Remote sensing monitoring method based on AI intelligent guidance and related equipment
Through the AI-enabled remote sensing monitoring method, deep learning technology and human-machine collaborative intelligent interpretation, the problems of low production efficiency and low automation in remote sensing monitoring are solved, and efficient identification and organization of change patterns are achieved, which improves the Monitor accuracy and efficiency.
Patent Information
- Application Number
- CN202411061075.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-05
- Publication Date
- 2024-11-15
AI Technical Summary
There are problems in the existing remote sensing monitoring business with low production efficiency, low degree of automation, and poor model adaptability, mainly due to the high labor intensity, low efficiency of manual drawing methods and the lack of highly customized and automated operating processes.
Using AI-enabled remote sensing monitoring method, the data ready processing and preset classification recognition of changing maps is generated by collecting initial satellite remote sensing image data, using deep learning technology to perform preliminary screening and review and editing of maps to generate those that meet the preset technical requirements. The target result file combines geographical partitioning, time phase and topographical and topographic classification to automatically identify and organize changing maps to realize intelligent interpretation of human-machine collaboratively.
The production efficiency of remote sensing monitoring is improved, the degree of automation is enhanced, manual operations is reduced, model adaptability and recognition accuracy is improved, labor intensity and misidentification rate are reduced, and an efficient and intelligent monitoring process is formed.
Smart Images

Figure CN118968338A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of AI intelligent technology, and in particular to an AI intelligent-guided remote sensing monitoring method and related equipment. Background Art
[0002] Remote sensing monitoring is an important means of supervision and law enforcement, land change investigation, farmland protection monitoring, ecological protection monitoring and other businesses. It is widely used in the fields of natural resources, ecological environment, agriculture and rural areas, urban construction, emergency support, etc.
[0003] Artificial intelligence technology, especially deep learning technology, has developed rapidly in recent years and has been widely used in the field of remote sensing, driving remote sensing monitoring to develop in the direction of intelligence. However, in actual business, the operation mode is still a manual carpet monitoring mode, and artificial intelligence has not played its due value. There are still some problems that need to be solved. One is that the advanced artificial intelligence technology is not fully utilized to assist the operation. The main operation mode is still fishing net-style manual delineation, that is, the remote sensing image is divided into fishing net-like operation areas, and then completed by visual interpretation and manual delineation. This method is labor-intensive, inefficient, and highly subjective; second, there is a lack of highly customized and automated operation processes. All links in today's production operations, from image selection to task distribution, to delineation and results collation, require specific operations and implementation by professionals, and the connection between each link is not smooth enough. Therefore, it is necessary to propose an AI-guided remote sensing monitoring method to at least solve some of the above problems. Summary of the invention
[0004] A series of simplified concepts are introduced in the Summary of the Invention section, which will be further described in detail in the Detailed Description of the Invention section. The Summary of the Invention section of this application does not mean to attempt to define the key features and essential technical features of the claimed technical solution, nor does it mean to attempt to determine the scope of protection of the claimed technical solution.
[0005] This application specifically includes the following aspects:
[0006] In the first aspect, the present application proposes an AI intelligent guided remote sensing monitoring method, comprising:
[0007] Collecting initial satellite remote sensing image data, and performing data-ready processing on the initial satellite remote sensing image data to obtain target satellite remote sensing image data; the initial satellite remote sensing image data is the satellite remote sensing image data of the base period and the current period;
[0008] Identify the change spots in the target satellite remote sensing image data according to preset classifications, and perform job distribution operations based on the number of image slices or the number of change spots to obtain multiple change spot sets; the preset classifications include geographic division classification, temporal phase classification, topographic classification, and model type classification;
[0009] Performing preliminary screening and spot review and editing on the change spots in each of the change spot sets to obtain a target change spot set; the preliminary screening is to eliminate misidentified change spots and change spots irrelevant to the operation; the spot review and editing includes delineating the spot boundaries and filling in image-related attributes;
[0010] The target change pattern set is collated and checked to obtain a target result file that meets the preset technical requirements.
[0011] In a feasible implementation manner, initial satellite remote sensing image data is collected, and data-ready processing is performed on the initial satellite remote sensing image data to obtain target satellite remote sensing image data, including:
[0012] Draw image mosaic lines based on the satellite remote sensing image coverage each quarter and generate map files;
[0013] Generate an operation area grid according to the scope of the mission area and the acquisition of the initial satellite remote sensing image data;
[0014] In combination with the image mosaic lines and the work area grid, a slicing tool is used to cut the current and base period images to obtain image slices; the image slices are the target satellite remote sensing image data.
[0015] In a feasible implementation, in the geographical zoning classification, Northeast China, North China, East China, Central China, South China, Southwest China, and Northwest China are used for geographical zoning; in the phase classification, four seasons of a year are used for phase classification; in the landform classification, plateaus, mountains, plains, hills, and basins are used for landform classification; in the model type classification, small models, medium models, and large models are divided; different geographical divisions, different seasons, and different landforms are respectively set with different models to identify regional change elements; different models correspond to different recognition accuracy and running speeds.
[0016] In a feasible implementation, the change spots in each of the change spot sets are preliminarily screened, including:
[0017] Using a single change pattern as a calculation unit mask, respectively calculating the texture features inside the mask of the previous and next phase images; the texture features include contrast, correlation and entropy;
[0018] averaging the texture features of the before and after phase images within the mask to obtain object-level texture features;
[0019] The confidence of the change spots is calculated by using the object-level texture features, so as to eliminate the misidentified change spots and the change spots irrelevant to the operation based on the values of the confidence.
[0020] In a feasible implementation, the brightness difference between adjacent pixels in the image is determined by the contrast; the spatial consistency or correlation of pixel values in the image is determined by the correlation; and the uncertainty of pixel value distribution in the image is determined by the entropy.
[0021] In a feasible implementation manner, the target change pattern set is sorted out, including:
[0022] Fill in the relevant attributes of the target change map to form county-level results;
[0023] For the target change pattern, the image blocks of the base period and the current period are respectively intercepted, and the monitoring pattern screenshot image file is produced for the target change pattern by using automatic processing technology;
[0024] The monitoring spot screenshot image file generates corresponding metadata to organize all the results into a result file in a standard format.
[0025] In a feasible implementation manner, after obtaining the achievement file, obtaining the target achievement file by checking the achievement file specifically includes:
[0026] Check the target change spots in the achievement file, remove misidentified change spots and change spots irrelevant to the operation, and form the final achievement of the change spots;
[0027] The final result is automatically checked to obtain the target result file that meets the preset technical requirements.
[0028] In the second aspect, the present application proposes an AI intelligent guided remote sensing monitoring system, comprising:
[0029] A data-ready processing module is used to collect initial satellite remote sensing image data and perform data-ready processing on the initial satellite remote sensing image data to obtain target satellite remote sensing image data; the initial satellite remote sensing image data is the satellite remote sensing image data of the base period and the current period;
[0030] The change spot automatic identification module is used to identify the change spots in the target satellite remote sensing image data according to preset classifications, and perform job distribution operations based on the number of image slices or the number of change spots to obtain multiple change spot sets; the preset classifications include geographical division classification, temporal phase classification, topographic classification and model type classification;
[0031] The knowledge-assisted spot confirmation module is used to perform preliminary screening and spot review and editing on the change spots in each of the change spot sets to obtain a target change spot set; the preliminary screening is to eliminate misidentified change spots and change spots irrelevant to the operation; the spot review and editing includes the delineation of spot boundaries and the filling of image-related attributes;
[0032] The result sorting and checking module is used to sort and check the target change pattern set to obtain a target result file that meets the preset technical requirements.
[0033] In the third aspect, the present application proposes an electronic device, comprising: a memory, a processor, and a computer program stored in the above-mentioned memory and executable on the above-mentioned processor, wherein the above-mentioned processor is used to implement the steps of the AI intelligently guided remote sensing monitoring method as any one of the above-mentioned first aspects when executing the computer program stored in the memory.
[0034] In a fourth aspect, the present application further proposes a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the AI intelligently guided remote sensing monitoring method of any one of the first aspects is implemented.
[0035] In summary, the AI intelligently guided remote sensing monitoring method of the embodiment of the present application includes: collecting initial satellite remote sensing image data, and performing data-ready processing on the above-mentioned initial satellite remote sensing image data to obtain target satellite remote sensing image data; the above-mentioned initial satellite remote sensing image data is the satellite remote sensing image data of the base period and the current period; identifying the change spots in the above-mentioned target satellite remote sensing image data according to the preset classification, and performing job distribution operations based on the number of image slices or the number of change spots to obtain multiple change spot sets; the above-mentioned preset classification includes geographical zoning classification, time phase classification, topography classification and model type classification; performing preliminary screening and spot review and editing on the change spots in each of the above-mentioned change spot sets to obtain a target change spot set; the above-mentioned preliminary screening is to eliminate misidentified change spots and change spots that are not related to the job; the above-mentioned spot review and editing includes the delineation of spot boundaries and the filling of image-related attributes; the above-mentioned target change spot set is collated and checked for results to obtain a target result file that meets the preset technical requirements. The AI intelligent-guided remote sensing monitoring method proposed in the embodiment of the present application utilizes human-machine collaborative intelligent interpretation technology to solve the problems faced by remote sensing monitoring, such as low production efficiency, low degree of automation, and low model adaptability.
[0036] The AI intelligent-guided remote sensing monitoring method proposed in this application and other advantages, objectives and features of this application will be partially reflected through the following description, and will also be partially understood by technical personnel in this field through research and practice of this application. BRIEF DESCRIPTION OF THE DRAWINGS
[0037] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present specification. Also, the same reference symbols are used throughout the accompanying drawings to represent the same components. In the accompanying drawings:
[0038] Figure 1 A schematic diagram of a process of remote sensing monitoring method guided by AI provided in an embodiment of the present application;
[0039] Figure 2 A schematic diagram of the functional modules of an AI intelligently guided remote sensing monitoring system provided in an embodiment of the present application;
[0040] Figure 3 A schematic diagram of the structure of an AI intelligently guided remote sensing monitoring electronic equipment provided in an embodiment of the present application. DETAILED DESCRIPTION
[0041] In order to better understand the technical solutions provided by the embodiments of this specification, the technical solutions of the embodiments of this specification are described in detail below through the accompanying drawings and specific embodiments. It should be understood that the embodiments of this specification and the specific features in the embodiments are detailed descriptions of the technical solutions of the embodiments of this specification, rather than limitations on the technical solutions of this specification. In the absence of conflict, the embodiments of this specification and the technical features in the embodiments can be combined with each other.
[0042] In this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply that there is any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that the process, method, article or equipment including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method, article or equipment. In the absence of more restrictions, the elements limited by the statement "comprise one..." do not exclude the existence of other identical elements in the process, method, article or equipment including the elements. The term "more than two" includes two or more than two situations.
[0043] This application provides an AI-guided remote sensing monitoring method, which uses deep learning, human-computer collaboration and other technologies to develop a human-computer collaborative intelligent interpretation system to solve the problems of low production efficiency, low degree of automation, and low model adaptability in existing remote sensing monitoring services.
[0044] See also Figure 1 , which is a schematic diagram of a process of remote sensing monitoring method guided by AI provided in an embodiment of the present application, which may specifically include:
[0045] S110, collecting initial satellite remote sensing image data, and performing data-ready processing on the initial satellite remote sensing image data to obtain target satellite remote sensing image data; the initial satellite remote sensing image data is the satellite remote sensing image data of the base period and the current period.
[0046] Exemplarily, the base period and current period image data are analyzed, an image mapping file is prepared, an operation grid is generated according to the operation scope, and the two period images are sliced.
[0047] S120. Identify the change spots in the target satellite remote sensing image data according to preset classifications, and perform job distribution operations based on the number of image slices or the number of change spots to obtain multiple change spot sets; the preset classifications include geographic zoning classification, temporal classification, topographic classification, and model type classification.
[0048] Exemplarily, a model library for remote sensing monitoring services is constructed based on geographical division, time phase, topography, and model type. Specifically, the geographical division adopts seven major geographical divisions: Northeast China, North China, East China, Central China, South China, Southwest China, and Northwest China; the time phase adopts the classification method of four seasons in a year; the topography is classified into five major topography: plateau, mountain, plain, hill, and basin; the model type is divided into small models (such as Tiny-CD), medium models (such as SNUNet-CD), and large models (such as TransUNet-CD). After the intelligent change discovery is completed, the distribution of tasks is automatically completed based on the number of image blocks or the number of change spots.
[0049] S130, performing preliminary screening and spot review and editing on each of the change spots in the change spot set to obtain a target change spot set; the preliminary screening is to eliminate misidentified change spots and change spots irrelevant to the operation; the spot review and editing includes outlining the spot boundaries and filling in image-related attributes.
[0050] For example, the changes in two phases of images are compared, the type of change is comprehensively judged, and intelligent guided boundary delineation of the map spots is carried out, and key attributes such as the type before the change, the type after the change, and the name of the ecological area are filled in.
[0051] Furthermore, SAM (SegmentAnything Model) is used for interactive feature extraction. SAM has strong zero-sample learning and generalization capabilities, which can not only improve the extraction efficiency, but also ensure the accuracy of the extraction results, especially the boundaries, through artificial interactive prompt learning.
[0052] S140, organizing and checking the target change pattern set to obtain a target result file that meets preset technical requirements.
[0053] In summary, the AI intelligently guided remote sensing monitoring method proposed in the embodiment of the present application includes: analyzing the base period and current image data, making an image mapping file, generating an operation grid according to the operation range, and slicing the two-phase image; automatically identifying the change spots of the two-phase image to obtain suspected change spots, and then obtaining the circumscribed rectangle of the change spots required for the business through threshold filtering, vector processing, etc., and automatically completing the distribution of tasks based on the number of image blocks or the number of change spots; through human-computer collaboration, quickly screening and eliminating pseudo-change spots and business-irrelevant spots, performing intelligently guided boundary delineation and filling in key attributes of spots; organizing and checking the results obtained, and automatically checking the final results.
[0054] In some examples, collecting initial satellite remote sensing image data and performing data-ready processing on the initial satellite remote sensing image data to obtain target satellite remote sensing image data includes:
[0055] Draw image mosaic lines based on the satellite remote sensing image coverage each quarter and generate map files;
[0056] Generate an operation area grid according to the scope of the mission area and the acquisition of the initial satellite remote sensing image data;
[0057] In combination with the image mosaic lines and the work area grid, a slicing tool is used to cut the current and base period images to obtain image slices; the image slices are the target satellite remote sensing image data.
[0058] For example, image mosaic lines are drawn based on the coverage of satellite remote sensing images each quarter, and a map file is generated. When drawing mosaic lines, images that meet the phase requirements, have no cloud or snow coverage, and have good image quality should be given priority. When the same area is covered by images of multiple resolutions, images with higher resolutions should be given priority in principle. When high-quality images are already available for the same area, later high-quality images will not be considered.
[0059] Furthermore, according to the scope of the mission area and the image acquisition situation, the mission area grid is generated. The grid size is recommended to be set to an integer multiple of 1024. When the image results span less than three projection zones, it is recommended to select the projection zone with the largest number of images as the projection zone of the grid vector. When the work area spans more than three projection zones, it can be divided into multiple projection zones to generate grid vectors. If the mission area is not adjusted, the mission area grid does not need to be generated repeatedly within a year, and the grid of the previous quarter can be used in the next quarter.
[0060] Finally, the current and base period images are cropped using the slicing tool in combination with the mission area grid and the image mosaic line. If the grid has not been changed, there is no need to repeat the cropping if the base period image has been cropped in the previous quarter, and only the current image needs to be cropped.
[0061] In some examples, in the geographic zoning classification, Northeast China, North China, East China, Central China, South China, Southwest China, and Northwest China are used for geographical zoning; in the phase classification, four seasons of a year are used for phase classification; in the landform classification, plateaus, mountains, plains, hills, and basins are used for landform classification; in the model type classification, small models, medium models, and large models are divided; different geographical divisions, different seasons, and different landforms are respectively set with different models to identify regional change elements; different models correspond to different recognition accuracy and running speeds.
[0062] Exemplarily, the geographical division adopts seven major geographical divisions, namely Northeast China, North China, East China, Central China, South China, Southwest China, and Northwest China. Since there are great differences in the main crop planting types, architectural styles and other factors in different geographical divisions, it is necessary to set up models separately to identify the changing factors; the time phase adopts the classification method of four seasons a year. Taking into account the large time span of winter and summer, the winter phase is set to December-March, and the summer phase is designed to be June-September, so the spring phase is April-May, and the autumn phase is October-November. The main reason for using quarters as the time phase classification method is that crops have a certain phenological period, and the image characteristics such as color and texture of cultivated land areas in different seasons are quite different. For example, most cultivated land areas in the north are green in summer, while cultivated land in autumn is mainly yellow. If the same model is used for cultivated land The changing elements of the regional area are prone to misidentification; the landforms are classified into five major landforms: plateau, mountain, plain, hill, and basin. There are certain differences in the land use types and changes in common land use types of different landforms, so it is necessary to set up corresponding models for identification; the model types are divided into small models (such as Tiny-CD), medium models (such as SNUNet-CD) and large models (such as TransUNet-CD). The three models have certain differences in recognition accuracy and running speed. After testing on the WHU building change detection public dataset, the IOU accuracy evaluation index of the large model is about 83%, the IOU accuracy evaluation index of the medium model is about 79%, and the IOU accuracy evaluation index of the small model is about 75%.
[0063] Furthermore, when identifying change patches, the accuracy of the model automatically selected by AI is usually based on the following key factors: data quality and characteristics: the AI system will evaluate the quality of the data, including image resolution, spectral characteristics, noise level, etc., to determine the appropriate model accuracy; change type: different types of changes (such as building expansion, vegetation cover changes, etc.) may require models of different accuracy to accurately identify; prior knowledge: AI can use historical data and prior knowledge to predict the recognition effect of different models on specific types of changes; model performance indicators: AI will consider performance indicators such as model accuracy, recall rate, F1 score, and their performance on verification data; computing resources: AI will evaluate the available computing resources and select the model that can achieve the best performance under resource constraints; training time: AI It will weigh the relationship between model training time and accuracy, and select a model that achieves the required accuracy within a reasonable time; Model complexity: AI will analyze the complexity of the model and select an appropriate model based on the complexity of the data; Automated feature selection: AI can automatically select the features that are most helpful in identifying change patches, reducing unnecessary computational burden; Cross-validation: AI may use techniques such as cross-validation to evaluate the performance of different models on different subsets to select the most robust model; User feedback: In some cases, the AI system may adjust the model selection strategy based on user feedback; Model update: The AI system may continuously update and optimize the model selection strategy based on new data and feedback; Cost-benefit analysis: AI will consider the cost-effectiveness of model selection and select the model that achieves the best balance between cost and accuracy.
[0064] In some examples, performing a preliminary screening on the change spots in each of the change spot sets includes:
[0065] Using a single change pattern as a calculation unit mask, respectively calculating the texture features inside the mask of the previous and next phase images; the texture features include contrast, correlation and entropy;
[0066] averaging the texture features of the before and after phase images within the mask to obtain object-level texture features;
[0067] The confidence of the change spots is calculated by using the object-level texture features, so as to eliminate the misidentified change spots and the change spots irrelevant to the operation based on the values of the confidence.
[0068] Exemplarily, a single change pattern is first used as a calculation unit mask, and three texture features inside the image mask before and after are calculated respectively, namely, contrast, correlation, and entropy.
[0069] The three texture features are extracted through the gray symbiotic matrix (GLCM), and the calculation formula is as follows:
[0070]
[0071] Among them, i, j are the row and column numbers of the image, P(i, j) is the gray-level co-occurrence matrix, Mean is the mean of the image pixel values, and Variance is the variance of the image pixel values.
[0072] After the feature calculation, the average of each feature of the front and back images is calculated inside the mask, thus completing the acquisition of three sets of object-level features.
[0073] Furthermore, the difference between the three groups of object-level features of the before and after images is determined. If the feature difference is too small, it means that there is a high probability that the mask area of the before and after images has not changed. The confidence calculation formula is as follows:
[0074]
[0075] Among them, A and B are the before and after images, there are n object-level confidence features in total, and i is the i-th feature. The higher the confidence C, the higher the probability of change. After the confidence calculation of all vector patches is completed, the patches with low confidence are automatically screened and eliminated.
[0076] In some examples, the contrast is used to determine the brightness difference between adjacent pixels in the image; the correlation is used to determine the spatial consistency or correlation of pixel values in the image; and the entropy is used to determine the uncertainty of the distribution of pixel values in the image.
[0077] For example, contrast measures the brightness difference between adjacent pixels in an image, correlation reflects the spatial consistency or correlation of pixel values in an image, and entropy reflects the uncertainty of the distribution of pixel values in an image. The principle of selecting texture features as the features used in confidence calculation is that obviously misidentified change spots are usually due to large differences in the tones of the previous and subsequent images. For example, the same wheat field area appears green in spring and golden in summer. This area is easily identified by AI as a change area, and such situations can be effectively removed by filtering with texture features.
[0078] In some examples, the target change pattern set is collated, including:
[0079] Fill in the relevant attributes of the target change map to form county-level results;
[0080] For the target change pattern, the image blocks of the base period and the current period are respectively intercepted, and the monitoring pattern screenshot image file is produced for the target change pattern by using automatic processing technology;
[0081] The monitoring spot screenshot image file generates corresponding metadata to organize all the results into a result file in a standard format.
[0082] For example, fill in the administrative region code, administrative region name, phase, scene number and other attributes of the change map spot. If the map spot crosses administrative boundaries, it will be sorted according to the principle of the largest area to form the county results. Then, for each change map spot, the base period and current image blocks are captured separately, and the monitoring map spot screenshot image file is automatically created, metadata is generated, and the results are organized into a standard format.
[0083] In some examples, after obtaining the achievement file, obtaining the target achievement file by checking the achievement file specifically includes:
[0084] Check the target change spots in the achievement file, remove misidentified change spots and change spots irrelevant to the operation, and form the final achievement of the change spots;
[0085] The final result is automatically checked to obtain the target result file that meets the preset technical requirements.
[0086] For example, after the results are sorted out, further check whether the changed spots in the results file meet the requirements for spot extraction, continue to remove pseudo-changed spots and business-irrelevant spots, and form the final results of changed spots. Automatically check the final results to check whether the results file format and content, attribute field settings, attribute content filling, spot topology, image format and naming meet the technical requirements.
[0087] It should be noted that the above embodiments are only the best examples and are not limitations on the implementation methods of the present application.
[0088] Furthermore, the present application also proposes an AI intelligent guided remote sensing monitoring system, which is applied to any of the above AI intelligent guided remote sensing monitoring methods, specifically as follows: Figure 2 As shown, this application proposes a functional module schematic diagram of an AI intelligent guided remote sensing monitoring system, including:
[0089] The data-ready processing module 21 is used to collect initial satellite remote sensing image data and perform data-ready processing on the initial satellite remote sensing image data to obtain target satellite remote sensing image data; the initial satellite remote sensing image data is the satellite remote sensing image data of the base period and the current period;
[0090] The change spot automatic identification module 22 is used to identify the change spots in the target satellite remote sensing image data according to preset classifications, and perform job distribution operations based on the number of image slices or the number of change spots to obtain multiple change spot sets; the preset classifications include geographical division classification, temporal classification, topographic classification, and model type classification;
[0091] The knowledge-assisted spot confirmation module 23 is used to perform preliminary screening and spot review and editing on each of the change spots in the change spot set to obtain a target change spot set; the preliminary screening is to eliminate misidentified change spots and change spots irrelevant to the operation; the spot review and editing includes the delineation of spot boundaries and the filling of image-related attributes;
[0092] The achievement arrangement and inspection module 24 is used to arrange and inspect the achievement of the target change pattern set to obtain a target achievement file that meets the preset technical requirements.
[0093] Exemplarily, the data-ready processing module 21 realizes functions such as making a drop map file, generating a work grid and image slices; the change spot automatic identification module 22 realizes functions such as model training, image prediction, accuracy evaluation, and automatic task distribution; the knowledge-assisted spot confirmation module 23 realizes the comparison and screening of the change spots and their corresponding before and after image areas, quickly removes false changes, and semi-automatically extracts spots, collects and edits, and performs quality inspection. The results sorting and inspection module 24 realizes functions such as results sorting, troubleshooting, and automatic inspection.
[0094] like Figure 3 As shown, an embodiment of the present application also provides an electronic device 300, including a processor 310, a memory 320, and a computer program 321 stored in the memory 320 and executable on the processor, and when the processor 310 executes the computer program 321, the steps of any one of the above-mentioned methods of AI intelligent-guided remote sensing monitoring are implemented.
[0095] Since the electronic device introduced in this embodiment is a device used to implement an AI intelligent-guided remote sensing monitoring method in the embodiment of the present application, based on the method introduced in the embodiment of the present application, technical personnel in this field can understand the specific implementation mode of the electronic device of this embodiment and its various variations. Therefore, how the electronic device implements the method in the embodiment of the present application is not introduced in detail here. As long as the equipment used by technical personnel in this field to implement the method in the embodiment of the present application falls within the scope of protection of this application.
[0096] In the specific implementation process, when the computer program 321 is executed by the processor, it can achieve Figure 1 Any implementation manner in the corresponding embodiments.
[0097] It should be noted that in the above embodiments, the description of each embodiment has its own emphasis, and for parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.
[0098] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-readable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-readable program code.
[0099] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, 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, an embedded computer, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.
[0100] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.
[0101] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.
[0102] An embodiment of the present application also provides a computer program product, which includes computer software instructions. When the computer software instructions are executed on a processing device, the processing device executes the process of the AI intelligent-guided remote sensing monitoring method.
[0103] The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on the computer, the process or function according to the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from a website site, a computer, a server or a data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (digital subscriber line, DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) mode to another website site, computer, server or data center. The computer-readable storage medium can be any available medium that a computer can store or a data storage device such as a server or a data center that includes one or more available media integration. The available medium can be a magnetic medium, (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid state drive (SSD)), etc.
[0104] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.
[0105] In the several embodiments provided in the present application, it should be understood that the disclosed devices, apparatuses and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of units is only a logical function division, and there may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.
[0106] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
[0107] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.
[0108] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the various embodiments of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (Read-Only Memory, ROM), random access memory (Random Access Memory, RAM), disk or optical disk and other media that can store program codes.
[0109] The above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.
[0110] Although the preferred embodiments of this specification have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of this specification.
[0111] Obviously, those skilled in the art can make various changes and modifications to this specification without departing from the spirit and scope of this specification. Thus, if these modifications and variations of this specification fall within the scope of the claims of this specification and their equivalents, this specification is also intended to include these modifications and variations.
Claims
1. An AI-guided remote sensing monitoring method, characterized in that: include: Collecting initial satellite remote sensing image data, and performing data-ready processing on the initial satellite remote sensing image data to obtain target satellite remote sensing image data; the initial satellite remote sensing image data is the satellite remote sensing image data of the base period and the current period; Identify the change spots in the target satellite remote sensing image data according to preset classifications, and perform job distribution operations based on the number of image slices or the number of change spots to obtain multiple change spot sets; the preset classifications include geographic division classification, temporal phase classification, topographic classification, and model type classification; Performing preliminary screening and spot review and editing on the change spots in each of the change spot sets to obtain a target change spot set; the preliminary screening is to eliminate misidentified change spots and change spots irrelevant to the operation; the spot review and editing includes delineating the spot boundaries and filling in image-related attributes; The target change pattern set is collated and checked to obtain a target result file that meets the preset technical requirements.
2. The AI intelligent guided remote sensing monitoring method according to claim 1 is characterized in that: Collecting initial satellite remote sensing image data and performing data-ready processing on the initial satellite remote sensing image data to obtain target satellite remote sensing image data, including: Draw image mosaic lines based on the satellite remote sensing image coverage each quarter and generate drop map files; Generate an operation area grid according to the scope of the mission area and the acquisition of the initial satellite remote sensing image data; In combination with the image mosaic lines and the work area grid, a slicing tool is used to cut the current and base period images to obtain image slices; the image slices are the target satellite remote sensing image data.
3. The AI intelligent guided remote sensing monitoring method according to claim 1 is characterized in that: In the geographical zoning classification, Northeast China, North China, East China, Central China, South China, Southwest China and Northwest China are used for geographical zoning; in the phase classification, four seasons of a year are used for phase classification; in the landform classification, plateaus, mountains, plains, hills and basins are used for landform classification; in the model type classification, small models, medium models and large models are divided; different geographical divisions, different seasons and different landforms are respectively set with different models to identify regional change elements; different models correspond to different recognition accuracy and running speed.
4. The AI intelligent guided remote sensing monitoring method according to claim 1, characterized in that: Preliminary screening of the change spots in each of the change spot sets includes: Using a single change pattern as a calculation unit mask, respectively calculating the texture features inside the masks of the previous and next phase images; the texture features include contrast, correlation and entropy; averaging the texture features of the before and after phase images within the mask to obtain object-level texture features; The confidence of the change spots is calculated by using the object-level texture features, so as to eliminate the misidentified change spots and the change spots irrelevant to the operation based on the values of the confidence.
5. The AI intelligent guided remote sensing monitoring method according to claim 4 is characterized in that: The brightness difference between adjacent pixels in the image is determined by the contrast; the spatial consistency or correlation of pixel values in the image is determined by the correlation; and the uncertainty of the distribution of pixel values in the image is determined by the entropy.
6. The AI intelligent guided remote sensing monitoring method according to claim 1, characterized in that: The target change patch set is collated, including: Fill in the relevant attributes of the target change map to form county-level results; For the target change pattern, the image blocks of the base period and the current period are respectively intercepted, and the monitoring pattern screenshot image file is produced for the target change pattern by using automatic processing technology; The monitoring spot screenshot image file generates corresponding metadata to organize all the results into a result file in a standard format.
7. The AI intelligent guided remote sensing monitoring method according to claim 6, characterized in that: After obtaining the achievement file, the target achievement file is obtained by checking the achievement file, which specifically includes: Check the target change spots in the achievement file, remove misidentified change spots and change spots irrelevant to the operation, and form the final achievement of the change spots; The final result is automatically checked to obtain the target result file that meets the preset technical requirements.
8. An AI-guided remote sensing monitoring system, characterized in that: include: A data-ready processing module is used to collect initial satellite remote sensing image data and perform data-ready processing on the initial satellite remote sensing image data to obtain target satellite remote sensing image data; the initial satellite remote sensing image data is the satellite remote sensing image data of the base period and the current period; The change spot automatic identification module is used to identify the change spots in the target satellite remote sensing image data according to preset classifications, and perform job distribution operations based on the number of image slices or the number of change spots to obtain multiple change spot sets; the preset classifications include geographical division classification, temporal phase classification, topographic classification and model type classification; The knowledge-assisted spot confirmation module is used to perform preliminary screening and spot review and editing on the change spots in each of the change spot sets to obtain a target change spot set; the preliminary screening is to eliminate misidentified change spots and change spots irrelevant to the operation; the spot review and editing includes the delineation of spot boundaries and the filling of image-related attributes; The result sorting and checking module is used to sort and check the target change pattern set to obtain a target result file that meets the preset technical requirements.
9. An electronic device, comprising: A memory and a processor, characterized in that the processor is used to implement the steps of the AI intelligent-guided remote sensing monitoring method as described in any one of claims 1 to 7 when executing the computer program stored in the memory.
10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the AI intelligent-guided remote sensing monitoring method as described in any one of claims 1 to 7 are implemented.
Citation Information
Cited By
Change detection pattern spot optimization method, electronic equipment and computer product
CN119887833A
Change detection patch optimization method, electronic device, and computer product
CN119887833B
Intelligent detection method and system for urban land cover change image, terminal and storage medium
CN122116056A
