Method for identifying cloud and cloud-like ground objects in remote sensing image based on geometric parallax

Abstract

The invention belongs to the technical field of remote sensing image processing, and particularly discloses a geometric parallax-based method for identifying cloud and cloud-like ground objects in a remote sensing image, and the method comprises the steps: carrying out the coarse extraction of a cloud region based on the spectral features of each region in a target remote sensing image, and obtaining a cloud contour binary image; performing homonymous cloud point matching on the cloud contour binary image to obtain parallax values of a plurality of homonymous cloud regions; and clustering the parallax values of the plurality of homonymous cloud regions, and determining cloud and cloud-like ground features in the target remote sensing image. According to the method, the recognition precision of cloud and cloud-like ground objects can be improved.

Description

Technical Field

[0001] The present application belongs to the technical field of remote sensing image processing, and more specifically, relates to a method for identifying clouds and cloud-like objects in remote sensing images based on geometric parallax. Background Art

[0002] With the advancement of remote sensing, digital imaging, and space technology, the application of remote sensing satellites has penetrated into all aspects of national economic construction and social development, particularly in surveying and mapping, environmental monitoring, military strikes, disaster warning, and meteorological observation. Currently, there are over 4,000 satellites in orbit worldwide, but the utilization of optical remote sensing imagery has been limited, resulting in a significant waste of resources. A major factor in this is cloud cover. Global cloud cover data shows that over 60% of the Earth's surface is obscured by clouds, resulting in a significant loss of ground object information and a significant loss of the usefulness of satellite data. This inconveniences subsequent image radiometric processing, target recognition, and change detection.

[0003] Currently, most cloud detection methods rely solely on image features, such as cloud distribution, spectral characteristics, texture structure, and image features. These are significantly affected by cloud-like objects, limiting detection accuracy. Spectral thresholding, the most widely used method in engineering, is particularly effective. This method relies on remote sensing imagery rich in spectral information. It exploits cloud characteristics such as high reflectivity and low temperatures, applying multispectral physical properties to individual pixels for detection. However, this method is susceptible to cloud-like objects with similar spectral characteristics, such as bright artificial objects, snow, and ice. These cloud-like objects can easily be misidentified as clouds, resulting in low recognition accuracy. Summary of the Invention

[0004] In response to the shortcomings of the existing technology, the purpose of this application is to provide a method for identifying clouds and cloud-like objects in remote sensing images based on geometric parallax, aiming to solve the problem of low cloud recognition accuracy caused by the fact that most existing cloud detection methods are only based on the image features of the image.

[0005] To achieve the above objectives, in a first aspect, the present application provides a method for identifying clouds and cloud-like objects in remote sensing images based on geometric parallax, comprising: Based on the spectral characteristics of each area in the target remote sensing image, the cloud area is roughly extracted to obtain a cloud contour binary image; Matching cloud points with the same name on the cloud contour binary image to obtain disparity values ​​of multiple cloud areas with the same name; Clustering is performed on the disparity values ​​of the multiple cloud regions with the same name to determine clouds and cloud-like objects in the target remote sensing image.

[0006] Based on the physical characteristics of high cloud reflectivity and the differences in spatial positions of clouds and cloud-like objects, this application uses the spectral characteristics in remote sensing images to extract cloud contour binary images, further obtain cloud area parallax values ​​and cluster them. The spectral characteristics and geometric properties of remote sensing images can be combined to identify clouds and cloud-like objects. This can overcome the problem of low recognition accuracy caused by existing technologies that only use spectral characteristics for cloud detection, and improve the recognition accuracy of clouds and cloud-like objects.

[0007] According to a method for identifying clouds and cloud-like objects in remote sensing images based on geometric parallax provided by the present application, the cloud area is roughly extracted based on the spectral characteristics of each area in the target remote sensing image to obtain a cloud contour binary image, including: Based on the spectral characteristics of each area in the target remote sensing image, cloud detection is performed on the target remote sensing image based on a preset threshold to obtain a cloud detection result; Edge extraction is performed on the cloud detection result to obtain the cloud contour binary image.

[0008] According to a method for identifying clouds and cloud-like objects in remote sensing images based on geometric parallax provided by the present application, matching cloud points with the same name on the cloud contour binary image to obtain disparity values ​​of multiple cloud areas with the same name includes: Extracting cloud area edge image points connected at the beginning and the end in the cloud contour binary image; Centralizing the edge image points of the cloud area connected end to end to obtain the central image point of the cloud area; The disparity value of the central image point of the cloud region is calculated as the disparity value of the cloud region with the same name.

[0009] According to a method for identifying clouds and cloud-like objects in remote sensing images based on geometric parallax provided by the present application, the method further includes: The imaging errors of the forward and backward remote sensing images of the same area are corrected to obtain the refined rational polynomial coefficient (RPC) parameters.

[0010] This application first corrects the imaging errors of the front- and rear-view target remote sensing images of the same area, eliminates the imaging errors, and obtains refined RPC parameters for subsequent processing.

[0011] According to a method for identifying clouds and cloud-like objects in remote sensing images based on geometric parallax provided by the present application, the imaging errors of the front-view and back-view remote sensing images of the same area are corrected, including: Construct an imaging error correction model; Calculating imaging error parameters in the imaging error correction model; Based on the imaging error parameters, the imaging errors of the front- and back-view remote sensing images of the same area are corrected to obtain refined rational polynomial coefficient RPC parameters.

[0012] According to a method for identifying clouds and cloud-like objects in remote sensing images based on geometric parallax provided by the present application, the method further includes: Resample the front- and back-view target remote sensing images of the same area.

[0013] This application resamples the front and back view remote sensing images to unify the resolution of the images for subsequent processing.

[0014] In a second aspect, the present application provides a device for identifying clouds and cloud-like objects in remote sensing images based on geometric parallax, comprising: The coarse extraction module is used to perform coarse extraction of cloud areas based on the spectral characteristics of each area in the target remote sensing image to obtain a cloud contour binary image; A matching module is used to match cloud points with the same name on the cloud contour binary image to obtain disparity values ​​of multiple cloud areas with the same name; The recognition module is used to cluster the disparity values ​​of the multiple cloud areas with the same name to determine the clouds and cloud-like objects in the target remote sensing image.

[0015] In a third aspect, the present application provides an electronic device comprising: at least one memory for storing programs; and at least one processor for executing the programs stored in the memory. When the programs stored in the memory are executed, the processor is used to execute the method for identifying clouds and cloud-like objects in remote sensing images based on geometric parallax as described in the first aspect or any possible implementation of the first aspect.

[0016] In a fourth aspect, the present application provides a computer-readable storage medium, which stores a computer program. When the computer program runs on a processor, the processor executes the method for identifying clouds and cloud-like objects in remote sensing images based on geometric parallax described in the first aspect or any possible implementation of the first aspect.

[0017] In a fifth aspect, the present application provides a computer program product, which, when running on a processor, enables the processor to execute the method for identifying clouds and cloud-like objects in remote sensing images based on geometric parallax as described in the first aspect or any possible implementation of the first aspect.

[0018] It can be understood that the beneficial effects of the second to fifth aspects mentioned above can be found in the relevant description of the first aspect mentioned above, and will not be repeated here.

[0019] In general, the above technical solutions conceived by this application have the following beneficial effects compared with the existing technologies: Based on the physical characteristics of high cloud reflectivity and the differences in spatial positions of clouds and cloud-like objects, this application uses the spectral characteristics in remote sensing images to extract cloud contour binary images, further obtain cloud area parallax values ​​and cluster them. The spectral characteristics and geometric properties of remote sensing images can be combined to identify clouds and cloud-like objects. This can overcome the problem of low recognition accuracy caused by existing technologies that only use spectral characteristics for cloud detection, and improve the recognition accuracy of clouds and cloud-like objects. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solutions in the present application or the prior art, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0021] Figure 1 This is one of the flow charts of the method for identifying clouds and cloud-like objects in remote sensing images based on geometric parallax provided in an embodiment of the present application; Figure 2 This is the second flow chart of the method for identifying clouds and cloud-like objects in remote sensing images based on geometric parallax provided in an embodiment of the present application; Figure 3 1 is a schematic structural diagram of a device for identifying clouds and cloud-like objects in remote sensing images based on geometric parallax provided by an embodiment of the present application; Figure 4 It is a structural diagram of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0022] In order to make the purpose, technical solutions and advantages of this application more clear, the following further describes this application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0023] The term "and / or" as used herein describes an association between related objects, indicating that three possible relationships exist. For example, "A and / or B" can represent: A exists alone, A and B exist simultaneously, or B exists alone. The symbol " / " as used herein indicates that the related objects are in an "or" relationship, for example, A / B means either A or B.

[0024] In the embodiments of this application, words such as "exemplary" or "for example" are used to indicate examples, illustrations, or descriptions. Any embodiment or design described as "exemplary" or "for example" in the embodiments of this application should not be interpreted as being preferred or advantageous over other embodiments or designs. Rather, the use of words such as "exemplary" or "for example" is intended to present the relevant concepts in a concrete manner.

[0025] In the description of the embodiments of the present application, unless otherwise specified, "multiple" means two or more, for example, multiple processing units means two or more processing units, etc.; multiple elements means two or more elements, etc.

[0026] First, let’s introduce the following contents: With the advancement of remote sensing, digital imaging, and space technology, the application of remote sensing satellites has penetrated into all aspects of national economic construction and social development, particularly in surveying and mapping, environmental monitoring, military strikes, disaster warning, and meteorological observation. Currently, there are over 4,000 satellites in orbit worldwide, but the utilization of optical remote sensing imagery has been limited, resulting in a significant waste of resources. A major factor in this is cloud cover. Global cloud cover data shows that over 60% of the Earth's surface is obscured by clouds, resulting in a significant loss of ground object information and a significant loss of the usefulness of satellite data. This inconveniences subsequent image radiometric processing, target recognition, and change detection.

[0027] Therefore, cloud detection plays a crucial role in remote sensing image processing. Accurate and efficient cloud identification and detection in optical remote sensing images facilitates subsequent image processing and application. As an integral part of remote sensing image processing, cloud detection is a hot topic within the field. Advances in remote sensing, image processing, and computer technology have yielded numerous outstanding achievements that are well-suited to the changing times. Mainstream cloud detection methods can be broadly categorized into three categories: spectral threshold-based methods, classical machine learning-based methods, and deep learning-based methods.

[0028] However, most of these methods rely solely on image features, such as cloud distribution, spectral characteristics, texture structure, and image features. These methods are significantly affected by cloud-like objects, limiting detection accuracy. Spectral thresholding, the most widely used method in engineering, is particularly effective. This method relies on remote sensing imagery rich in spectral information. It exploits cloud characteristics such as high reflectivity and low temperatures, applying multispectral physical properties to individual pixels for detection. However, this method is susceptible to cloud-like objects with similar spectral characteristics, such as bright artificial objects, snow, and ice. These cloud-like objects can easily be misidentified as clouds, leading to incorrect detection.

[0029] In fact, compared to ordinary images, remote sensing images not only have ultra-high spatiotemporal resolution, but also contain relevant information such as geometric properties due to their unique imaging mechanism. Since clouds are generally much higher in elevation than other land features and are moving targets, their geographical location can change over a short period of time. This characteristic makes their geometric representation in remote sensing images significantly different from that of land features. Effectively utilizing this characteristic can significantly improve the accuracy of cloud detection.

[0030] In summary, in view of the fact that current cloud detection algorithms have difficulty in distinguishing cloud-like objects, and most of them only use multispectral image information as features while ignoring the geometric features in remote sensing information, this application provides a method for identifying clouds and cloud-like objects in remote sensing images based on geometric parallax. It fully combines the spectral characteristics and geometric properties of remote sensing images to distinguish between true cloud areas and cloud-like objects, effectively removes the influence of cloud-like objects, and provides an effective and feasible method for high-precision cloud detection in remote sensing images.

[0031] Next, combine Figure 1-Figure 2 The present invention introduces a method for identifying clouds and cloud-like objects in remote sensing images based on geometric parallax provided in an embodiment of the present application.

[0032] Figure 1 This is one of the flow charts of the method for identifying clouds and cloud-like objects in remote sensing images based on geometric parallax provided by the embodiment of the present application, such as Figure 1 As shown, the method includes the following steps: Step 100, based on the spectral characteristics of each area in the target remote sensing image, the cloud area is roughly extracted to obtain a cloud contour binary image; The basic idea of ​​this application is to first perform a rough extraction of the cloud area based on the physical characteristics of cloud high reflectivity to obtain a preliminary cloud area range, namely a cloud contour binary map.

[0033] Clouds vary widely in size, shape, and posture, but their spectral characteristics are particularly pronounced, manifesting as extremely high reflectivity in remote sensing images, often appearing as "highlights." This spectral signature allows for a rough extraction of cloud areas, and a binary image of the cloud outline is generated by extracting the cloud edges. This method is efficient and accurate, distinguishing clouds from most other ground objects and generating an initial cloud mask for subsequent processing.

[0034] Step 110 , matching cloud points with the same name on the cloud outline binary image to obtain disparity values ​​of multiple cloud areas with the same name; Then, the cloud points with the same name are matched and their object coordinates and disparity values ​​are calculated. Since clouds have a high elevation and may change their spatial position in a short period of time, there will be a large disparity in the cloud images at different viewing angles. However, ground objects usually have a low elevation and their spatial position remains basically unchanged in a short period of time, so their disparity is small.

[0035] Specifically, cloud points with the same name are matched in the cloud contour binary image, and cloud areas are divided according to the rules. Several cloud points with the same name in each connected cloud area are classified into the same category. The disparity of cloud points with the same name in the same category is calculated, and the disparity value of each cloud area is obtained through cluster analysis and mean calculation.

[0036] Step 120 : clustering the disparity values ​​of multiple cloud regions with the same name to determine clouds and cloud-like objects in the target remote sensing image.

[0037] Finally, the real cloud area and cloud-like area are distinguished by analyzing the parallax of the stereo image pairs of the rough cloud area extraction results, thereby eliminating the influence of cloud-like objects and obtaining refined cloud detection results.

[0038] Specifically, the disparity values ​​of multiple cloud areas with the same name are clustered, and the disparity threshold between clouds and cloud-like objects is obtained through cluster analysis. Areas with disparity values ​​less than the threshold are considered to be cloud-like objects, and areas with disparity values ​​greater than the threshold are considered to be real cloud areas.

[0039] The results of this step are combined with the initial rough extraction cloud detection results to obtain refined cloud detection results.

[0040] The method for identifying clouds and cloud-like objects in remote sensing images based on geometric parallax provided in this application is based on the physical characteristics of high cloud reflectivity and the differences in spatial position between clouds and cloud-like objects. It utilizes the spectral characteristics in remote sensing images to extract cloud outline binary images, further obtains and clusters cloud area parallax values, and can combine the spectral characteristics and geometric properties of remote sensing images to identify clouds and cloud-like objects. This method can overcome the low recognition accuracy problem caused by existing technologies that only detect clouds through spectral characteristics, thereby improving the recognition accuracy of clouds and cloud-like objects.

[0041] In some embodiments, step 100 specifically includes: Step 1001, based on the spectral characteristics of each area in the target remote sensing image and a preset threshold, cloud detection is performed on the target remote sensing image to obtain a cloud detection result; Step 1002: perform edge extraction on the cloud detection result to obtain a cloud contour binary image.

[0042] Optionally, the cloud area can be roughly extracted by the spectral threshold method. The spectral threshold method sets a threshold T, identifies pixels in the remote sensing image with grayscale values ​​greater than the T value as clouds, and performs edge extraction on the cloud detection results to obtain a preliminary cloud area contour binary map, i.e., the cloud detection result.

[0043] In some embodiments, step 110 specifically includes: Step 1101, extracting cloud edge image points connected at the beginning and end in the cloud contour binary image; Step 1102: Centering the edge image points of the cloud area connected end to end to obtain the center image point of the cloud area; Step 1103 , calculating the disparity value of the central image point of the cloud region as the disparity value of the cloud region with the same name.

[0044] Optionally, the preliminary cloud area contour binary images of the front and back images can be extracted by chain code encoding to obtain the cloud area edge image points connected at the beginning and end of the two images.

[0045] Chain code encoding is a method of describing a curve or boundary using the coordinates of the curve starting point and the direction code of the boundary point. It is an encoding method based on boundary tracking, which connects the pixel points on the boundary in sequence to form a chain code.

[0046] Optionally, edge image points of the cloud area connected end to end in the two images obtained may be centered using a mean value method to obtain a central image point of the cloud area, and the disparity value of the central image point of the cloud area may be calculated.

[0047] Since the actual object space coordinates of the cloud area are difficult to obtain, the image disparity value is converted into the difference of the object space coordinates of the central image point. Specifically, the object space coordinates of the central image point of the cloud area are calculated by combining the refined RPC parameters, and then the disparity value is calculated by the following formula:

[0048] in, 、 They represent the object space coordinates of the central image point of the cloud area in the fore and aft images respectively.

[0049] In some embodiments, the method further comprises: Step 130 , correcting imaging errors of the front- and rear-view target remote sensing images of the same area.

[0050] Figure 2 This is the second method for identifying clouds and cloud-like objects in remote sensing images based on geometric parallax provided by the embodiment of the present application, such as Figure 2 As shown in Figure 1, before the cloud area is roughly extracted, the imaging errors of the front- and back-view target remote sensing images of the same area are first corrected to eliminate the imaging errors and obtain refined RPC parameters for subsequent processing.

[0051] Because satellite position and sensor attitude angle observations often contain systematic errors, primarily constant errors, the rigorous geometric processing model established based on these position and attitude parameters inevitably contains systematic errors. Consequently, the virtual control grid generated using the rigorous geometric processing model and the RPC parameters calculated using the terrain-independent solution are bound to contain systematic errors. Furthermore, errors caused by sensor lens distortion and other unknown factors will also propagate into the RPC parameters as the rigorous geometric processing model is established.

[0052] In some embodiments, step 130 specifically includes: Construct an imaging error correction model; Calculate imaging error parameters in the imaging error correction model; The imaging errors of the forward and backward remote sensing images of the same area are corrected based on the imaging error parameters.

[0053] First, an imaging error correction model is constructed. Specifically, when the imaging error is corrected using the image space correction scheme, the relationship between the image point coordinates (R, C) and the ground point coordinates (X, Y, Z) is corrected to:

[0054] in, is the normalized coordinate of the ground point (X, Y, Z) after translation and scaling, and its value is between [-1, 1]; each polynomial P i ( i =1, 2, 3, 4) for each coordinate component The maximum power of is no more than 3, and the sum of the powers of each coordinate component of each item is no more than 3. For example, its specific form is shown in the following formula.

[0055]

[0056] Where, It is the RPC parameter.

[0057] The purpose of using regularized coordinates is to improve the stability of the solution of each coefficient in the model and reduce the data rounding error caused by the large data difference during the calculation process. The regularization formula is:

[0058]

[0059] Where, X o , Y o , Z o , R o , C o is the regularized translation parameter; X s , Y s , Z s , R s ,C s is the regularization proportional coefficient, which can be calculated by the following formula.

[0060]

[0061]

[0062] in, m is the number of control points.

[0063]

[0064]

[0065] ( , ) is the imaging error correction value of the image point coordinate (R, C), and its calculation formula is:

[0066] in, is the imaging error correction parameter.

[0067] When the imaging error correction amount ( , ) is taken to the first order, the imaging error correction model is the affine transformation model:

[0068] Alternatively, the imaging error parameters can be solved by the adjustment method. Specifically, if the affine transformation coefficients in the imaging error correction model and the object space coordinates of the target point are taken as unknowns, the bundle adjustment error equation matrix can be obtained as follows:

[0069] in, is the residual vector of the row and column coordinate observation values ​​of the image point; is the incremental vector of the image point coordinate system error correction parameter; ; ,in are the row and column coordinates of the image point calculated using the approximate value of the unknown number; , which is expanded as follows:

[0070]

[0071]

[0072]

[0073]

[0074]

[0075] in,

[0076] Among them, the partial derivative values ​​obtained by taking polynomial P1 as an example are:

[0077] in, It is the RPC parameter.

[0078] Two error equations can be established for each image point. When enough image points are measured, the normal equation can be formed:

[0079] According to the least squares adjustment principle, the unknowns can be obtained:

[0080] After obtaining the system error correction parameters for each scene image, the system error can be further integrated into the RPC parameters to obtain refined RPC parameters for use in subsequent steps.

[0081] In some embodiments, the method further comprises: Resample the front- and back-view target remote sensing images of the same area.

[0082] like Figure 2 As shown in FIG, the remote sensing images of the foreground and background targets in the same area are resampled to unify the resolution of the images for easy subsequent processing.

[0083] Figure 3 is a schematic diagram of the structure of a device for identifying clouds and cloud-like objects in remote sensing images based on geometric parallax, as provided in an embodiment of the present application. Figure 3 As shown, the system includes a coarse extraction module 310, a matching module 320 and a recognition module 330, wherein: A coarse extraction module 310 is used to perform coarse extraction of cloud areas based on the spectral characteristics of each area in the target remote sensing image to obtain a cloud contour binary image; Matching module 320, used to match cloud points with the same name on the cloud outline binary image to obtain disparity values ​​of multiple cloud areas with the same name; The identification module 330 is used to cluster the disparity values ​​of multiple cloud areas with the same name to determine the clouds and cloud-like objects in the target remote sensing image.

[0084] Based on the method in the above embodiment, Figure 4 An example of a physical structure diagram of an electronic device is shown below. Figure 4 As shown, an embodiment of the present application provides an electronic device, which may include: a processor 410, a communication interface 420, a memory 430, and a communication bus 440, wherein the processor 410, the communication interface 420, and the memory 430 communicate with each other via the communication bus 440. The processor 410 can call the logic instructions in the memory 430 to execute the method for identifying clouds and cloud-like objects in remote sensing images based on geometric parallax in the above embodiment.

[0085] In addition, the logic instructions in the aforementioned memory 430 can be implemented in the form of a software functional unit and, when sold or used as an independent product, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the portion that contributes to the prior art, or the portion of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method for identifying clouds and cloud-like objects in remote sensing images based on geometric parallax as described in various embodiments of the present application.

[0086] Based on the method in the above embodiment, an embodiment of the present application provides a computer-readable storage medium, which stores a computer program. When the computer program runs on a processor, the processor executes the method for identifying clouds and cloud-like objects in remote sensing images based on geometric parallax in the above embodiment.

[0087] Based on the method in the above embodiment, an embodiment of the present application provides a computer program product. When the computer program product runs on a processor, the processor executes the method for identifying clouds and cloud-like objects in remote sensing images based on geometric parallax in the above embodiment.

[0088] It is understood that the processor in the embodiments of the present application may be a central processing unit (CPU), other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field programmable gate arrays (FPGA), other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. The general-purpose processor may be a microprocessor or any conventional processor.

[0089] The method steps in the embodiments of the present application can be implemented by hardware or by a processor executing software instructions. The software instructions can be composed of corresponding software modules, which can be stored in random access memory (RAM), flash memory, read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), registers, hard disks, mobile hard disks, CD-ROMs, or any other form of storage medium known in the art. An exemplary storage medium is coupled to the processor so that the processor can read information from the storage medium and write information to the storage medium. Of course, the storage medium can also be an integral part of the processor. The processor and the storage medium can be located in an ASIC.

[0090] The above embodiments can be implemented in whole or in part through software, hardware, firmware, or any combination thereof. When implemented using software, they can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When loaded and executed on a computer, the computer program instructions fully or partially produce the processes or functions described in the embodiments of this application. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted via the computer-readable storage medium. The computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that can be accessed by a computer, or a data storage device such as a server or data center that integrates one or more available media. The available medium can be magnetic media (e.g., floppy disk, hard disk, tape), optical media (e.g., DVD), or semiconductor media (e.g., solid-state drive (SSD)).

[0091] It will be understood that the various numerical numbers involved in the embodiments of the present application are merely distinctions for the convenience of description and are not intended to limit the scope of the embodiments of the present application.

[0092] It is easy for those skilled in the art to understand that the above is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present application should be included in the scope of protection of the present application.

Claims

1. A method for identifying clouds and cloud-like objects in remote sensing images based on geometric parallax, characterized in that: include: Based on the spectral characteristics of each area in the target remote sensing image, the cloud area is roughly extracted to obtain a cloud contour binary image; Matching cloud points with the same name on the cloud contour binary image to obtain disparity values ​​of multiple cloud areas with the same name; Clustering is performed on the disparity values ​​of the multiple cloud regions with the same name to determine clouds and cloud-like objects in the target remote sensing image.

2. The method for identifying clouds and cloud-like objects in remote sensing images based on geometric parallax according to claim 1, characterized in that: The method of performing rough extraction of cloud areas based on the spectral characteristics of each area in the target remote sensing image to obtain a cloud contour binary image includes: Based on the spectral characteristics of each area in the target remote sensing image, cloud detection is performed on the target remote sensing image based on a preset threshold to obtain a cloud detection result; Edge extraction is performed on the cloud detection result to obtain the cloud contour binary image.

3. The method for identifying clouds and cloud-like objects in remote sensing images based on geometric parallax according to claim 1, characterized in that: The matching of cloud points with the same name on the cloud contour binary image to obtain disparity values ​​of multiple cloud areas with the same name includes: Extracting cloud area edge image points connected at the beginning and the end in the cloud contour binary image; Centralizing the edge image points of the cloud area connected end to end to obtain the central image point of the cloud area; The disparity value of the central image point of the cloud region is calculated as the disparity value of the cloud region with the same name.

4. The method for identifying clouds and cloud-like objects in remote sensing images based on geometric parallax according to claim 1, characterized in that: The method further comprises: The imaging errors of the forward and backward remote sensing images of the same area are corrected to obtain the refined rational polynomial coefficients (RPC) parameters.

5. The method for identifying clouds and cloud-like objects in remote sensing images based on geometric parallax according to claim 4, characterized in that: The imaging errors of the front- and rear-view target remote sensing images of the same area are corrected to obtain the refined rational polynomial coefficient RPC parameters, including: Construct an imaging error correction model; Calculating imaging error parameters in the imaging error correction model; Based on the imaging error parameters, the imaging errors of the front- and back-view remote sensing images of the same area are corrected to obtain refined rational polynomial coefficient RPC parameters.

6. The method for identifying clouds and cloud-like objects in remote sensing images based on geometric parallax according to claim 1, characterized in that: The method further comprises: Resample the front- and back-view target remote sensing images of the same area.

7. A device for identifying clouds and cloud-like objects in remote sensing images based on geometric parallax, characterized in that: include: The coarse extraction module is used to perform coarse extraction of cloud areas based on the spectral characteristics of each area in the target remote sensing image to obtain a cloud contour binary image; A matching module is used to match cloud points with the same name on the cloud contour binary image to obtain disparity values ​​of multiple cloud areas with the same name; The recognition module is used to cluster the disparity values ​​of the multiple cloud areas with the same name to determine the clouds and cloud-like objects in the target remote sensing image.

8. An electronic device, characterized in that: include: at least one memory for storing a computer program; At least one processor is used to execute the program stored in the memory. When the program stored in the memory is executed, the processor is used to execute the method for identifying clouds and cloud-like objects in remote sensing images based on geometric parallax as described in any one of claims 1 to 6.

9. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed on a processor, the processor is enabled to execute the method for identifying clouds and cloud-like objects in remote sensing images based on geometric parallax according to any one of claims 1 to 6.

10. A computer program product, characterized in that When the computer program product runs on a processor, the processor is enabled to execute the method for identifying clouds and cloud-like objects in remote sensing images based on geometric parallax according to any one of claims 1 to 6.

Images