A method and system for identifying idle land development based on GIS

Abstract

The application discloses a GIS-based idle land construction development identification method and system, relates to the land supervision identification field, and comprises the following steps: performing multi-source heterogeneous fusion on remote sensing image data and aerial image data of a region to be researched to obtain a fused image data set; performing land block identification on the fused image data set based on a deep convolutional neural network and in combination with a geographic information system to obtain a land block segmentation coordinate set; performing spatial superposition analysis on the land block segmentation coordinate set in combination with Internet of Things sensing data to obtain a construction development state prediction layer, generating a human activity heat map layer based on a target tracking algorithm according to the Internet of Things sensing data and the aerial image data; and identifying idle land in a land block to be identified in the region to be researched according to the construction development state prediction layer and the human activity heat map layer in combination with public procedure data to obtain an idle land construction development identification result. The application can improve the idle land construction development identification accuracy.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of land supervision and identification, and particularly relates to a GIS-based idle land construction development identification method and system. BACKGROUND

[0002] Idle land is a typical type of "supply but not use" land, and "supply but not use" land causes serious waste of land resources. Therefore, land resource management of "supply but not use" land is conducive to promoting sustainable development of cities and optimizing land space layout.

[0003] At present, as for "supply but not use" land, especially idle land, its construction development or disposal situation depends on manual patrol and self-reporting mode, supplemented by simple remote sensing image checking and supervision means. However, this idle land construction development identification method has obvious defects: manual patrol checking consumes too much manpower and material resources, is low in efficiency, and is difficult to realize large-scale and normalized supervision; there is a situation of not timely reporting in some areas, resulting in too large difference between the increase speed and disposal speed of idle land; traditional remote sensing technology is limited by the image data itself, such as weather factors affecting the acquisition and resolution of remote sensing images, so that when idle land construction development is identified through remote sensing images, there is a situation of insufficient identification accuracy. The above defects all lead to insufficient accuracy of existing idle land construction development identification, and how to improve the accuracy of idle land construction development identification is still a difficult problem to be solved in the prior art. SUMMARY

[0004] The present application provides a GIS-based idle land construction development identification method and system to solve the technical problem of insufficient accuracy of existing idle land construction development identification.

[0005] According to a first aspect of the embodiment of the present application, a GIS-based idle land construction development identification method is provided, comprising:

[0006] Multi-source heterogeneous fusion is performed on remote sensing image data and aerial image data of a to-be-studied region to obtain a fused image data set;

[0007] Based on a deep convolutional neural network and in combination with a geographic information system, parcel identification is performed on the fused image data set to obtain a parcel segmentation coordinate set;

[0008] According to the parcel segmentation coordinate set, spatial superposition analysis is performed in combination with Internet of Things sensing data of the to-be-studied region to obtain a construction development state prediction layer corresponding to the to-be-studied region, and a human activity heat map layer corresponding to the to-be-studied region is generated based on a target tracking algorithm according to the Internet of Things sensing data and the aerial image data;

[0009] According to the construction development state prediction layer and the human activity heat map layer, in combination with the public procedure data of the region to be researched, idle land identification is performed on the land block to be identified in the region to be researched, and idle land construction development identification results of the land block to be identified are obtained.

[0010] The application performs multi-source heterogeneous fusion on remote sensing image data and aerial image data of the region to be researched to obtain a fused image data set, and then performs land block identification based on a deep convolutional neural network to obtain a land block segmentation coordinate set. Through multi-source heterogeneous fusion of remote sensing image data and aerial image data, the comprehensiveness of analysis data can be improved, and identification errors caused by single type data can be avoided. Then, a construction development state prediction layer is obtained through spatial superposition analysis in combination with Internet of Things sensing data of the region to be researched, and a human activity heat map layer is generated based on a target tracking algorithm according to the Internet of Things sensing data and the aerial image data. Through layer analysis of the Internet of Things sensing data and the aerial image data, the efficiency of analyzing the region to be researched can be improved, and the identification efficiency of the land block to be identified can be improved. At the same time, multiple types of data are added, the comprehensiveness of analysis data is improved, and the identification accuracy of the land block to be identified is improved. Then, idle land identification is performed on the land block to be identified in combination with the public procedure data of the region to be researched. The idle land identification results of the land block to be identified can be verified and corrected through the public procedure data, and the accuracy of idle land construction development identification is further improved.

[0011] In some embodiments of the application, the multi-source heterogeneous fusion of the remote sensing image data and the aerial image data of the region to be researched to obtain a fused image data set specifically includes:

[0012] According to the time stamp, the remote sensing image data and the aerial image data are time axis aligned to obtain first remote sensing image data and first aerial image data;

[0013] According to the spatial coordinates, the first remote sensing image data and the first aerial image data are spatial coordinate aligned to obtain second remote sensing image data and second aerial image data;

[0014] The second remote sensing image data and the second aerial image data are fused to obtain a fused image data set.

[0015] According to the time stamp, the remote sensing image data and the aerial image data are time axis aligned, and then according to the spatial coordinates, the spatial coordinates are aligned, and then fused to obtain a fused image data set. Through two batches of alignment of the time stamp and the spatial coordinates, multi-source heterogeneous data can be converted into multi-source homogeneous data, which provides favorable conditions for subsequent fusion.

[0016] In some embodiments of the present application, the deep convolutional neural network-based, in combination with the geographic information system, performs parcel recognition on the fused image dataset to obtain a parcel segmentation coordinate set, specifically comprising:

[0017] Based on the deep convolutional neural network, the fused image dataset is subjected to semantic segmentation to obtain a feature classification set of ground objects.

[0018] According to the feature classification set of ground objects, the fused image dataset is subjected to feature matching recognition based on an edge detection algorithm to obtain a parcel classification image set, and according to the parcel classification image set, in combination with the provided parcel spatial coordinate map obtained by the geographic information system, the association between the parcel and the coordinate is constructed to obtain the parcel segmentation coordinate set.

[0019] The present application first performs semantic segmentation on the fused image dataset based on the deep convolutional neural network to obtain a feature classification set of ground objects, which can separate each feature of ground objects from the fused image dataset, and then performs feature matching recognition on the fused image dataset based on the edge detection algorithm, which can identify and segment different types of parcels according to each feature of ground objects, and further construct the association between the parcel and the coordinate in combination with the provided parcel spatial coordinate map obtained by the geographic information system to obtain the parcel segmentation coordinate set, thereby providing conditions for subsequent analysis and recognition.

[0020] In some embodiments of the present application, according to the parcel segmentation coordinate set, in combination with the Internet of Things sensing data of the region to be studied, spatial superposition analysis is performed to obtain a development state prediction layer corresponding to the region to be studied, and according to the Internet of Things sensing data and the aerial image data, a human activity heat map layer corresponding to the region to be studied is generated based on a target tracking algorithm, specifically comprising:

[0021] According to the parcel segmentation coordinate set, in combination with the Internet of Things sensing data of the region to be studied, spatial superposition analysis is performed based on the layer superposition method to obtain the development state prediction layer.

[0022] According to the Internet of Things sensing data, first human activity trajectory data is extracted, and according to the aerial image data, second human activity trajectory data is obtained based on a target tracking algorithm.

[0023] Based on the layer superposition method, spatial superposition analysis is performed on the first human activity trajectory data and the second human activity trajectory data to obtain the human activity heat map layer.

[0024] The application first divides a coordinate set according to a plot, combines Internet of Things sensing data of a region to be studied, and obtains a construction and development state prediction layer based on a layer superposition method; then extracts first human activity track data according to the Internet of Things sensing data, and obtains second human activity track data based on a target tracking algorithm according to aerial image data, and then obtains a human activity heat layer based on a layer superposition method, so that layer analysis is performed through the Internet of Things sensing data and the aerial image data, the efficiency of analyzing the region to be studied is improved, the identification efficiency of the plot to be identified is improved, multiple types of data are added, the comprehensiveness of the analysis data is improved, and the identification accuracy of the plot to be identified is improved.

[0025] In some embodiments of the application, the idle land of the plot to be identified is identified in the region to be studied according to the construction and development state prediction layer and the human activity heat layer, and combined with public procedure data of the region to be studied, to obtain an idle land construction and development identification result of the plot to be identified, specifically including:

[0026] According to the construction and development state prediction layer, the plot distribution type of the plot to be identified is determined;

[0027] According to the human activity heat layer, a human activity identification result of the plot to be identified is determined;

[0028] According to the plot distribution type and the human activity identification result, a first idle land identification result is determined;

[0029] According to the public procedure data of the region to be studied, the first idle land identification result is verified and corrected to obtain an idle land construction and development identification result of the plot to be identified.

[0030] According to the construction and development state prediction layer and the human activity heat layer, the plot distribution type and the human activity identification result of the plot to be identified are determined respectively, and then the first idle land identification result is determined, and then the idle land construction and development identification result of the plot to be identified is obtained by combining the public procedure data, so that the accuracy of idle land construction and development identification is improved.

[0031] According to a second aspect of the embodiment of the application, an idle land construction and development identification system based on GIS is provided, which comprises a multi-source fusion module, a plot identification module, a layer analysis module and an idle land identification module.

[0032] The multi-source fusion module is used for multi-source heterogeneous fusion of remote sensing image data and aerial image data of a region to be studied to obtain a fused image data set.

[0033] The land parcel identification module is configured to identify land parcels based on a deep convolutional neural network and in combination with geographic information systems, to obtain a land parcel segmentation coordinate set from the fused image data set.

[0034] The layer analysis module is configured to perform spatial superposition analysis according to the land parcel segmentation coordinate set and in combination with Internet of Things (IoT) sensing data of the region to be studied, to obtain a construction and development state prediction layer corresponding to the region to be studied, and to generate a human activity heat map layer corresponding to the region to be studied based on a target tracking algorithm and according to the IoT sensing data and the aerial image data.

[0035] The idle land identification module is configured to identify idle land in a land parcel to be identified according to the construction and development state prediction layer and the human activity heat map layer and in combination with public procedure data of the region to be studied, to obtain an idle land construction and development identification result of the land parcel to be identified.

[0036] In some embodiments of the present application, the multi-source fusion module includes a time axis alignment unit, a spatial coordinate alignment unit, and a data fusion unit.

[0037] The time axis alignment unit is configured to perform time axis alignment on the remote sensing image data and the aerial image data according to timestamps, to obtain first remote sensing image data and first aerial image data.

[0038] The spatial coordinate alignment unit is configured to perform spatial coordinate alignment on the first remote sensing image data and the first aerial image data according to spatial coordinates, to obtain second remote sensing image data and second aerial image data.

[0039] The data fusion unit is configured to fuse the second remote sensing image data and the second aerial image data, to obtain a fused image data set.

[0040] In some embodiments of the present application, the land parcel identification module includes a semantic segmentation identification unit and a feature matching identification unit.

[0041] The semantic segmentation identification unit is configured to perform semantic segmentation on the fused image data set based on a deep convolutional neural network, to obtain a land feature classification set.

[0042] The feature matching identification unit is configured to perform feature matching identification on the fused image data set based on an edge detection algorithm according to the land feature classification set, to obtain a land parcel classification image set, and to construct an association between land parcels and coordinates according to the land parcel classification image set and an already-supplied land parcel spatial coordinate map obtained by a geographic information system, to obtain a land parcel segmentation coordinate set.

[0043] In some embodiments of the present application, the layer analysis module comprises a plot distribution analysis unit, an activity trajectory analysis unit and an activity heat analysis unit;

[0044] The plot distribution analysis unit is configured to perform spatial superposition analysis based on a layer superposition method according to the plot segmentation coordinate set and in combination with the Internet of Things sensing data of the region to be studied, to obtain the development status prediction layer.

[0045] The activity trajectory analysis unit is configured to extract first human activity trajectory data according to the Internet of Things sensing data, and obtain second human activity trajectory data based on a target tracking algorithm according to the aerial image data.

[0046] The activity heat analysis unit is configured to perform spatial superposition analysis on the first human activity trajectory data and the second human activity trajectory data based on a layer superposition method, to obtain the human activity heat layer.

[0047] In some embodiments of the present application, the idle land identification module comprises a plot type identification unit, a human activity identification unit, an idle land identification unit and an identification verification correction unit.

[0048] The plot type identification unit is configured to determine the plot distribution type of the plot to be identified according to the development status prediction layer.

[0049] The human activity identification unit is configured to determine the human activity identification result of the plot to be identified according to the human activity heat layer.

[0050] The idle land identification unit is configured to determine the first idle land identification result according to the plot distribution type and the human activity identification result.

[0051] The identification verification correction unit is configured to verify and correct the first idle land identification result according to the public procedure data of the region to be studied, to obtain the idle land development identification result of the plot to be identified.

[0052] The remote sensing image data and the aerial image data of the to-be-studied region are fused to obtain a fused image data set, and then land parcel recognition is performed based on a deep convolutional neural network to obtain a land parcel segmentation coordinate set. Through multi-source heterogeneous fusion of the remote sensing image data and the aerial image data, the comprehensiveness of the analysis data can be improved, and recognition errors caused by a single type of data can be avoided. Then, a construction development state prediction layer is obtained through spatial superposition analysis of the Internet of Things (IoT) sensing data of the to-be-studied region. A human activity heat map layer is generated based on a target tracking algorithm according to the IoT sensing data and the aerial image data. Through layer analysis of the IoT sensing data and the aerial image data, the efficiency of analyzing the to-be-studied region can be improved, and the recognition efficiency of the to-be-recognized land parcel is improved. Meanwhile, multiple types of data are added, the comprehensiveness of the analysis data is improved, and the recognition accuracy of the to-be-recognized land parcel is improved. Then, idle land recognition is performed on the to-be-recognized land parcel in combination with public procedure data of the to-be-studied region. The idle land recognition result of the to-be-recognized land parcel can be verified and corrected through the public procedure data, and the accuracy of idle land construction development recognition is further improved. BRIEF DESCRIPTION OF DRAWINGS

[0053] Figure 1 FIG. 1 is a flowchart of a GIS-based idle land construction development recognition method according to some embodiments of the present application;

[0054] Figure 2 FIG. 3 is a block diagram of a GIS-based idle land construction development recognition system according to some embodiments of the present application. DETAILED DESCRIPTION

[0055] The embodiments of the present application are described in detail below with reference to the accompanying drawings. Examples of the embodiments are shown in the drawings, in which the same or similar reference numerals represent the same or similar elements throughout. The embodiments described below with reference to the drawings are exemplary and are only used to explain some embodiments of the present application, and cannot be understood as limiting the embodiments of the present application. Based on the embodiments shown in the present application, all other embodiments obtained by those of ordinary skill in the art without creative labor fall within the scope of protection of the present application.

[0056] In the description of the present application, it should be understood that the terms "first" and "second" are only used for description purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" can explicitly or implicitly include at least one of the features. In the description of the present application, unless otherwise explicitly and specifically limited, the meaning of "multiple" and "several" is two or more.

[0057] The construction development or disposal of existing idle land relies on manual patrol and self-reporting, supplemented by simple remote sensing image verification supervision. However, this idle land construction development identification method has obvious defects: manual patrol verification is too laborious and inefficient, and it is difficult to achieve large-scale and normalized supervision; there are cases of not timely reporting in some areas, resulting in a large difference between the increase and disposal speed of idle land; traditional remote sensing technology is limited by image resolution and single data source, and it is difficult to accurately distinguish the true use state of the land. The above defects all lead to the inaccuracy of the existing idle land construction development identification, and how to improve the accuracy of idle land construction development identification is still a difficult problem to be solved in the existing technology.

[0058] Based on the above technical background, see Figure 1 The embodiment of the present application provides a GIS-based idle land construction development identification method, which comprises steps S101 to S104, and each step is as follows:

[0059] Step S101: Multi-source heterogeneous fusion is performed on remote sensing image data and aerial image data of a to-be-researched region to obtain a fused image data set.

[0060] In some embodiments of the present application, the multi-source heterogeneous fusion of the remote sensing image data and the aerial image data of the to-be-researched region to obtain the fused image data set specifically comprises:

[0061] According to the time stamp, the remote sensing image data and the aerial image data are time axis aligned to obtain first remote sensing image data and first aerial image data;

[0062] According to the spatial coordinates, the first remote sensing image data and the first aerial image data are spatial coordinate aligned to obtain second remote sensing image data and second aerial image data;

[0063] The second remote sensing image data and the second aerial image data are fused to obtain the fused image data set.

[0064] In some embodiments of the present application, the time axis alignment according to the time stamp can be a time stamp interpolation method, and the preferred embodiment is a time stamp linear interpolation method.

[0065] In some embodiments of the present application, the spatial coordinate alignment according to the spatial coordinates can be a spatial coordinate projection method, and the preferred embodiment is a three-dimensional rectangular spatial coordinate projection method.

[0066] In some embodiments of the present application, the implementation of fusing the second remote sensing image data and the second aerial image data can be splicing each piece of data in the second remote sensing image data with the data of the corresponding time stamp in the second aerial image data according to the time stamp.

[0067] The present application first aligns the remote sensing image data and the aerial image data according to the time stamp, and then aligns the spatial coordinates according to the spatial coordinates, and further fuses to obtain a fused image data set. Through the two batches of alignment of the time stamp and the spatial coordinates, the multi-source heterogeneous data can be converted into multi-source homogeneous data, which provides favorable conditions for subsequent fusion.

[0068] Step S102: based on a deep convolutional neural network, in combination with a geographic information system, performing land parcel identification on the fused image data set to obtain a land parcel segmentation coordinate set.

[0069] In some embodiments of the present application, the land parcel identification on the fused image data set based on the deep convolutional neural network in combination with the geographic information system to obtain the land parcel segmentation coordinate set specifically includes:

[0070] based on a deep convolutional neural network, performing semantic segmentation on the fused image data set to obtain a land feature classification set;

[0071] According to the land feature classification set, based on an edge detection algorithm, performing feature matching identification on the fused image data set to obtain a land parcel classification image set, and according to the land parcel classification image set, in combination with the provided land parcel spatial coordinate map obtained by the geographic information system, constructing the association between the land parcel and the coordinate to obtain the land parcel segmentation coordinate set.

[0072] In some embodiments of the present application, the edge detection algorithm includes but is not limited to Canny algorithm, Sobel algorithm and Prewitt algorithm, and the preferred implementation is Canny algorithm.

[0073] Generally, those skilled in the art can easily understand that the geographic information system (Geographic Information System) is GIS.

[0074] The present application first performs semantic segmentation on the fused image data set based on a deep convolutional neural network to obtain a land feature classification set, which can separate each land feature from the fused image data set. Then, based on an edge detection algorithm, feature matching identification is performed on the fused image data set, which can identify and segment different types of land parcels according to each land feature. Furthermore, in combination with the provided land parcel spatial coordinate map obtained by the geographic information system, the association between the land parcel and the coordinate is constructed to obtain the land parcel segmentation coordinate set, thereby providing conditions for subsequent analysis and identification.

[0075] Step S103: according to the plot segmentation coordinate set, combined with the Internet of Things sensing data of the region to be studied, spatial superposition analysis is performed to obtain a development state prediction layer corresponding to the region to be studied, and based on a target tracking algorithm, a human activity heat map layer corresponding to the region to be studied is generated according to the Internet of Things sensing data and the aerial image data.

[0076] In some embodiments of the present application, the development state prediction layer corresponding to the region to be studied is obtained by spatial superposition analysis based on the plot segmentation coordinate set and the Internet of Things sensing data of the region to be studied, and the human activity heat map layer corresponding to the region to be studied is generated based on a target tracking algorithm according to the Internet of Things sensing data and the aerial image data, which specifically includes:

[0077] According to the plot segmentation coordinate set, combined with the Internet of Things sensing data of the region to be studied, spatial superposition analysis is performed based on a layer superposition method to obtain the development state prediction layer.

[0078] According to the Internet of Things sensing data, first human activity trajectory data is extracted, and second human activity trajectory data is obtained based on a target tracking algorithm according to the aerial image data.

[0079] Based on the layer superposition method, spatial superposition analysis is performed on the first human activity trajectory data and the second human activity trajectory data to obtain the human activity heat map layer.

[0080] In some embodiments of the present application, the first human activity trajectory data is extracted according to the Internet of Things sensing data, specifically: a plurality of individual activity trajectories are determined according to the use logs of each sensor in the Internet of Things sensing data; and the first human activity trajectory data is extracted according to the plurality of individual activity trajectories.

[0081] According to the plot segmentation coordinate set, combined with the Internet of Things sensing data of the region to be studied, spatial superposition analysis is performed based on a layer superposition method to obtain a development state prediction layer; then first human activity trajectory data is extracted according to the Internet of Things sensing data, and second human activity trajectory data is obtained based on a target tracking algorithm according to the aerial image data, and then spatial superposition analysis is performed based on a layer superposition method to obtain a human activity heat map layer. Through layer analysis of the Internet of Things sensing data and the aerial image data, the efficiency of analyzing the region to be studied can be improved, and the identification efficiency of the plot to be identified can be improved, while multiple types of data are added, the comprehensiveness of the analysis data is improved, and the identification accuracy of the plot to be identified is improved.

[0082] Step S104: according to the construction development state prediction layer and the human activity heat map, in combination with the public procedure data of the region to be studied, idle land identification is performed on the to-be-identified land block in the region to be studied, and an idle land construction development identification result of the to-be-identified land block is obtained.

[0083] In some embodiments of the present application, the idle land identification is performed on the to-be-identified land block in the region to be studied according to the construction development state prediction layer and the human activity heat map in combination with the public procedure data of the region to be studied, and an idle land construction development identification result of the to-be-identified land block is obtained, which specifically includes:

[0084] According to the construction development state prediction layer, the land block distribution type of the to-be-identified land block is determined.

[0085] According to the human activity heat map, a human activity identification result of the to-be-identified land block is determined.

[0086] According to the land block distribution type and the human activity identification result, a first idle land identification result is determined.

[0087] According to the public procedure data of the region to be studied, the first idle land identification result is verified and corrected, and an idle land construction development identification result of the to-be-identified land block is obtained.

[0088] According to the construction development state prediction layer and the human activity heat map, the land block distribution type and the human activity identification result of the to-be-identified land block are determined, and then the first idle land identification result is determined, and then the idle land construction development identification result of the to-be-identified land block is obtained by verification and correction in combination with the public procedure data, which can improve the accuracy of idle land construction development identification.

[0089] Compared with the prior art, the remote sensing image data and the aerial image data of the to-be-studied region are subjected to multi-source heterogeneous fusion to obtain a fused image data set, and then land parcel recognition is performed based on a deep convolutional neural network to obtain a land parcel segmentation coordinate set. Through multi-source heterogeneous fusion of the remote sensing image data and the aerial image data, the comprehensiveness of the analysis data can be improved, and recognition errors caused by a single type of data can be avoided. Then, spatial superposition analysis is performed on the to-be-studied region in combination with Internet of Things (IoT) sensing data to obtain a development status prediction layer. Based on target tracking algorithm, a human activity heat map layer is generated from the IoT sensing data and the aerial image data. Through layer analysis of the IoT sensing data and the aerial image data, the efficiency of analyzing the to-be-studied region can be improved, and the recognition efficiency of the to-be-recognized land parcel can be improved. At the same time, multiple types of data are added, the comprehensiveness of the analysis data is improved, and the recognition accuracy of the to-be-recognized land parcel is improved. Then, in combination with public procedure data of the to-be-studied region, idle land recognition is performed on the to-be-recognized land parcel. The idle land recognition result of the to-be-recognized land parcel can be verified and corrected through the public procedure data, and the accuracy of idle land development recognition is further improved.

[0090] Corresponding to the foregoing method, please refer to Figure 2 The embodiment of the application provides a GIS-based idle land development recognition system, which comprises a multi-source fusion module 210, a land parcel recognition module 220, a layer analysis module 230 and an idle land recognition module 240.

[0091] The multi-source fusion module 210 is configured to perform multi-source heterogeneous fusion on remote sensing image data and aerial image data of a to-be-studied region to obtain a fused image data set.

[0092] The land parcel recognition module 220 is configured to perform land parcel recognition on the fused image data set based on a deep convolutional neural network and in combination with a geographic information system to obtain a land parcel segmentation coordinate set.

[0093] The layer analysis module 230 is configured to perform spatial superposition analysis on the to-be-studied region in combination with IoT sensing data of the to-be-studied region according to the land parcel segmentation coordinate set to obtain a development status prediction layer corresponding to the to-be-studied region, and generate a human activity heat map layer corresponding to the to-be-studied region based on a target tracking algorithm according to the IoT sensing data and the aerial image data.

[0094] The idle land recognition module 240 is configured to perform idle land recognition on a to-be-recognized land parcel in the to-be-studied region in combination with public procedure data of the to-be-studied region according to the development status prediction layer and the human activity heat map layer to obtain an idle land development recognition result of the to-be-recognized land parcel.

[0095] In some embodiments of the present application, the multi-source fusion module 210 comprises a time axis alignment unit, a spatial coordinate alignment unit and a data fusion unit;

[0096] The time axis alignment unit is configured to perform time axis alignment on the remote sensing image data and the aerial video data according to timestamps to obtain first remote sensing image data and first aerial video data;

[0097] The spatial coordinate alignment unit is configured to perform spatial coordinate alignment on the first remote sensing image data and the first aerial video data according to spatial coordinates to obtain second remote sensing image data and second aerial video data;

[0098] The data fusion unit is configured to fuse the second remote sensing image data and the second aerial video data to obtain a fused image data set.

[0099] In some embodiments of the present application, the plot recognition module 220 comprises a semantic segmentation recognition unit and a feature matching recognition unit;

[0100] The semantic segmentation recognition unit is configured to perform semantic segmentation on the fused image data set based on a deep convolutional neural network to obtain a feature classification set of ground objects;

[0101] The feature matching recognition unit is configured to perform feature matching recognition on the fused image data set based on an edge detection algorithm according to the feature classification set of ground objects to obtain a plot classification image set, and to construct an association between plots and coordinates according to the plot classification image set and a spatial coordinate map of supplied plots obtained by a geographic information system to obtain a plot segmentation coordinate set.

[0102] In some embodiments of the present application, the layer analysis module 230 comprises a plot distribution analysis unit, an activity trajectory analysis unit and an activity heat analysis unit;

[0103] The plot distribution analysis unit is configured to perform spatial superposition analysis based on a layer superposition method according to the plot segmentation coordinate set and in combination with Internet of Things (IoT) sensing data of a region to be studied to obtain the under-construction development state prediction layer;

[0104] The activity trajectory analysis unit is configured to extract first human activity trajectory data according to the IoT sensing data, and to obtain second human activity trajectory data based on a target tracking algorithm according to the aerial video data;

[0105] The activity heat analysis unit is configured to perform spatial superposition analysis on the first human activity trajectory data and the second human activity trajectory data based on a layer superposition method to obtain a human activity heat layer.

[0106] In some embodiments of the present application, the idle land identification module 240 comprises a plot type identification unit, a human activity identification unit, an idle land identification unit, and an identification verification correction unit.

[0107] The plot type identification unit is configured to determine the plot distribution type of the to-be-identified plot according to the construction and development state prediction layer.

[0108] The human activity identification unit is configured to determine the human activity identification result of the to-be-identified plot according to the human activity heat map layer.

[0109] The idle land identification unit is configured to determine the first idle land identification result according to the plot distribution type and the human activity identification result.

[0110] The identification verification correction unit is configured to verify and correct the first idle land identification result according to the public procedure data of the to-be-studied region, to obtain the idle land construction and development identification result of the to-be-identified plot.

[0111] The present application performs multi-source heterogeneous fusion on the remote sensing image data and the aerial image data of the to-be-studied region to obtain a fused image data set, and then performs plot identification based on a deep convolutional neural network to obtain a plot segmentation coordinate set. Through multi-source heterogeneous fusion of remote sensing image data and aerial image data, the comprehensiveness of the analysis data can be improved, and the identification error caused by a single type of data can be avoided. Then, the space superposition analysis is performed on the to-be-studied region in combination with the Internet of Things sensing data to obtain a construction and development state prediction layer. Based on the target tracking algorithm, the human activity heat map layer is generated according to the Internet of Things sensing data and the aerial image data. Through layer analysis of the Internet of Things sensing data and the aerial image data, the efficiency of analyzing the to-be-studied region can be improved, and the identification efficiency of the to-be-identified plot can be improved. At the same time, multiple types of data are added, the comprehensiveness of the analysis data is improved, and the identification accuracy of the to-be-identified plot is improved. Then, in combination with the public procedure data of the to-be-studied region, the idle land identification of the to-be-identified plot is performed. The idle land identification result of the to-be-identified plot can be verified and corrected through the public procedure data, and the accuracy of the idle land construction and development identification is further improved.

[0112] It should be understood that the system provided by the embodiments of the present application corresponds to the foregoing method. The GIS-based idle land construction and development identification system provided by the embodiments of the present application can implement the GIS-based idle land construction and development identification method provided by any one of the embodiments of the present application.

[0113] Adaptively, the embodiments of the present application further provide a computer device and a computer readable storage medium.

[0114] The computer device comprises a processor, a memory and a computer program stored in the memory and configured to be executed by the processor;

[0115] The processor implements the GIS-based idle land construction development identification method of the application when executing the computer program.

[0116] The computer readable storage medium stores a plurality of instructions, which are suitable for being loaded by the processor to execute the GIS-based idle land construction development identification method of the application.

[0117] The above is part of the embodiments of the application, which further details the purpose, technical solutions and beneficial effects of the application. It should be clear that the above part of the embodiments of the application cannot be understood as the limitation of the application. It is particularly pointed out that any change, modification, equivalent replacement and variation, etc. made by the person skilled in the art within the spirit and principle of the application shall be included in the protection scope of the application.

Claims

1. A GIS-based method for identifying the commencement of development on idle land, characterized in that, include: Multi-source heterogeneous fusion of remote sensing image data and aerial image data of the study area is performed to obtain a fused image dataset. Based on a deep convolutional neural network and combined with a geographic information system, land parcel identification is performed on the fused image dataset to obtain a set of land parcel segmentation coordinates; Based on the plot segmentation coordinate set, spatial overlay analysis is performed using IoT sensor data of the area under study to obtain a construction and development status prediction layer corresponding to the area under study. Based on the IoT sensor data and the aerial image data, a human activity heat map corresponding to the area under study is generated using a target tracking algorithm. Based on the construction and development status prediction layer and the human activity heat map layer, and combined with the publicly available permit data of the area to be studied, idle land is identified in the area to be identified, and the results of the construction and development identification of the idle land of the area to be identified are obtained.

2. The method for identifying the commencement of development on idle land based on GIS according to claim 1, characterized in that, The remote sensing image data and aerial image data of the area to be studied are subjected to multi-source heterogeneous fusion to obtain a fused image dataset, which specifically includes: Based on the timestamp, the remote sensing image data and the aerial image data are aligned along the time axis to obtain the first remote sensing image data and the first aerial image data. Based on spatial coordinates, the first remote sensing image data and the first aerial image data are aligned to obtain the second remote sensing image data and the second aerial image data. The second remote sensing image data and the second aerial image data are fused to obtain a fused image dataset.

3. The method for identifying the commencement of development on idle land based on GIS according to claim 1, characterized in that, The process of identifying land parcels from the fused image dataset using a deep convolutional neural network and a geographic information system to obtain a set of land parcel segmentation coordinates specifically includes: Based on a deep convolutional neural network, semantic segmentation is performed on the fused image dataset to obtain a set of ground feature classifications; Based on the land feature classification set, feature matching and recognition are performed on the fused image dataset using an edge detection algorithm to obtain a land parcel classification image set. Based on the land parcel classification image set and combined with the spatial coordinate map of the supplied land parcels obtained from the geographic information system, the association between the land parcels and the coordinates is constructed to obtain a land parcel segmentation coordinate set.

4. The method for identifying the commencement of development on idle land based on GIS according to claim 1, characterized in that, The process involves performing spatial overlay analysis based on the plot segmentation coordinate set and IoT sensor data of the area under study to obtain a construction and development status prediction layer for the area under study. Furthermore, based on the IoT sensor data and the aerial imagery data, and using a target tracking algorithm, a human activity heatmap for the area under study is generated. Specifically, this includes: Based on the plot segmentation coordinate set, combined with the IoT sensor data of the area to be studied, spatial overlay analysis is performed based on the layer overlay method to obtain the construction and development status prediction layer. Based on the IoT sensor data, first human activity trajectory data is extracted, and based on the aerial image data and a target tracking algorithm, second human activity trajectory data is obtained. Based on the layer overlay method, the first human activity trajectory data and the second human activity trajectory data are spatially overlaid and analyzed to obtain the human activity heat map.

5. The method for identifying the commencement of development on idle land based on GIS according to claim 1, characterized in that, The process involves identifying idle land in the area under study based on the construction and development status prediction layer and the human activity heat map, combined with publicly available permit data. This process yields the results of identifying the construction and development of idle land in the area under study. Specifically, this includes: Based on the construction and development status prediction layer, determine the land parcel distribution type of the land parcels to be identified; Based on the aforementioned human activity heat map, determine the human activity identification results for the plot to be identified; Based on the land parcel distribution type and the human activity identification results, the first idle land identification result is determined; Based on publicly available procedural data for the area under study, the first idle land identification result is verified and corrected to obtain the idle land development identification result for the plot to be identified.

6. A GIS-based system for identifying the commencement of development on idle land, characterized in that, It includes a multi-source fusion module, a land parcel identification module, a layer analysis module, and an idle land identification module; The multi-source fusion module is used to perform multi-source heterogeneous fusion of remote sensing image data and aerial image data of the area to be studied to obtain a fused image dataset. The land parcel identification module is used to identify land parcels based on the fused image dataset using a deep convolutional neural network and in conjunction with a geographic information system, thereby obtaining a set of land parcel segmentation coordinates. The layer analysis module is used to perform spatial overlay analysis based on the plot segmentation coordinate set and the IoT sensor data of the area under study to obtain the construction and development status prediction layer corresponding to the area under study, and to generate the human activity heat map layer corresponding to the area under study based on the IoT sensor data and the aerial image data and the target tracking algorithm. The idle land identification module is used to identify idle land in the area to be identified based on the construction and development status prediction layer and the human activity heat map layer, combined with the publicly available permit data of the area to be studied, and to obtain the idle land construction and development identification results of the plots to be identified.

7. A GIS-based system for identifying the commencement of development on idle land according to claim 6, characterized in that, The multi-source fusion module includes a time axis alignment unit, a spatial coordinate alignment unit, and a data fusion unit; The time axis alignment unit is used to align the remote sensing image data and the aerial image data according to the timestamp to obtain the first remote sensing image data and the first aerial image data. The spatial coordinate alignment unit is used to align the first remote sensing image data and the first aerial image data according to spatial coordinates to obtain the second remote sensing image data and the second aerial image data. The data fusion unit is used to fuse the second remote sensing image data and the second aerial image data to obtain a fused image dataset.

8. A GIS-based system for identifying the commencement of development on idle land according to claim 6, characterized in that, The land parcel identification module includes a semantic segmentation identification unit and a feature matching identification unit; The semantic segmentation and recognition unit is used to perform semantic segmentation on the fused image dataset based on a deep convolutional neural network to obtain a set of ground feature classifications. The feature matching and recognition unit is used to perform feature matching and recognition on the fused image dataset based on the feature classification set and the edge detection algorithm to obtain a land parcel classification image set. Based on the land parcel classification image set and the spatial coordinate map of the supplied land parcels obtained by the geographic information system, the unit constructs the association between the land parcels and the coordinates to obtain a land parcel segmentation coordinate set.

9. A GIS-based system for identifying the commencement of development on idle land according to claim 6, characterized in that, The layer analysis module includes a land parcel distribution analysis unit, an activity trajectory analysis unit, and an activity thermal analysis unit; The land parcel distribution analysis unit is used to perform spatial overlay analysis based on the land parcel segmentation coordinate set and the IoT sensor data of the area to be studied, and to obtain the construction and development status prediction layer based on the layer overlay method. The activity trajectory analysis unit is used to extract first human activity trajectory data based on the IoT sensor data, and to obtain second human activity trajectory data based on the aerial image data and a target tracking algorithm. The activity thermal analysis unit is used to perform spatial overlay analysis on the first human activity trajectory data and the second human activity trajectory data based on the layer overlay method to obtain the human activity thermal layer.

10. A GIS-based system for identifying the commencement of development on idle land according to claim 6, characterized in that, The idle land identification module includes a land parcel type identification unit, a human activity identification unit, an idle land identification unit, and an identification verification and correction unit; The land parcel type identification unit is used to determine the land parcel distribution type of the land parcel to be identified based on the construction and development status prediction layer; The human activity identification unit is used to determine the human activity identification result of the plot to be identified based on the human activity heat map. The idle land identification unit is used to determine the first idle land identification result based on the land parcel distribution type and the human activity identification result; The identification verification and correction unit is used to verify and correct the first idle land identification result based on the publicly available procedural data of the area to be studied, so as to obtain the idle land development identification result of the plot to be identified.

Images