analysis on the first satellite image and the second satellite image comprises the following steps: extracting grid feature data of the first satellite image and the second satellite image by using a convolutional neural network; calculating Mahalanobis distances of multiple image grids of the second satellite image compared with the first satellite image according to the grid feature data; comparing the Mahalanobis distances of the multiple image grids with a region segmentation threshold, and selecting multiple image grids with a Mahalanobis distance less than the region segmentation threshold to form the change region.

3. The method of claim 2, wherein, The method of determining the region segmentation threshold comprises the following steps: obtaining a reference sample image and a labeled sample image corresponding to the reference sample image; calculating Mahalanobis distances of a change region and an unchanged region of the labeled sample image compared with the reference sample image to generate a Mahalanobis distance histogram; determining the region segmentation threshold according to the Mahalanobis distance histogram.

4. The method of claim 1, wherein, The method of performing image enhancement transformation on the first satellite image and the second satellite image comprises the following steps: performing piecewise linear transformation on image gray scales of the first satellite image and the second satellite image; performing Gaussian low-pass filter transformation on frequency domain data corresponding to the first satellite image and the second satellite image.

5. The method of claim 4, wherein, The transformation formula of the piecewise linear transformation on the image gray scales is as follows: wherein h represents a pixel gray scale value after transformation, f(x, y) represents a gray scale value of a pixel (x, y) before transformation, L represents a maximum gray scale of an image, and a, b, c, and d represent gray scale segmentation values.

6. The method of claim 4, wherein, The method of performing Gaussian low-pass filter transformation on the frequency domain data corresponding to the first satellite image and the second satellite image comprises the following steps: performing filter transformation on the frequency domain data of the frequency domain data corresponding to the first satellite image and the second satellite image by using a two-dimensional Gaussian low-pass filter; the transfer function of the two-dimensional Gaussian low-pass filter is as follows: wherein D(u, v) represents a distance between a point (u, v) in a frequency spectrum image and a Fourier transform origin, and D0 represents a cutoff frequency.

7. The method of claim 1, wherein, The method of determining a cloud target corresponding to the tracer cloud of the original region in the change region by performing image matching on the first enhanced image and the second enhanced image comprises the following steps: determining a target window in the first enhanced image according to the original region, determining a search window in the second enhanced image according to the changed region, calling the target window to match the search window, and determining a forward matching point set; determining a target window in the second enhanced image according to the changed region, determining a search window in the first enhanced image according to the original region, calling the target window to match the search window, and determining a reverse matching point set; taking an intersection of the forward matching point set and the reverse matching point set as a successful matching point set; determining a plurality of target points corresponding to the tracer cloud of the original region in the changed region according to the successful matching point set, and determining the cloud target corresponding to the tracer cloud according to the plurality of target points.

8. The method of claim 7, wherein, The calling the target window to match the search window comprises: calculating a correlation coefficient between a plurality of pixels in the target window and a plurality of pixels in the search window, and determining a matching point between the target window and the search window according to the correlation coefficient; wherein the correlation coefficient is: wherein ρ represents the correlation coefficient, g(c, r) represents a gray value of a pixel (c, r), (c1, r1) represents a pixel coordinate of the target window, (c2, r2) represents a pixel coordinate of the search window, i and j are pixel indexes of the target window and the search window respectively, and the value ranges of i and j are both 1-n. comprising:

9. A satellite remote sensing data-based cloud tracking device, characterized by, an image acquisition module configured to acquire a first satellite image and a second satellite image from high-resolution satellite remote sensing video data; a similarity analysis module configured to determine a changed region in the second satellite image and an original region corresponding to the changed region in the first satellite image by performing similarity analysis on the first satellite image and the second satellite image, the original region containing a tracer cloud; an image enhancement module configured to perform image enhancement transformation on the first satellite image and the second satellite image, and correspondingly generate a first enhanced image and a second enhanced image; and a target tracking module configured to determine a cloud target corresponding to the tracer cloud of the original region in the changed region by performing image matching on the first enhanced image and the second enhanced image. The processor implements the method according to any one of claims 1-8 when executing the program. The processor implements the method according to any one of claims 1-8 when executing the program.

10. An electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, ​

Images