A natural resource data analysis method and system

Abstract

The present invention relates to a natural resource data analysis method and system. The natural resource data analysis method includes, after acquiring remote sensing image data, dividing the remote sensing image data into regions according to known geographical spots of natural resources and classifying the divided image regions according to the categories of natural resources; for image regions of natural resources of different categories, using an analysis model targeted to the categories to analyze the remote sensing image data of the image regions. Preliminary classification identification can be completed without accurate recognition of remote sensing image data, and the computing resources consumed are extremely low; it also has good analysis accuracy.

Description

Technical Field

[0001] The present invention relates to the field of natural resource planning, and in particular to a natural resource data analysis method and system. Background Art

[0002] The preparation of a plan for the protection and use of natural resource assets owned by the whole people is an important task in the current rights and interests management of the natural resource department. In the production process of the natural resource asset protection and use plan, it is necessary to integrate the basic data catalog list through production application (management) software, integrate spatial analysis according to the planning evaluation link, realize the automatic or semi-automatic processing of data and results, and display the results, so as to realize the informatization of asset planning, improve work efficiency and reduce human resources. In the current natural resource data processing, it is necessary to manually mark and analyze the remote sensing image data, and the automated processing method will consume a lot of computing resources, and the processing accuracy of the remote sensing image data is very high, otherwise the results of the automated processing will be very rough. Summary of the invention

[0003] The purpose of this application is to propose a natural resource data analysis method and system to solve the problem that it is difficult to obtain relatively accurate analysis results with low computing resource consumption when automatically processing remote sensing image data of natural resources.

[0004] The natural resource data analysis method of this application includes:

[0005] After acquiring the remote sensing image data, the remote sensing image data is divided into regions according to known geographical spots of natural resources and the divided image regions are classified according to the categories of natural resources;

[0006] For the image areas of natural resources of different categories, an analysis model targeted at the categories is used to analyze the remote sensing image data of the image areas.

[0007] Furthermore, for the image areas of natural resources of different categories, using an analysis model targeted at the categories to analyze the remote sensing image data of the image areas includes:

[0008] Determine a first image region corresponding to a first category and its associated image regions, and merge them into a first temporary analysis region;

[0009] Determine the categories of major natural resources within the first temporary analysis area;

[0010] The first temporary analysis area is analyzed by using an analysis model that is targeted at the categories of the main natural resources.

[0011] Furthermore, after analyzing the first temporary analysis area by using an analysis model that is targeted at the categories of the main natural resources, the method further includes:

[0012] In the subsequent analysis of other image areas, the image areas of other types of natural resources other than the main natural resources in the first temporary analysis area are eliminated.

[0013] Further, if the analysis result indicates that an abnormality is suspected to exist in the image area of ​​other categories of natural resources other than the main natural resources, the abnormality is marked;

[0014] In the subsequent process of identifying the category of the abnormal image region, the abnormal image region is not eliminated, and a targeted analysis model is directly used to analyze the abnormal image region.

[0015] Furthermore, the determining of the categories of the main natural resources in the first temporary analysis area includes:

[0016] Determine that there are at least two categories of major natural resources within the first temporary analysis area;

[0017] The analyzing the first temporary analysis area by using an analysis model targeted at the categories of the main natural resources includes:

[0018] The first temporary analysis area is analyzed by a composite analysis model, wherein the composite analysis model targets at least two of the main natural resources at the same time.

[0019] Further, the determining of the first image region corresponding to the first category and its associated image region includes:

[0020] Determining whether there are other images of natural resources around the first image area;

[0021] If yes, determining the correlation between the natural resources in the first image area and other surrounding natural resources;

[0022] Image regions of other natural resources whose relevance to the natural resources at the first image region is greater than a set threshold are determined as image regions associated with the first image region.

[0023] Furthermore, the determining of the correlation between the natural resources in the first image area and other surrounding natural resources includes:

[0024] Determine the natural development relevance of the categories of other surrounding natural resources to the first category of natural resources;

[0025] Determine the asset planning relevance of the categories of other surrounding natural resources to the first category of natural resources;

[0026] The correlation between the categories of other surrounding natural resources and the first category of natural resources is determined based on the natural development correlation, the planning correlation and the distance attenuation factor.

[0027] Further, the determining of the first image region corresponding to the first category and its associated image regions, and merging them into the first temporary analysis region comprises:

[0028] Determining a minimum basic graphic covering the first image area and its associated image areas;

[0029] The image data within the minimum basic graphic forms the first temporary analysis area.

[0030] Furthermore, the dividing the remote sensing image data into regions according to the known geographical spots of natural resources and classifying the divided image regions according to the categories of natural resources includes:

[0031] Get historical geographic patches;

[0032] Based on the historical geographic spots, generating predicted geographic spots according to planning data and prediction data;

[0033] The remote sensing image data is divided into regions according to the predicted geographic spots and the divided image regions are classified according to the categories of natural resources.

[0034] On the other hand, the present application also provides a natural resource data analysis system, including a classification identification module and an analysis module;

[0035] The classification identification module is used to divide the remote sensing image data into regions according to known geographical spots of natural resources after the system acquires the remote sensing image data and classify the divided image regions according to the categories of natural resources;

[0036] The analysis module is used to analyze the remote sensing image data of the image area of ​​different categories of natural resources by using an analysis model that is targeted to the category.

[0037] The natural resource data analysis method provided in the present application, after acquiring the remote sensing image data, divides the remote sensing image data into regions according to the known geographical spots and classifies the divided image regions according to the categories of natural resources; thus, the preliminary classification identification can be completed without the need to accurately identify the remote sensing image data, avoiding the indiscriminate fine identification of the remote sensing image data, and consuming extremely low computing resources. For the image regions of natural resources of different categories, the remote sensing image data of the image regions are analyzed using an analysis model that is targeted to their categories; thus, accurate analysis of the remote sensing image data is achieved and the accuracy of the analysis is ensured; and the use of a targeted analysis model can simplify the complexity of the model and reduce the overall computing consumption. It can be seen that in some applications, the above-mentioned natural resource data analysis method simplifies the complexity of a single analysis model without consuming a large amount of computing resources, and also has good analysis accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 is a basic flow chart of the natural resource data analysis method of this embodiment;

[0039] Figure 2 A schematic diagram of the remote sensing image data and historical geographic map of this embodiment;

[0040] Figure 3 Based on Figure 2 Schematic diagram of regional division of remote sensing image data;

[0041] Figure 4 Schematic diagram of the structure of the natural resource data analysis system of this embodiment. DETAILED DESCRIPTION

[0042] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0043] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0044] Embodiment 1:

[0045] This embodiment provides a natural resource data analysis method, which includes but is not limited to Figure 1 The following steps:

[0046] S101, after acquiring remote sensing image data, dividing the remote sensing image data into regions according to known geographical spots of natural resources and classifying the divided image regions according to the categories of natural resources;

[0047] S102, for image areas of natural resources of different categories, using an analysis model that is targeted at the categories to analyze remote sensing image data of the image areas;

[0048] Remote sensing image data can be large-scale national land images collected by image satellites, or a large number of images reflecting land conditions collected by drones or other sensing equipment. In the conventional remote sensing image data analysis process, users are often required to manually label to identify natural resources in remote sensing image data. If image recognition is used to determine natural resources in remote sensing image data, it is necessary to directly identify the image. In this process, an intelligent model that can distinguish various types of natural resources is required, which will consume computing resources. In addition, the intelligent model for identifying the type of natural resources has great difficulty in identifying some implicit natural resources (such as mineral resources, etc.), and cannot carry out further analysis on various natural resources well, resulting in a large amount of computing resources being consumed in the process of automated analysis of remote sensing image data.

[0049] The natural resource data analysis method of this embodiment directly uses the known geographical spots of natural resources to divide and classify the remote sensing image data. That is, when determining the natural resources in the remote sensing image data, they are directly marked according to the known spots. Of course, it is understandable that the known geographical spots of natural resources are based on the historical data already in the system and are not completely consistent with the actual situation. Therefore, the classification performed in this process is preliminary. However, in the subsequent analysis process, the actual situation of the remote sensing image data will be analyzed and obtained, and because of the preliminary classification and division, the relevant analysis model is selected in a targeted manner, and the analysis results are relatively accurate.

[0050] It should be noted that a targeted analysis model refers to an analysis model that has a more accurate analysis result for an image of a certain category of natural resources relative to other analysis models that can be selected by the system. That is, the system will have at least two different analysis models, and the analysis model that is targeted for a certain category of natural resources is one of the ones with better analysis capabilities. A targeted analysis model for a certain category of natural resources is usually formed by training mainly based on images of natural resources of that category. For example, the analysis model for image recognition of natural resources such as woodlands and grasslands is mainly trained based on remote sensing images of such natural resources. When analyzing remote sensing image data of natural resources such as mining development and water resources, the accuracy of the results will be relatively poor. The analysis performed by the analysis model in this embodiment can be to automatically update the geographic map; or it can also be used for abnormal monitoring, such as analyzing whether natural resources are damaged or inconsistent with the plan based on remote sensing image data; in some applications, the analysis model can also perform other analysis projects, or obtain multiple analysis results at the same time. In fact, the embodiments of the present application do not limit the specific analysis process performed by the analysis model.

[0051] In practical applications, some similar natural resources can also be analyzed using the same analysis model. That is, the natural resources targeted by the analysis model can be multiple. Or in some applications, the granularity of classification when analyzing remote sensing image data can be larger, and a single category can be a large category covering multiple specific natural resource categories.

[0052] The geographical spots of known natural resources may be geographical spots that have been entered and determined in the system. These geographical spots may be data generated based on previous land surveys and other projects, and entered and stored in the system. Exemplarily, the information contained in the geographical spots includes but is not limited to its geographical coordinates, corresponding natural resources (if any), land planning conditions, data acquisition time, etc. In practical applications, natural resources usually do not change a lot, and their changes usually have traces to follow, so the present application can directly perform preliminary classification of remote sensing image data based on known geographical spots.

[0053] It should be understood that, in addition to being obtained through historical actual surveys, the geographical maps of known natural resources can also be obtained based on planning or predictions of the natural development of natural resources. For example, planning data of natural resources can be entered into the system, and then geographical maps after the planned development can be generated according to the planned natural resources. In some applications, the system can also predict the natural changes of natural resources in the geographical maps. For example, based on changes in climate and surrounding environment, it can predict the situation of some natural resources after a period of time; specifically, for natural resources such as woodlands and grasslands, changes in their coverage area can be predicted based on historical data.

[0054] As an example, in some implementations, dividing the remote sensing image data into regions according to known geographic spots of natural resources and classifying the divided image regions according to the categories of natural resources includes the following steps:

[0055] S201, obtaining historical geographical spots;

[0056] S202, based on the historical geographical spots, generating predicted geographical spots according to the planning data and the prediction data;

[0057] S203, dividing the remote sensing image data into regions according to the predicted geographic spots and classifying the divided image regions according to the categories of natural resources;

[0058] It is understandable that the above steps S201 and S202 may be performed in the previous data analysis process, or the results may be directly input. That is, the above steps do not need to be performed after the remote sensing image data is newly acquired. Of course, for natural resources that have not been planned or are not easy to change, the historical geographical spots can be directly used as the benchmark.

[0059] For a better understanding, an example is given here. Suppose there is a historical geographical spot 1, which is a spot of wasteland, and in the plan, the wasteland will be changed to cultivated land. After acquiring the remote sensing image data, it can be determined whether the time of acquiring the remote sensing image data is after the planning is completed. If so, the geographical spot after the planning is completed, that is, the type of the historical geographical spot 1 can be directly changed to forest land. The corresponding area of ​​the remote sensing image data is determined by the scope of the planned forest land, and marked as the type of forest land. Of course, in actual applications, the shape and range of the spot to be formed after planning may also be different from the original historical geographical spot 1, but based on the planning situation, a new geographical spot will be formed.

[0060] Similarly, if there is predicted data on changes in natural resources, it can also be decided whether to use the predicted geographic spots as the benchmark based on the time of the remote sensing image data. If the time of the remote sensing image data is closer to the predicted time, the predicted geographic spots can be used as the benchmark; if the time of the remote sensing image data is closer to the time of the historically recorded geographic spots, the historically recorded geographic spots can be selected without considering the predicted changes.

[0061] It can be seen that in the process of preliminary classification, only simple judgment rules are needed, and in-depth image analysis cannot be performed, so the computational burden is small.

[0062] Optionally, in order to further improve the accuracy of the analysis, for image areas of natural resources of different categories, the step of analyzing the remote sensing image data of the image area using an analysis model targeted to the category may include:

[0063] S301, determining a first image region corresponding to a first category and its associated image regions, and merging them into a first temporary analysis region;

[0064] S302, determining the categories of major natural resources in the first temporary analysis area;

[0065] S303, analyzing the first temporary analysis area by using an analysis model that is targeted at the categories of major natural resources;

[0066] The main natural resources will be those with a proportion of more than 30%. In some applications, more than 40% can also be selected as the main natural resources. The proportion refers to the area ratio of each natural resource in the first temporary analysis area.

[0067] To simplify the description, it is assumed that the image areas associated with the first image area (hereinafter referred to as the second image area) are all natural resources of the second category. Then the first image area will merge the second image area associated with its surroundings into the first temporary analysis area, and the subsequent analysis will be based on the first temporary analysis area as a whole. That is, it can reflect the association between natural resources, have a more accurate analysis result, and can also reduce the number of areas to be analyzed, which is conducive to reducing the computational burden. In the first temporary analysis area, natural resources that account for more than 30% will be determined as the main natural resources. At this time, if only the first image area accounts for more than 30%, the analysis model that is targeted at the natural resources of the first category is selected for analysis. If only the second image area accounts for more than 30%, the analysis model that is targeted at the natural resources of the second category is selected for analysis. If the proportion of the first image area and the second image area is both more than 30%, the natural resources of the image area with a larger area can be selected as the main natural resources.

[0068] In some implementation processes, in order to obtain better comprehensive analysis results, it is determined that the categories of main natural resources in the first temporary analysis area may be no less than two. For example, when the proportions of the first image area and the second image area are both above 30%, the natural resources corresponding to the two are both determined as main natural resources. Then, analyzing the first temporary analysis area through an analysis model that is targeted at the categories of main natural resources includes: analyzing the first temporary analysis area through a composite analysis model, and the composite analysis model targets at least two of the main natural resources at the same time. For example, the analysis model can be formed by training based on the associated data of the first category of natural resources and the second category of natural resources. Of course, the training data of the analysis model can cover more situations and is not limited to this example.

[0069] In some embodiments, determining the first image region corresponding to the first category and its associated image region comprises:

[0070] S401, determining whether there are other natural resource images around the first image area, if so, executing S402;

[0071] S402, determining the correlation between the natural resources in the first image area and other surrounding natural resources;

[0072] S403, determining image regions of other natural resources whose relevance to the natural resources at the first image region is greater than a set threshold as image regions associated with the first image region;

[0073] The above step S401 can be judged by setting a distance threshold to determine whether there are other natural resources within the set distance. Based on known geographic spots, it is easy to determine whether there are other natural resources within the distance threshold range corresponding to the position of the first image area.

[0074] The association referred to in this application usually refers to the correlation between these natural resources in terms of planning, natural development or their location. Among them, the correlation in natural development can be based on the natural evolution of time. When a natural resource changes, it will have an impact on other surrounding natural resources. For example, when the coverage area of ​​water resources is reduced, the location covered by natural resources such as surrounding woodlands or grasslands may also be reduced to a certain extent. The correlation in position is reflected in adjacent natural resources. The increase in the coverage area of ​​one natural resource may mean a reduction in the coverage area of ​​another adjacent natural resource. Of course, for different natural resources that are not directly adjacent but close to each other, there may also be a certain degree of mutual influence, which can be represented by a distance attenuation factor.

[0075] In some embodiments, determining the correlation between the natural resources in the first image area and other surrounding natural resources includes:

[0076] S501, determining the natural development correlation between the categories of other surrounding natural resources and the first category of natural resources;

[0077] S502, determining the asset planning relevance between the categories of other surrounding natural resources and the first category of natural resources;

[0078] S503, determining the relevance of the categories of other surrounding natural resources to the first category of natural resources according to the natural development relevance, planning relevance and distance attenuation factor;

[0079] Exemplarily, the correlation R between the first category of natural resources and other surrounding natural resources can be expressed by the following formula 1.

[0080] Formula 1: R = (R1 + R2 + R3 ... + Rn) * a

[0081] Among them, R1 and R2 are the natural development correlation and asset planning correlation respectively; in practical applications, other correlation parameters can also be set according to the situation, which are represented by R3 to Rn in formula 1; a is the distance decay factor. It can be understood that the distance decay factor a is negatively correlated with the distance, that is, the greater the distance between natural resources, the smaller the distance decay factor a, and thus the smaller the correlation between natural resources. When the distance between natural resources exceeds the aforementioned distance threshold, the distance decay factor a can be considered to be 0.

[0082] The natural development correlation can be determined based on the categories of the two natural resources. The natural development correlation between the natural resources can be set in advance in the system. When calculating, the natural development correlation is determined directly based on the current category and the category of other surrounding resources. The asset planning correlation can be set based on the planning of natural resource assets. In the process of managing and trading natural resources, multiple natural resources may be formed into asset packages for unified management and linkage planning. The asset planning correlation can be determined based on the proportion of various natural resources in the asset package and the planning scheme. The rules for determining various correlations are set in the system. When analysis is required, the values ​​of each correlation can be calculated based on the rules. The calculation is also very simple and does not involve complex analysis models.

[0083] In this embodiment, the analysis of remote sensing image data will be performed in sequence based on the categories of natural resources. That is, the image area of ​​one category of natural resources is analyzed first, and then the image area of ​​another category of natural resources is analyzed, until the analysis of all remote sensing image data is completed. Of course, since this embodiment also includes its associated image areas when analyzing the image area of ​​natural resources, it is possible that some image areas are repeatedly analyzed in different analysis steps. These repeatedly analyzed image areas may obtain analysis results through different analysis models, so the relatively accurate analysis results can be selected as the final results, which is beneficial to the overall analysis accuracy.

[0084] In some embodiments, for the purpose of saving computing resources, after analyzing the first temporary analysis area using an analysis model that is targeted at the categories of major natural resources, the method further includes:

[0085] In the subsequent analysis of other image regions, image regions of other types of natural resources contained in the first temporary analysis region are eliminated.

[0086] Still taking the example that the first temporary analysis area includes natural resources of the first category and the second category, in the process of analyzing the first temporary analysis area, assuming that a targeted analysis model for natural resources of the first category is adopted, the remote sensing image data in the first temporary analysis area will all be analyzed by the analysis model. When the image area of ​​the second category of natural resources is subsequently analyzed, the area of ​​the natural resources belonging to the second category in the first temporary analysis area will no longer be analyzed again. Thus, it is possible to avoid the same image area being processed multiple times during the analysis and processing of multiple categories, thereby reducing the consumption of computing resources. Since the image areas of these other categories of natural resources contained in the first temporary analysis area only occupy a small proportion, even if they are eliminated in the subsequent analysis process, the impact on the overall analysis accuracy is also small.

[0087] Of course, in order to ensure the processing and identification of anomalies, if the analysis results indicate that there are suspected anomalies in the image area of ​​other categories of natural resources associated with the first category, the anomaly will be identified; and in the subsequent process of identifying the category of the image area with the abnormality, the image area with the abnormality will not be eliminated, and a targeted analysis model will be directly used to analyze the image area with the abnormality.

[0088] As an example, in the process of analyzing the first image area, the first temporary analysis area may include natural resources of the first category and natural resources of the second category, and the analysis is performed using an analysis model that is targeted at the first category of natural resources. At this time, if the analysis result shows that the image area corresponding to the second category of natural resources in the first temporary analysis area may have abnormalities, then in the subsequent analysis of the image area that may have abnormalities, the analysis model that is targeted at the second category is directly used for analysis. In this process, only the areas that may have abnormalities can be analyzed to reduce the amount of data; or the analysis can be performed directly based on the first temporary analysis area that has been divided, without the need to redefine the association and plan the analysis area, so that more image information and the correlation between natural resources are retained to obtain relatively accurate information; the above two methods can be selected according to actual conditions.

[0089] Of course, if no abnormality is found, deduplication can be performed in the subsequent analysis process. For example, each first image area of ​​the first category is analyzed first (which can be called the first round of analysis); then each second image area of ​​the second category is analyzed (called the second round of analysis). In the second round of analysis, the second image area that has been included in the analysis in the first round of analysis and the analysis results do not indicate abnormalities can be directly ignored.

[0090] In addition, in order to simplify the segmentation difficulty of remote sensing image data and facilitate the normalization of analyzed data, the process of determining the first image area corresponding to the first category and its associated image areas and merging them into the first temporary analysis area may include the following steps:

[0091] S601, determining a minimum basic graphic covering a first image area and its associated image areas;

[0092] S602, forming a first temporary analysis area with the image data in the minimum basic graph;

[0093] The minimum basic figure is usually selected as a square, and other relatively regular basic figures such as triangles and trapezoids may also be used in some implementation processes. The first temporary analysis area is formed directly with the minimum basic figure as a unit, so that the first temporary analysis area is more regular, avoiding the identification of fragmented boundaries, and is conducive to simplifying the data processing before inputting the analysis model.

[0094] In order to better understand the natural resource data analysis method of the present application, this embodiment is described below in conjunction with a specific example.

[0095] See also Figure 2As shown, a remote sensing image data is schematically shown, which includes an image area A corresponding to natural resources of a first category and an image area B corresponding to natural resources of a second category. Figure 2 In the figure, the actual natural resource coverage area is indicated by a solid line. It should be understood that the natural resource coverage area shown by the solid line is shown in this embodiment for the convenience of explanation and understanding. Since it has not been manually marked or automatically identified, there is no distinguishing mark for different natural resource areas in the actual remote sensing image data, and the remote sensing image data exists in the form of a whole or a combination of multiple images.

[0096] Figure 2 The known geographical spots are also indicated by dotted lines, where geographical spots A1 are spots of natural resources of the first category; geographical spots B1 are spots of natural resources of the second category. Of course, the remaining geographical spots may be spots of non-natural resources, such as villages, towns, cities, etc., which are not shown in this embodiment. In addition, Figure 2 If no dotted line is shown, it means that the known geographic map and the actual situation of the remote sensing image data are the same, and no dotted line is shown to simplify the figure, such as the area of ​​B (B1).

[0097] from Figure 2 It can be seen that the actual situation of natural resources in remote sensing image data may be different from the known geographical spots, which may be because natural resources have changed based on natural development or human development. In this example, the image area in the remote sensing image data is still divided according to the areas of geographical spots A1 and B1.

[0098] After completing the regional division and classification of the remote sensing image data, the first image area A2 and its associated image areas are determined for the first round of analysis. Figure 3 As shown in the figure, the first image area A2 and its associated second image area B2, the first image area A2 is actually the area corresponding to the geographic spot A1 in the remote sensing image data, and the second image area B2 is actually the area corresponding to the surrounding geographic spot B1 in the remote sensing image data. For ease of processing, for each first image area A2, the smallest quadrilateral that can encompass the first image area A2 and its associated second image area B2 is determined. Figure 3 As shown, a quadrilateral first temporary analysis area C1 can be determined as a basic unit of the input analysis model. For multiple first image areas A2 in the remote sensing image data, first temporary analysis areas C1 can be formed respectively. In other examples, if there is overlap between multiple first temporary analysis areas C1, these first temporary analysis areas C1 can also be merged to form a larger temporary analysis area. Figure 3It can be seen that the scope of the first temporary analysis area formed is larger than the map spots of each natural resource. It can be seen that by determining the minimum basic figure covering these areas to form a temporary analysis area, it can also help to determine whether the natural resources have expanded beyond the original map range.

[0099] In this example, the main natural resources in each first temporary analysis area C1 are natural resources of the first category. Therefore, the image data of the first temporary analysis area C1 is input into the analysis model that is targeted at the first category of natural resources for analysis to obtain the analysis result. In other examples, the main natural resources in the first temporary analysis area C1 may also be natural resources of the second category, and the analysis model that is targeted at the second category of natural resources will be used for analysis.

[0100] Assume that in this example, no suspected abnormality is found during the analysis of the first temporary analysis area C1. Then the next round of analysis can be started. Figure 3 As shown, in the remote sensing image data, a part of the second image area B2 has been merged into the first temporary analysis area C1 in the first round of analysis to complete the analysis. However, a part of the second image area B2 is not associated with the first image area A2 and has not been analyzed; for example, the second image area B2 on the right side of the illustrated direction. In this round of analysis, the unanalyzed second image area B2 will be analyzed using an analysis model that is targeted at the second category of natural resources. The second image area B2 that was merged into the first temporary analysis area C1 in the first round of analysis and completed the analysis can be ignored to avoid repeated analysis. In other examples, if there is a suspected abnormal second image area B2 in the first round of analysis, it will be analyzed again in the second round of analysis using an analysis model that is targeted at the second category of natural resources to obtain a more accurate analysis result.

[0101] The above example only takes two types of natural resources as an example. When there are three or more types of natural resources, the analysis process type is to complete the analysis of the second type of natural resources and then continue to analyze the third type of natural resources, and thus complete the data analysis of all natural resources.

[0102] This embodiment also provides a natural resource data analysis system 100, such as Figure 4 As shown, it includes a classification identification module 101 and an analysis module 102.

[0103] The classification identification module 101 is used to divide the remote sensing image data into regions according to the known geographical spots of natural resources after the system acquires the remote sensing image data and classify the divided image regions according to the categories of natural resources;

[0104] The analysis module 102 is used to analyze the remote sensing image data of the image areas of natural resources of different categories by using an analysis model that is targeted at the categories.

[0105] In addition, the specific execution process of the steps executed by each module of the natural resource data analysis system 100 of this embodiment can also refer to the steps of the natural resource data analysis method provided above in this embodiment, which will not be repeated in this embodiment.

[0106] In addition, although exemplary embodiments have been described herein, the scope includes any and all embodiments based on the present disclosure that are equivalent, modified, omitted, combined (e.g., various embodiments intersecting), adapted or changed. It is not limited to the examples described in this specification or during the implementation of this application, and its examples will be interpreted as non-exclusive.

[0107] The above description is intended to be illustrative rather than restrictive. For example, the above examples (or one or more of them) can be used in combination with each other. For example, those of ordinary skill in the art may use other embodiments when reading the above description.

[0108] The above are only preferred embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, the embodiments of the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A natural resource data analysis method, characterized in that: include: After acquiring the remote sensing image data, the remote sensing image data is divided into regions according to known geographical spots of natural resources and the divided image regions are classified according to the categories of natural resources; For the image areas of natural resources of different categories, the remote sensing image data of the image areas are analyzed using an analysis model that is targeted at the categories; The method of analyzing the remote sensing image data of the image areas of the natural resources of different categories by using an analysis model targeted at the categories thereof includes: Determine a first image region corresponding to a first category and its associated image regions, and merge them into a first temporary analysis region; Determine the categories of major natural resources within the first temporary analysis area; Analyzing the first temporary analysis area by using an analysis model that is targeted at the categories of the main natural resources; The determining of the first image region corresponding to the first category and its associated image region comprises: Determining whether there are other images of natural resources around the first image area; If yes, determining the correlation between the natural resources in the first image area and other surrounding natural resources; Determine image areas of other natural resources whose relevance to the natural resources at the first image area is greater than a set threshold as image areas associated with the first image area; The determining of the correlation between the natural resources in the first image area and other surrounding natural resources includes: Determine the natural development relevance of the categories of other surrounding natural resources to the first category of natural resources; Determine the asset planning relevance of the categories of other surrounding natural resources to the first category of natural resources; Determine the relevance of the categories of other surrounding natural resources to the first category of natural resources according to the natural development relevance, the planning relevance and the distance attenuation factor; Among them, the correlation R between the first category of natural resources and other surrounding natural resources satisfies: R=(R1+R2+R3……+Rn)*a Among them, R1 and R2 are the natural development correlation and asset planning correlation respectively, R3 to Rn represent the set correlation parameters; a is the distance attenuation factor, and the distance attenuation factor a is negatively correlated with the distance. The greater the distance between natural resources, the smaller the distance attenuation factor a.

2. The natural resource data analysis method according to claim 1, characterized in that: After analyzing the first temporary analysis area by using an analysis model targeted at the categories of the main natural resources, the method further includes: In the subsequent analysis of other image areas, the image areas of other types of natural resources other than the main natural resources in the first temporary analysis area are eliminated.

3. The natural resource data analysis method according to claim 2, characterized in that: If the analysis results indicate that there are suspected anomalies in the image area of ​​other categories of natural resources other than the main natural resources, the anomalies will be marked; In the subsequent process of identifying the category of the abnormal image region, the abnormal image region is not eliminated, and a targeted analysis model is directly used to analyze the abnormal image region.

4. The natural resource data analysis method according to claim 1, characterized in that: Determining the categories of the main natural resources in the first temporary analysis area includes: Determine that there are at least two categories of major natural resources within the first temporary analysis area; The analyzing the first temporary analysis area by using an analysis model targeted at the categories of the main natural resources includes: The first temporary analysis area is analyzed by a composite analysis model, wherein the composite analysis model targets at least two of the main natural resources at the same time.

5. The natural resource data analysis method according to claim 1, characterized in that: The step of determining the first image region corresponding to the first category and its associated image regions and merging them into the first temporary analysis region comprises: Determining a minimum basic graphic covering the first image area and its associated image areas; The image data within the minimum basic graphic forms the first temporary analysis area.

6. The natural resource data analysis method according to any one of claims 1 to 5, characterized in that: The step of dividing the remote sensing image data into regions according to the known geographical spots of natural resources and classifying the divided image regions according to the categories of natural resources comprises: Get historical geographic patches; Based on the historical geographic spots, generating predicted geographic spots according to planning data and prediction data; The remote sensing image data is divided into regions according to the predicted geographic spots and the divided image regions are classified according to the categories of natural resources.

7. A natural resource data analysis system, characterized in that: It includes a classification identification module and an analysis module; The classification identification module is used to divide the remote sensing image data into regions according to known geographical spots of natural resources after the system acquires the remote sensing image data and classify the divided image regions according to the categories of natural resources; The analysis module is used to analyze the remote sensing image data of the image area of ​​different categories of natural resources by using an analysis model that is targeted to the category; The analysis module is also used for: Determine a first image region corresponding to a first category and its associated image regions, and merge them into a first temporary analysis region; Determine the categories of major natural resources within the first temporary analysis area; Analyzing the first temporary analysis area by using an analysis model that is targeted at the categories of the main natural resources; The analysis module is also used for: Determining whether there are other images of natural resources around the first image area; If yes, determining the correlation between the natural resources in the first image area and other surrounding natural resources; Determine image areas of other natural resources whose relevance to the natural resources at the first image area is greater than a set threshold as image areas associated with the first image area; The analysis module is also used for: Determine the natural development relevance of the categories of other surrounding natural resources to the first category of natural resources; Determine the asset planning relevance of the categories of other surrounding natural resources to the first category of natural resources; Determine the relevance of the categories of other surrounding natural resources to the first category of natural resources according to the natural development relevance, the planning relevance and the distance attenuation factor; Among them, the correlation R between the first category of natural resources and other surrounding natural resources satisfies: R=(R1+R2+R3……+Rn)*a Among them, R1 and R2 are the natural development correlation and asset planning correlation respectively, R3 to Rn represent the set correlation parameters; a is the distance attenuation factor, and the distance attenuation factor a is negatively correlated with the distance. The greater the distance between natural resources, the smaller the distance attenuation factor a.

Images