A method and system for land spatial planning analysis based on digital twins

Abstract

This invention relates to the field of spatial data processing, specifically to a method and system for land spatial planning analysis based on digital twins. Based on the inherent positioning information of remote sensing images, the method determines the positioning data of marker elements within the remote sensing image of the target space, thereby generating a digital twin space. All marker elements are clustered to divide the space into several subspaces, facilitating a comprehensive and sufficient state representation of each subspace. Based on the correlation between the state representation quantities of the subspaces and several planning requirement information, the method determines the planning decision for the actual land area of ​​each subspace within the target space and generates a land use pattern model. This invention enables global digital twin-level segmentation and planning matching under different types of land resource scenarios, improving the efficiency and accuracy of land planning and utilization, and achieving comprehensive decision-making on land resources.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of spatial data processing, in particular to a land space planning analysis method and system based on digital twinning. BACKGROUND

[0002] Land, as an important urban development resource, affects the scale and layout of urban development. As the foundation of urban construction, with the acceleration of urbanization, under the condition of limited land resources, it is necessary to consider the spatial development utilization rate, environmental and ecological protection, residential comfort, and layout of different functional areas. The existing land resource planning is usually demand-oriented to find suitable plots on a large amount of available land to implement planning. The above-mentioned method can maximize the satisfaction of different land use demands, but the implementation process usually needs to analyze a large number of plots one by one, which not only consumes a lot of manpower and material resources, but also seriously affects the efficiency and progress of land use planning.

[0003] In addition, land resources have the characteristics of spatial continuous distribution and complex and variable three-dimensional form, and the above characteristics need to be considered during the implementation of land use planning. However, at present, there is no overall utilization planning for the spatial structure characteristics of land resources, and no repeatable digital model of land resources is constructed, which cannot realize accurate planning of land resources. Therefore, how to segment and plan matching according to the global spatial state of land resources at the level of digital twinning, and realize overall and omnidirectional construction decision of land resources, has very important significance for improving land utilization rate and planning accuracy. SUMMARY

[0004] Considering the continuous range of spatial distribution of land resources and the complex and variable three-dimensional form, in order to implement global digital twinning layer segmentation and planning matching in different types of land resource scenarios, improve the efficiency and accuracy of land planning, and realize overall and omnidirectional decision of land resources, the present application provides a land space planning analysis method based on digital twinning, which comprises the following steps:

[0005] S100: identifying a plurality of identification elements from a remote sensing image of a target space, determining the positioning data of the identification elements according to the internal positioning information of the remote sensing image, arranging all the identification elements according to the positioning data, and generating a digital twin space of the target space;

[0006] S200: clustering all the identification elements of the digital twin space according to the astronomical feature information of the target space, thereby segmenting to form a plurality of subspaces; updating the remote sensing image to obtain the morphological change characteristics of all the identification elements in the subspaces;

[0007] S300: According to the morphological change characteristics, the range correction and element detail information correction are performed on the subspace to generate a state representation of the subspace;

[0008] S400: According to the correlation between the state representation of the subspace and a plurality of planning requirement information, a planning decision of the subspace in the real land range of the target space is determined to generate a land use pattern model.

[0009] Preferably, in S100, a plurality of identification elements are identified from a remote sensing image of the target space, the positioning data of the identification elements is determined according to the internal positioning information of the remote sensing image, all identification elements are arranged according to the positioning data, and a digital twin space of the target space is generated, specifically:

[0010] The in-picture element occlusion state of a plurality of remote sensing images of the target space is identified, and a remote sensing image with the least occlusion interference is selected therefrom; object contour recognition is performed on the selected remote sensing image to obtain a plurality of identification elements; the identification elements refer to objects in the remote sensing image picture that meet the corresponding three-dimensional morphological conditions;

[0011] The positioning data of the identification elements is determined according to the geographical positioning information of a plurality of reference lines in the selected remote sensing image picture and the relative position relationship between the identification elements and all reference lines;

[0012] All identification elements are mapped and arranged in a virtual space with the same dimension size as the target space according to the positioning data, and a digital twin space of the target space is generated.

[0013] Preferably, in S200, all identification elements of the digital twin space are clustered according to the astronomical feature information of the target space to form a plurality of subspaces; the remote sensing image is updated to obtain the morphological change characteristics of all identification elements in the subspaces, specifically:

[0014] The sun-shining feature information of the target space is obtained according to the geographical range position of the target space; the sun-shining feature information includes sun elevation angle dynamic change information; the sun-shining occlusion relationship between different identification elements is determined according to the sun-shining feature information and all identification elements in the digital twin space, all identification elements are correspondingly divided into a plurality of clusters, and a plurality of subspaces are correspondingly segmented according to the occupied ranges of the clusters; the identification elements under different clusters will not be sun-shining occluded;

[0015] The obtained remote sensing image is compared with the last obtained remote sensing image to determine the identification elements that have spatial level changes, and the morphological change characteristics of all identification elements in the subspace are determined therefrom; wherein the morphological change characteristics include spatial position change characteristics and external size change characteristics of the identification elements.

[0016] Preferably, in S300, according to the morphological change characteristics, the range of the subspace is corrected and the element detail information is corrected to generate the state representation quantity of the subspace, specifically:

[0017] According to the spatial position change characteristics of the identification elements contained in the morphological change characteristics, the boundary position of the subspace is changed to correct the range of the subspace; according to the external size change characteristics of the identification elements contained in the morphological change characteristics, the external contour information of the corresponding identification elements in the subspace is corrected;

[0018] The spatial range boundary position of the subspace and the positions and three-dimensional morphological sizes of all identification elements in the subspace are semantically segmented and converted to generate a multi-modal representation matrix of the subspace.

[0019] Preferably, in S400, according to the correlation between the state representation quantity of the subspace and the planning demand information, the planning decision of the subspace in the real land range of the target space is determined to generate a land use pattern model, specifically:

[0020] The multi-modal representation matrix contained in the state representation quantity of the subspace and the semantic matrix corresponding to each planning demand information are semantically associated and matched one by one to determine the semantic correlation degree between the state representation quantity and each planning demand information; according to the semantic correlation degree, a plurality of planning demand information is selected as the matching planning demand information of the subspace;

[0021] According to the position adjacency relationship of all subspaces in the real land range of the target space, the matching planning demand information of all subspaces is subjected to secondary screening to generate the planning decision of each subspace in the real land range of the target space, thereby generating a land use pattern model.

[0022] On the other hand, the application provides a national space planning analysis system based on digital twinning, which comprises the following modules:

[0023] An element recognition module is used to recognize a plurality of identification elements from a remote sensing image of a target space, and to determine the positioning data of the identification elements according to the inherent positioning information of the remote sensing image;

[0024] a twin space generation module configured to arrange all the identified elements according to the positioning data to generate a digital twin space of the target space;

[0025] a segmentation module configured to cluster all the identified elements of the digital twin space according to astronomical feature information of the target space to segment to form a plurality of subspaces;

[0026] an element change identification module configured to update the remote sensing image to obtain morphological change features of all the identified elements in the subspaces;

[0027] a representation generation module configured to correct a range and element detail information of the subspaces according to the morphological change features to generate state representations of the subspaces;

[0028] a land use determination module configured to determine planning decisions of the subspaces in a real land range of the target space according to a correlation between the state representations of the subspaces and a plurality of planning demand information to generate a land use pattern model.

[0029] Preferably, the element identification module is configured to identify a plurality of identified elements from a remote sensing image of the target space, and determine positioning data of the identified elements according to inherent positioning information of the remote sensing image, specifically as follows:

[0030] perform in-picture element occlusion state identification on a plurality of remote sensing images of the target space, select a remote sensing image with minimum occlusion interference therefrom, perform object contour identification on the selected remote sensing image to obtain a plurality of identified elements, and the identified elements refer to objects in the remote sensing image picture that meet corresponding three-dimensional morphological conditions;

[0031] determine the positioning data of the identified elements according to geographical positioning information of a plurality of reference lines in the selected remote sensing image picture and relative positional relationships between the identified elements and all the reference lines;

[0032] The twin space generation module is configured to arrange all the identified elements according to the positioning data to generate a digital twin space of the target space, specifically as follows:

[0033] map and arrange all the identified elements in a virtual space with the same dimension size as the target space according to the positioning data to generate the digital twin space of the target space.

[0034] Preferably, the segmentation module is configured to cluster all the identified elements of the digital twin space according to astronomical feature information of the target space to segment to form a plurality of subspaces, specifically as follows:

[0035] According to the geographical range position of the target space, the sun shining feature information of the target space is acquired; wherein the sun shining feature information includes the sun elevation angle dynamic change information; according to the sun shining feature information and all the identification elements in the digital twin space, the sun shining shielding relationship between different identification elements is determined, so as to correspondingly divide all the identification elements into several clusters, thereby corresponding several clusters occupy the range corresponding to the segmentation to form several subspaces; wherein the identification elements under different clusters will not have sun shining shielding events;

[0036] The element change identification module is used to update the remote sensing image, so as to obtain the morphological change characteristics of all the identification elements in the subspace, specifically:

[0037] By comparing the acquired remote sensing image with the remote sensing image acquired last time, the identification elements that have spatial level changes are determined, and the morphological change characteristics of all the identification elements in the subspace are determined therefrom; wherein the morphological change characteristics include the spatial position change characteristics and the shape size change characteristics of the identification elements;

[0038] The representation generation module is used to correct the range of the subspace and the element detail information according to the morphological change characteristics, so as to generate the state representation of the subspace, specifically:

[0039] According to the spatial position change characteristics of the identification elements contained in the morphological change characteristics, the boundary position of the subspace is changed, so as to correct the range of the subspace; according to the shape size change characteristics of the identification elements contained in the morphological change characteristics, the shape contour information of the corresponding identification elements in the subspace is corrected;

[0040] The spatial range boundary position of the subspace and the positions and three-dimensional morphological sizes of all the identification elements in the subspace are semantically segmented and converted, and a multi-modal representation matrix of the subspace is generated.

[0041] Preferably, the land use determination module is used to determine the planning decision of the subspace in the real land range of the target space according to the correlation between the state representation of the subspace and the planning demand information, so as to generate a land use pattern model, specifically:

[0042] The multi-modal representation matrix contained in the state representation of the subspace and the semantic matrix corresponding to each planning demand information are semantically associated and matched one by one, so as to determine the semantic correlation degree between the state representation and each planning demand information; according to the semantic correlation degree, a plurality of planning demand information is selected as the matching planning demand information of the subspace;

[0043] According to the abutment relationship of the respective positions of all subspaces in the real land range of the target space, the matching planning demand information of all subspaces is subjected to secondary screening, so as to generate the planning decision of each subspace in the real land range of the target space, thereby generating a land use pattern model.

[0044] In addition, the application also provides a computer readable storage medium, wherein the computer readable storage medium stores a computer program, and the computer program realizes the method as described above when executed by a processor.

[0045] Compared with the prior art, the application has the following beneficial effects:

[0046] The land space planning analysis method and system based on digital twinning provided by the application are aimed at the characteristics of large continuous range of spatial distribution of land resources and complex and changeable three-dimensional form, implement global digital twinning level segmentation and planning matching under different types of land resource scenes, improve land planning and utilization efficiency and accuracy, and realize overall and omnidirectional decision of land resources. BRIEF DESCRIPTION OF DRAWINGS

[0047] In order to more clearly illustrate the technical solutions in the embodiments of the application, the drawings needed to be used in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the application, and other drawings can be obtained by those skilled in the art without creative effort. Among them:

[0048] Figure 1 is a flowchart of a land space planning analysis method based on digital twinning provided by the application.

[0049] Figure 2 is the distribution of identification elements in the target space.

[0050] Figure 3 is the digital twinning space of the target space.

[0051] Figure 4 is the sunlight blocking condition of the identification elements.

[0052] Figure 5 is the generation process of the multi-modal representation matrix.

[0053] Figure 6 is the semantic association matching process of the state representation quantity and the planning demand information.

[0054] Figure 7 is a land use pattern model.

[0055] Figure 8 is a structural diagram of a land space planning analysis system based on digital twinning provided by the application. DETAILED DESCRIPTION

[0056] In order to make the above objectives, features and advantages of the present application more apparent, specific embodiments of the present application will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are intended for explanation only, and are not a limitation of the present application. In addition, it should be noted that only the parts related to the present application are shown in the drawings for ease of description. Based on the embodiments of the present application, all other embodiments obtained by those of ordinary skill in the art without creative labor fall within the scope of the present application.

[0057] The terms "comprising" and "having" and any variations thereof herein are intended to cover a non-exclusive inclusion. For example, a process, method, system, product or apparatus that comprises a list of steps or units is not limited to the listed steps or units, but can optionally further include other steps or units not listed, or can optionally further include other steps or units inherent to such processes, methods, products or apparatus.

[0058] Reference herein to "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment can be included in at least one embodiment of the application. The appearance of the phrase in various places in the specification is not necessarily all referring to the same embodiment, nor is it necessarily referring to a separate or alternative embodiment to the other embodiments. It will be explicitly understood by those of ordinary skill in the art that the embodiments described herein can be combined with other embodiments.

[0059] Please refer to Figure 1 As shown in the drawings, the present application provides a land space planning analysis method based on digital twinning, which comprises the following steps:

[0060] S100: identifying a plurality of identification elements from a remote sensing image of a target space, determining the positioning data of the identification elements according to the internal positioning information of the remote sensing image, arranging all the identification elements according to the positioning data, and generating a digital twin space of the target space.

[0061] Further, in S100, a plurality of identification elements are identified from a remote sensing image of a target space, the positioning data of the identification elements is determined according to the internal positioning information of the remote sensing image, all the identification elements are arranged according to the positioning data, and a digital twin space of the target space is generated, specifically:

[0062] In-frame element occlusion state recognition is performed on a plurality of remote sensing images of the target space, and a remote sensing image with the least occlusion interference is selected therefrom; object contour recognition is performed on the selected remote sensing image to obtain a plurality of identification elements; the identification elements refer to objects in the remote sensing image that meet the corresponding three-dimensional form conditions;

[0063] determine the positioning data of the identification element according to the geographical positioning information of the selected reference lines in the remote sensing image and the relative position relationship of the identification element with all the reference lines;

[0064] arrange all the identification elements in a virtual space with the same dimension as the target space according to the positioning data, and generate a digital twin space of the target space.

[0065] In the operation of territorial space planning, in order to implement fine identification planning on the global territorial space, the global territorial space is first divided into grids to obtain a plurality of target spaces with the same shape and size. The purpose of territorial space planning is to determine the potential use of each target space according to the terrain, distribution of original artificial objects (such as man-made buildings) / non-artificial objects (such as vegetation, water bodies, etc.), lighting conditions, etc. of each target space, so as to facilitate subsequent precise and targeted construction planning on all target spaces in the global territorial space. Considering that the original artificial objects / non-artificial objects in each target space differ in type, location, shape and size, such differences affect the use type and construction scheme of each target space.

[0066] Considering that the area of each target space is wide, the number of objects contained in each target space is large, and the structure and morphology of the objects are quite different, if all the objects in each target space are identified and analyzed one by one, not only a lot of resources and time will be consumed, but also it cannot be guaranteed that the identification and analysis results of each object will have an impact on the land planning. Therefore, remote sensing vision needs to be performed on each target space to calibrate and identify the mutual spatial relationship between each object and different objects in each target space. Specifically, a plurality of remote sensing images corresponding to a plurality of shooting directions of the same target space are obtained, the in-picture element occlusion state of each remote sensing image is identified, that is, the occlusion proportion of all objects in each remote sensing image by atmospheric clouds is identified, and according to the above occlusion proportion, a remote sensing image with the smallest occlusion interference (i.e. with the smallest occlusion proportion) is selected, so as to ensure that the selected remote sensing image has high object identifiable characteristics. Then, the in-picture object contour recognition is performed on the above selected remote sensing image to obtain all objects in the remote sensing image that meet the corresponding three-dimensional morphology conditions. The above three-dimensional morphology conditions can include but are not limited to a plurality of preset three-dimensional shape types and / or a preset three-dimensional size range. Through the above method, each object in the target space is screened on the three-dimensional morphology level, and only part of the objects are calibrated as identification elements, which provides sufficient and accurate element object basis for subsequent generation of a digital twin space matching the target space.

[0067] The geographical position information of each boundary line of the selected remote sensing image is known in advance. By using the known latitude and longitude of each boundary line as a reference line, the relative position relationship between each identification element and all reference lines is obtained. The relative position relationship can include, but is not limited to, the relative distance and relative direction between each identification element and each reference line. The geographical position information of all reference lines and the relative position relationship between each identification element and each reference line are combined to calculate the position transformation, determine the positioning data of each identification element, and map all identification elements into the target space according to the positioning data to obtain the distribution of all identification elements. Please refer to Figure 2 The distribution of all identification elements in the target space with a length of 10km along the X direction and a length of 7.5km along the Y direction, wherein each irregularly shaped pattern corresponds to an identification element. It can be understood that Figure 2 The spatial distribution state of all identification elements in terms of position and shape size is represented. In order to implement all-round operation and processing of the target space and the identification elements inside it at the digital level, the position coordinates of each identification element are extracted from the positioning data of all identification elements, and all identification elements are mapped and arranged in a virtual space with the same dimension size as the target space according to the position coordinates, to generate a digital twin space of the target space. Please refer to Figure 3 As shown in the figure, the identification elements of the same color correspond to objects of the same type in the target space, and the identification elements of different colors correspond to objects of different types. By using Figure 3 The digital twin space can provide a basis for further analyzing the morphological changes of each identification element at the digital space level.

[0068] S200: According to the astronomical characteristic information of the target space, cluster all identification elements in the digital twin space to form a plurality of subspaces; update the remote sensing image to obtain the morphological change characteristics of all identification elements in the subspaces.

[0069] Further, in S200, according to the astronomical characteristic information of the target space, cluster all identification elements in the digital twin space to form a plurality of subspaces; update the remote sensing image to obtain the morphological change characteristics of all identification elements in the subspaces, specifically:

[0070] According to the geographical range position of the target space, the sun height angle dynamic change information and other sun characteristic information of the target space are obtained; wherein the sun characteristic information includes the sun height angle dynamic change information; according to the sun characteristic information and all the identified elements in the digital twin space, the sun shading relationship between different identified elements is determined, so as to correspondingly divide all the identified elements into several clusters, thereby corresponding to the segmentation of the range occupied by the several clusters to form several subspaces; wherein the identified elements under different clusters will not have sun shading events between them.

[0071] The obtained remote sensing image is compared with the remote sensing image obtained last time to determine the identified elements that have spatial level changes, and the morphological change characteristics of all the identified elements in the subspace are determined therefrom; wherein the morphological change characteristics include the spatial position change characteristics and the size change characteristics of the identified elements.

[0072] Sunlight is an important consideration factor for land space planning, and sun parameters such as sunlight intensity and duration directly affect the use planning of the corresponding land of the target space. Considering that the area range of the target space is large, the sun parameters at different positions in the target space are correspondingly different, so that the actual sun state of the identified elements at different positions in the target space is also different. In order to calibrate the sun distribution state of the target space in the full range, first, according to the longitude and latitude span range of the target space, the sun height angle dynamic change information and other sun characteristic information of the target space are obtained, and the sun shading relationship between different identified elements is determined in combination with the above sun characteristic information and the relative spatial distribution between all the identified elements in the above digital twin space. Please refer to Figure 4 , the actual sun shading situation of the identified elements at different positions in the target space, wherein the proportion of the sun shining on the first type of identified elements, the second type of identified elements and the third type of identified elements is reduced in turn. By determining the range size and position of the sun shading of each identified element itself, the sun shading relationship between different identified elements is determined, wherein the above sun shading relationship can be but is not limited to the sun shading relationship formed by a certain identified element to other identified elements. According to the sun shading relationship between different identified elements, several clusters are divided; wherein any two identified elements in each cluster will have sun shading events, and two identified elements belonging to different clusters will not have sun shading events. By performing the above cluster division on all the identified elements, and dividing the range occupied by each cluster in the target space as a subspace, it is convenient for subsequent implementation of individual identified element morphological change identification for each subspace.

[0073] Considering that the shape of the identified element in the target space changes due to its own action or the action of the external environment, such shape change can be reflected in the change of terrain height and slope, and once the shape of the target space changes, the land use planning of the target space will also change. Therefore, the land use planning of the target space is updated in time, another remote sensing image of the target space is obtained, and is compared with the remote sensing image obtained last time to determine the identified element whose shape changes in the space level, so as to obtain the spatial position change characteristics and shape size change characteristics of the identified element whose shape changes in each subspace, and provide basis for generating the multi-modal representation matrix of the identified element in each subspace.

[0074] S300: According to the shape change characteristics, the range of the subspace is corrected and the element detail information is corrected, so as to generate the state representation quantity of the subspace.

[0075] Further, in S300, according to the shape change characteristics, the range of the subspace is corrected and the element detail information is corrected, so as to generate the state representation quantity of the subspace, specifically:

[0076] According to the spatial position change characteristics of the identified element contained in the shape change characteristics, the boundary position of the subspace is changed, so that the range of the subspace is corrected; according to the shape size change characteristics of the identified element contained in the shape change characteristics, the shape contour information of the corresponding identified element in the subspace is corrected;

[0077] The spatial range boundary position of the subspace and the position and three-dimensional shape size of all identified elements in the subspace are subjected to semantic segmentation and conversion to generate a multi-modal representation matrix of the subspace.

[0078] Through the above analysis, it can be known that when the identified element in each subspace changes in space shape, the spatial position of the identified element changes, so that the range of each subspace changes accordingly, and if the original range of the subspace is still used as the basis for multi-modal analysis at this time, it will inevitably lead to a large analysis error. Therefore, the range of the subspace needs to be corrected in real time. Please refer to Figure 5, first, according to the spatial position change characteristics of the identification elements contained in the morphological change characteristics, the boundary position of the subspace is changed, so as to correct the range of the subspace; and according to the size change characteristics of the identification elements contained in the morphological change characteristics, the shape contour information of the corresponding identification elements in the subspace is corrected, so as to realize the accurate correction of the range boundary of the subspace and the shape contour of the identification elements in the subspace. Then, the spatial range boundary position of the corrected subspace and the position and three-dimensional shape size (i.e. the shape contour information of the identification elements) of all identification elements are subjected to semantic segmentation and conversion by using a deep learning network model, to generate a multi-modal representation matrix of each subspace, wherein the multi-modal representation matrix refers to a matrix whose elements are the position coordinates and three-dimensional shape size of all identification elements in each subspace. The multi-modal representation matrix of the subspace can comprehensively and accurately represent the spatial state of the subspace, and provide a reliable basis for subsequent planning and decision-making of the target space.

[0079] S400: According to the correlation between the state representation of the subspace and the planning demand information, the planning decision of the subspace in the real land range of the target space is determined, so as to generate a land use pattern model.

[0080] Further, in S400, according to the correlation between the state representation of the subspace and the planning demand information, the planning decision of the subspace in the real land range of the target space is determined, so as to generate a land use pattern model, specifically:

[0081] The multi-modal representation matrix contained in the state representation of the subspace and the semantic matrix corresponding to each planning demand information are subjected to semantic association matching one by one, to determine the semantic correlation between the state representation and each planning demand information; according to the semantic correlation, a plurality of planning demand information is selected as the matching planning demand information of the subspace;

[0082] According to the position adjacency relationship of all subspaces in the real land range of the target space, the matching planning demand information of all subspaces is subjected to secondary screening, to generate the planning decision of each subspace in the real land range of the target space, so as to generate a land use pattern model.

[0083] From the above analysis, it can be seen that the multi-modal representation matrix contained in the state representation of each subspace has matrix elements corresponding to the position coordinates and three-dimensional shape size of all identification elements in each subspace, so that the multi-modal representation matrix of each subspace can fully and accurately reflect the real identification element spatial morphology of the subspace. Please refer to Figure 6The multi-modal representation matrix of the subspace is matched with the semantic matrix corresponding to each planning requirement information one by one to determine the semantic correlation between the state representation and each planning requirement information, and the planning requirement information corresponding to the first few highest semantic similarities is taken as the matching planning requirement information of the subspace. It can be understood that each planning requirement information includes planning use type, land construction volume rate, building ground range and the like. According to the above process, each subspace corresponds to several matching planning requirement information, that is, each subspace can be planned for multiple different land uses. In order to avoid that a large number of adjacent subspace in the target space are planned for the same land use and / or the land use planned by the adjacent two subspace conflicts, according to the position adjacency relationship of all subspace in the real land range of the target space, the matching planning requirement information of all subspace is subjected to secondary screening to generate the planning decision of each subspace in the real land range of the target space, and finally the land use pattern model of the whole target space is generated. Please refer to Figure 7 Any one of the planning decisions of the residential land planning, commercial land planning, green land planning, industrial land planning and reserved land planning is set for each subspace in the target space, so that the target space as a whole obtains sufficient and accurate planning use, improves the efficiency and accuracy of land planning use, and realizes overall and omnidirectional decision of land resources.

[0084] In addition, the above land use pattern model can be digitally converted to generate a digital map, and the digital map is edited in layers corresponding to each subspace, and the corresponding street address information, legal status information, geographic space information, construction project (progress) information and the like are added to the digital map, so as to facilitate users to query and edit the corresponding digital map in the cloud in real time, and realize flexible and visual operation of land space planning.

[0085] Please refer to Figure 8 As shown in the drawings, the present application provides a land space planning analysis system based on digital twinning, which comprises the following modules:

[0086] An element recognition module is used to recognize a plurality of identification elements from the remote sensing image of the target space, and to determine the positioning data of the identification elements according to the internal positioning information of the remote sensing image.

[0087] A twinning space generation module is used to arrange all the identification elements according to the positioning data to generate a digital twinning space of the target space.

[0088] A segmentation module is used to cluster all the identification elements of the digital twinning space according to the astronomical feature information of the target space to segment to form a plurality of subspaces.

[0089] An element change recognition module is configured to update the remote sensing image, so as to obtain morphological change characteristics of all identified elements in the subspace;

[0090] A representation generation module is configured to correct the range of the subspace and the element detail information according to the morphological change characteristics, so as to generate the state representation of the subspace;

[0091] A land use determination module is configured to determine a planning decision of the real land range of the subspace in the target space according to the correlation between the state representation of the subspace and a plurality of planning demand information, so as to generate a land use pattern model.

[0092] Further, the element recognition module is configured to identify a plurality of identified elements from the remote sensing image of the target space, and determine the positioning data of the identified elements according to the internal positioning information of the remote sensing image, specifically:

[0093] The element recognition module is configured to perform in-picture element occlusion state recognition on a plurality of remote sensing images of the target space, select a remote sensing image with the least occlusion interference, perform object contour recognition on the selected remote sensing image, and obtain a plurality of identified elements; the identified element refers to an object in the remote sensing image picture that meets the corresponding three-dimensional morphological condition;

[0094] The element recognition module is configured to determine the positioning data of the identified elements according to the geographical positioning information of a plurality of reference lines in the selected remote sensing image picture and the relative position relationship between the identified elements and all reference lines;

[0095] A twin space generation module is configured to arrange all identified elements according to the positioning data, and generate a digital twin space of the target space, specifically:

[0096] The twin space generation module is configured to arrange all identified elements according to the positioning data in a virtual space with the same dimension size as the target space, and generate a digital twin space of the target space.

[0097] Further, the segmentation module is configured to cluster all identified elements of the digital twin space according to astronomical characteristic information of the target space, so as to segment to form a plurality of subspaces, specifically:

[0098] The segmentation module is configured to obtain the sunlight characteristic information of the target space according to the geographical range position of the target space; the sunlight characteristic information includes the dynamic change information of the solar elevation angle; the sunlight occlusion relationship between different identified elements is determined according to the sunlight characteristic information and all identified elements in the digital twin space, so as to divide all identified elements into a plurality of clusters, thereby corresponding to a plurality of clusters occupying ranges to segment to form a plurality of subspaces; the identified elements under different clusters will not have sunlight occlusion events;

[0099] The element change recognition module is configured to update the remote sensing image, so as to obtain morphological change characteristics of all identified elements in the subspace, specifically:

[0100] By comparing the updated remote sensing image with the previously acquired remote sensing image, the marker elements that have undergone spatial changes are identified, and the morphological change characteristics of all marker elements within the subspace are determined. The morphological change characteristics include the spatial position change characteristics and the external size change characteristics of the marker elements.

[0101] The representation quantity generation module is used to correct the range and element detail information of the subspace based on the morphological change characteristics, thereby generating the state representation quantity of the subspace, specifically:

[0102] Based on the spatial position change characteristics of the identifier elements contained in the morphological change characteristics, the boundary position of the subspace is changed, thereby correcting the range of the subspace; based on the external size change characteristics of the identifier elements contained in the morphological change characteristics, the external outline information of the corresponding identifier elements in the subspace is corrected.

[0103] Semantic segmentation and transformation are performed on the spatial boundary location of the subspace, as well as the location and three-dimensional shape size of all identifier elements within the subspace, to generate a multimodal representation matrix of the subspace.

[0104] Furthermore, the land use determination module is used to determine the planning decision of the actual land area of ​​the subspace in the target space based on the correlation between the state representation quantity of the subspace and several planning demand information, thereby generating a land use pattern model, specifically:

[0105] Semantic association matching is performed one by one between the multimodal representation matrix contained in the state representation quantity of the subspace and the semantic matrix corresponding to each of the several planning requirement information to determine the semantic relevance between the state representation quantity and each planning requirement information; based on the semantic relevance, multiple planning requirement information are selected as the matching planning requirement information of the subspace.

[0106] Based on the adjacency relationships of each subspace within the actual land area of ​​the target space, a secondary screening is performed on the matching planning requirements information of all subspaces to generate planning decisions for the actual land area of ​​each subspace within the target space, thereby generating a land use pattern model.

[0107] The operation and effects of the digital twin-based land spatial planning analysis system of the present invention are consistent with those of the aforementioned digital twin-based land spatial planning analysis method, and will not be described again here.

[0108] In one embodiment of the present invention, the present invention also provides a computer-readable storage medium on which a computer program is stored, the computer program implementing the method described above when executed by a processor.

[0109] In an embodiment of the present application, the present application also provides a computer device comprising at least a memory and a processor, wherein a computer program is stored on the memory, and the computer program, when executed by the processor, implements the method as described above.

[0110] Those skilled in the art can clearly understand that the embodiments can be implemented by means of the necessary universal hardware platforms, and of course can also be implemented by means of the combination of hardware and software. Based on such understanding, the above technical solutions can be embodied in the form of computer products, and the present application can be embodied in the form of computer program products implemented on one or more computer usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer usable program codes.

[0111] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit it, and other embodiments can also be used; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement for part of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A digital-twin-based land space planning analysis method, characterized in that, The method comprises the following steps: S100: identifying a plurality of identification elements from a remote sensing image of a target space, determining positioning data of the identification elements according to inherent positioning information of the remote sensing image, arranging all the identification elements according to the positioning data, and generating a digital twin space of the target space; S200: clustering all the identification elements in the digital twin space according to astronomical feature information of the target space, thereby segmenting to form a plurality of subspaces; and updating the remote sensing image to obtain morphological change characteristics of all the identification elements in the subspaces, specifically: obtaining solar radiation feature information of the target space according to a geographical range position of the target space; wherein the solar radiation feature information comprises solar elevation angle dynamic change information; determining solar radiation shielding relationships between different identification elements according to the solar radiation feature information and all the identification elements in the digital twin space, thereby corresponding division of all the identification elements into a plurality of clusters, so as to segment a plurality of subspaces corresponding to the range occupied by the plurality of clusters; wherein the identification elements under different clusters will not have solar radiation shielding events; comparing the obtained remote sensing image with a remote sensing image obtained last time to determine identification elements having spatial level changes, and determining morphological change characteristics of all the identification elements in the subspaces therefrom; wherein the morphological change characteristics comprise spatial position change characteristics and external size change characteristics of the identification elements; S300: range correction and element detail information correction of the subspaces according to the morphological change characteristics, thereby generating state representation quantities of the subspaces; S400: determining planning decisions of the subspaces in real land ranges of the target space according to a correlation between the state representation quantities of the subspaces and a plurality of planning requirement information, thereby generating a land use pattern model, specifically: performing semantic correlation matching between a multi-modal representation matrix contained in the state representation quantities of the subspaces and semantic matrices corresponding to the plurality of planning requirement information respectively, determining semantic correlation degrees between the state representation quantities and each planning requirement information, and selecting a plurality of planning requirement information according to the semantic correlation degrees as matching planning requirement information of the subspaces; implementing secondary screening of the matching planning requirement information of all the subspaces according to position adjacency relationships of all the subspaces in the real land ranges of the target space, thereby generating planning decisions of each subspace in the real land ranges of the target space, and generating a land use pattern model.

2. The digital twin-based national space planning analysis method according to claim 1, wherein in S100, a plurality of identification elements are identified from a remote sensing image of a target space, positioning data of the identification elements is determined according to inherent positioning information of the remote sensing image, and all the identification elements are arranged according to the positioning data, thereby generating a digital twin space of the target space, specifically: ​ The in-picture element occlusion state of a plurality of remote sensing images of a target space is recognized, and a remote sensing image with the least occlusion interference is selected from the remote sensing images; object contour recognition is performed on the selected remote sensing image to obtain a plurality of identification elements; the identification elements refer to objects in the remote sensing image picture that meet corresponding three-dimensional shape conditions; According to the geographical positioning information of a plurality of reference lines in the selected remote sensing image picture and the relative position relationship between the identification elements and all reference lines, the positioning data of the identification elements is determined; According to the positioning data, all identification elements are arranged one by one in a virtual space with the same dimension size as the target space to generate a digital twin space of the target space.

3. The digital twin-based land space planning analysis method according to claim 1, characterized in that, In S300, according to the shape change characteristics, the range of the subspace is corrected and the element detail information is corrected to generate the state representation quantity of the subspace, specifically: According to the spatial position change characteristics of the identification elements contained in the shape change characteristics, the boundary position of the subspace is changed to correct the range of the subspace; according to the shape size change characteristics of the identification elements contained in the shape change characteristics, the shape contour information of the corresponding identification elements in the subspace is corrected; The spatial range boundary position of the subspace and the position and three-dimensional shape size of all identification elements in the subspace are subjected to semantic segmentation and conversion to generate a multi-modal representation matrix of the subspace.

4. A digital-twin-based land space planning analysis system, characterized by, The system comprises the following modules: An element recognition module is configured to recognize a plurality of identification elements from a remote sensing image of a target space, and determine positioning data of the identification elements according to inherent positioning information of the remote sensing image; A twin space generation module is configured to arrange all identification elements according to the positioning data to generate a digital twin space of the target space; A segmentation module is configured to cluster all identification elements of the digital twin space according to astronomical characteristic information of the target space to form a plurality of subspaces, specifically: According to the geographical range position of the target space, the sun exposure characteristic information of the target space is obtained; wherein the sun exposure characteristic information comprises sun elevation angle dynamic change information; according to the sun exposure characteristic information and all identification elements in the digital twin space, the sun exposure occlusion relationship between different identification elements is determined, so that all identification elements are correspondingly divided into a plurality of clusters, thereby a plurality of subspaces are correspondingly segmented according to the range occupied by the clusters; wherein the identification elements under different clusters will not have sun exposure occlusion events; An element change recognition module is configured to update the remote sensing image to obtain shape change characteristics of all identification elements in the subspace, specifically: The obtained remote sensing image is compared with the remote sensing image obtained last time to determine the identification elements that have spatial level changes, and the shape change characteristics of all identification elements in the subspace are determined therefrom; wherein the shape change characteristics comprise spatial position change characteristics and shape size change characteristics of the identification elements; An element change recognition module is configured to update the remote sensing image to obtain shape change characteristics of all identification elements in the subspace, specifically: The obtained remote sensing image is compared with the remote sensing image obtained last time to determine the identification elements that have spatial level changes, and the shape change characteristics of all identification elements in the subspace are determined therefrom; wherein the shape change characteristics comprise spatial position change characteristics and shape size change characteristics of the identification elements; The state characterization quantity generation module is configured to correct the range and element details of the sub-space according to the morphological change characteristics, so as to generate the state characterization quantity of the sub-space. The land use determination module is configured to determine the planning decision of the real land range of the target space according to the correlation between the state characterization quantity of the sub-space and the planning demand information, so as to generate the land use pattern model, specifically: The semantic correlation between the state characterization quantity and each planning demand information is determined by performing semantic association matching between the multi-modal characterization matrix contained in the state characterization quantity of the sub-space and the semantic matrix corresponding to each planning demand information; and the semantic correlation degree is used to select a plurality of planning demand information as the matching planning demand information of the sub-space. The matching planning demand information of all sub-spaces is subjected to secondary screening according to the location adjacency relationship of each sub-space in the real land range of the target space, so as to generate the planning decision of each sub-space in the real land range of the target space, thereby generating the land use pattern model.

5. The digital-twin-based national space planning analysis system according to claim 4, wherein The element recognition module is configured to recognize a plurality of identification elements from the remote sensing image of the target space, and determine the positioning data of the identification elements according to the internal positioning information of the remote sensing image, specifically: The element recognition module is configured to recognize a plurality of identification elements from the remote sensing image of the target space, and determine the positioning data of the identification elements according to the internal positioning information of the remote sensing image, specifically: The positioning data of the identification elements is determined according to the geographical positioning information of a plurality of reference lines in the selected remote sensing image and the relative position relationship between the identification elements and all reference lines. The twin space generation module is configured to arrange all identification elements according to the positioning data, so as to generate the digital twin space of the target space, specifically: The digital twin space of the target space is generated by mapping and arranging all identification elements in a virtual space with the same dimension size as the target space according to the positioning data.

6. The digital-twin-based national space planning analysis system according to claim 4, wherein The state characterization quantity generation module is configured to correct the range and element details of the sub-space according to the morphological change characteristics, so as to generate the state characterization quantity of the sub-space, specifically: The range of the sub-space is changed according to the spatial position change characteristics of the identification elements contained in the morphological change characteristics, so as to correct the range of the sub-space; and the contour information of the corresponding identification elements in the sub-space is corrected according to the size change characteristics of the identification elements contained in the morphological change characteristics. The spatial range boundary position of the sub-space and the positions and three-dimensional morphological sizes of all identification elements in the sub-space are subjected to semantic segmentation and conversion, so as to generate the multi-modal characterization matrix of the sub-space.

7. A computer readable storage medium having stored thereon a computer program, characterized in that The computer program, when executed by a processor, implements the digital-twin-based national space planning analysis method according to any one of claims 1-3.

Images