A remote sensing monitoring method and system for the current status of development and construction land for linear engineering projects based on boundary constraints

Abstract

The present invention discloses a remote sensing monitoring method and system for the current status of development and construction land for linear engineering projects based on boundary constraints, belonging to the field of remote sensing geographic information technology; including: obtaining project approval information; generating boundary constraints according to the project approval information; obtaining remote sensing images; inputting remote sensing images into an interpretation model for interpretation to obtain a development and construction land area; post-processing the development and construction land area to obtain a development and construction map patch; processing the development and construction map patch according to the boundary constraints to obtain a final development and construction current status range. The present invention is based on multi-period high-resolution remote sensing images and approval boundary information of highway and railway linear engineering projects, and uses computer deep learning and modeling technology to effectively improve the work efficiency of extracting development and construction boundary range information of highway and railway linear engineering projects, and also provides time-effective monitoring data for the implementation and management of approval land for major construction projects.

Description

Technical Field

[0001] The invention relates to the field of remote sensing geographic information technology, and in particular to a remote sensing monitoring method and system for the current status of development and construction land for a linear engineering project based on boundary constraints. Background Art

[0002] Major linear engineering construction projects such as highways and railways approved annually in my country account for about 80% of the total approved projects. The implementation status of the development and construction land of these approved linear engineering projects is extremely important for both land use and investment supervision, so how to quickly and accurately monitor and extract the land development boundary information of such projects is of great significance.

[0003] At present, based on high-resolution remote sensing images, linear information such as roads can be identified and extracted through manual visual recognition or existing remote sensing image intelligent extraction model methods. However, manual methods require a lot of manpower and cannot meet the timeliness requirements of rapid monitoring.

[0004] The boundary information extracted by the existing intelligent extraction model is quite different from the real boundary and cannot meet the monitoring accuracy and quality requirements. Summary of the invention

[0005] The purpose of the present invention is to provide an efficient remote sensing monitoring method and system for the current status of development and construction land of a linear engineering project based on boundary constraints.

[0006] In order to solve the above technical problems, the present invention provides a remote sensing monitoring method for the current status of development and construction land of a linear engineering project based on boundary constraints, comprising the following steps:

[0007] Obtain project approval information;

[0008] Generate boundary constraints based on project approval information;

[0009] Acquisition of remote sensing images;

[0010] Input the remote sensing image into the interpretation model for interpretation to obtain the development and construction land area;

[0011] Post-process the development and construction land area to obtain the development and construction map patch;

[0012] According to the boundary constraints, the development and construction map patches are processed to obtain the final development and construction status scope.

[0013] Preferably, before the remote sensing image is input into the interpretation model for interpretation, the following steps are also included:

[0014] Based on boundary constraints, the remote sensing images are filtered for invalid ranges and cropped with the minimum bounding rectangle.

[0015] Preferably, the training method of the interpretation model comprises the following steps, specifically comprising the following steps:

[0016] Construct a sample dataset;

[0017] Based on Transformer, the interpretation model training is carried out under the PyTorch framework.

[0018] Preferably, constructing a sample data set specifically includes the following steps:

[0019] Determine the land use type based on the project approval information; the land use type includes typical roads and railway project roads;

[0020] According to the land use type, sample features are established to obtain the corresponding remote sensing images with completed sample features and uncompleted remote sensing images;

[0021] The historical remote sensing images with labeled sample features are input into the data set model for training to obtain a trained data set model;

[0022] The unlabeled remote sensing images are input into the trained dataset model for labeling to obtain the training sample dataset.

[0023] Preferably, the sample features include roadbed, bridge, interchange and slope.

[0024] Preferably, the number of training iterations of the interpretation model is set to 100 epochs, the initialization learning rate parameter is set to 0.001, cosine annealing is used to reduce the learning rate, cross entropy is used to calculate the loss, and the model optimization uses the adm optimizer.

[0025] Preferably, the post-processing includes small object filtering, short-distance disconnection repair and edge smoothing.

[0026] Preferably, post-processing the development and construction land area to obtain the development and construction map patch specifically includes the following steps:

[0027] Searching for a continuous pixel area with the same color value from the development and construction land area to obtain a processing target; the processing target is a small target or a hole;

[0028] Use the structural element to perform expansion or erosion operations on the processing target to obtain a filled image; the expansion operation adds the pixel points of the neighboring area in the development and construction land area to the processing target, and the erosion operation adds the pixel points of the processing target in the development and construction land area to the neighboring area;

[0029] Based on the filled image and the development and construction land area, edge detection is performed to obtain the irregular image edge, and the irregular image edge is corrected to obtain an edge-corrected image;

[0030] Binarization is performed on the edge-corrected image to obtain a binarized image;

[0031] The size of the connected area in the binarized image is calculated, and the connected area is filtered according to the size of the connected area to obtain the development and construction map spots.

[0032] Preferably, the development and construction map patches are processed according to the boundary constraints to obtain the final development and construction status range, which specifically includes the following steps:

[0033] Conduct an intersection analysis on boundary constraints and development and construction map patches to obtain the final development and construction status scope.

[0034] The present invention also provides a remote sensing monitoring system for the current status of development and construction land for linear engineering projects based on boundary constraints, comprising:

[0035] The first acquisition module is used to obtain project approval information;

[0036] A boundary constraint condition generation module is used to generate boundary constraint conditions based on project approval information;

[0037] The second acquisition module is used to acquire remote sensing images;

[0038] The interpretation module is used to input the remote sensing image into the interpretation model for interpretation to obtain the development and construction land area;

[0039] A post-processing module is used to post-process the development and construction land area to obtain the development and construction map spots;

[0040] The intersection screening module is used to process the development and construction map patches according to the boundary constraints to obtain the final development and construction status scope.

[0041] Compared with the prior art, the present invention has the following beneficial effects:

[0042] Based on multi-period high-resolution remote sensing images and approval boundary information of highway and railway linear engineering projects, the present invention uses computer deep learning and modeling technology to effectively improve the work efficiency of extracting development and construction boundary range information of highway and railway linear engineering projects, and also provides timely monitoring data for the implementation management of land use approval for major construction projects. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] The specific implementation modes of the present invention are further described in detail below with reference to the accompanying drawings.

[0044] Figure 1 The present invention is a flow chart of a remote sensing monitoring method for the current status of development and construction land of a linear engineering project based on boundary constraints. DETAILED DESCRIPTION

[0045] Many specific details are set forth in the following description to facilitate a full understanding of the present invention. However, the present invention can be implemented in many other ways than those described herein, and those skilled in the art can make similar generalizations without violating the connotation of the present invention, so the present invention is not limited to the specific implementation disclosed below.

[0046] The terms used in one or more embodiments of this specification are only for the purpose of describing specific embodiments, and are not intended to limit one or more embodiments of this specification. The singular forms of "a", "said" and "the" used in one or more embodiments of this specification and the appended claims are also intended to include plural forms, unless the context clearly indicates other meanings. It should also be understood that the term "and / or" used in one or more embodiments of this specification refers to and includes any or all possible combinations of one or more associated listed items.

[0047] It should be understood that although the terms first, second, etc. may be used to describe various information in one or more embodiments of this specification, this information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of one or more embodiments of this specification, the first may also be referred to as the second, and similarly, the second may also be referred to as the first. Depending on the context, the word "if" as used herein may be interpreted as "at the time of" or "when" or "in response to determining".

[0048] The present invention is further described in detail below in conjunction with the accompanying drawings:

[0049] The present invention provides a remote sensing monitoring method for the current status of development and construction land of a linear engineering project based on boundary constraints, comprising the following steps:

[0050] Obtain project approval information;

[0051] Generate boundary constraints based on project approval information;

[0052] Acquisition of remote sensing images;

[0053] Input the remote sensing image into the interpretation model for interpretation to obtain the development and construction land area;

[0054] Post-process the development and construction land area to obtain the development and construction map patch;

[0055] According to the boundary constraints, the development and construction map patches are processed to obtain the final development and construction status scope.

[0056] Preferably, before the remote sensing image is input into the interpretation model for interpretation, the following steps are also included:

[0057] Based on boundary constraints, the remote sensing images are filtered for invalid ranges and cropped with the minimum bounding rectangle.

[0058] Preferably, the training method of the interpretation model comprises the following steps, specifically comprising the following steps:

[0059] Construct a sample dataset;

[0060] Based on Transformer, the interpretation model training is carried out under the PyTorch framework.

[0061] Preferably, constructing a sample data set specifically includes the following steps:

[0062] Determine the land use type based on the project approval information; the land use type includes typical roads and railway project roads;

[0063] According to the land use type, sample features are established to obtain the corresponding remote sensing images with completed sample features and uncompleted remote sensing images;

[0064] The historical remote sensing images with labeled sample features are input into the data set model for training to obtain a trained data set model;

[0065] The unlabeled remote sensing images are input into the trained dataset model for labeling to obtain the training sample dataset.

[0066] Preferably, the sample features include roadbed, bridge, interchange and slope.

[0067] Preferably, the number of training iterations of the interpretation model is set to 100 epochs, the initialization learning rate parameter is set to 0.001, cosine annealing is used to reduce the learning rate, cross entropy is used to calculate the loss, and the model optimization uses the adm optimizer.

[0068] Preferably, the post-processing includes small object filtering, short-distance disconnection repair and edge smoothing.

[0069] Preferably, post-processing the development and construction land area to obtain the development and construction map patch specifically includes the following steps:

[0070] Searching for a continuous pixel area with the same color value from the development and construction land area to obtain a processing target; the processing target is a small target or a hole;

[0071] Use the structural element to perform expansion or erosion operations on the processing target to obtain a filled image; the expansion operation adds the pixel points of the neighboring area in the development and construction land area to the processing target, and the erosion operation adds the pixel points of the processing target in the development and construction land area to the neighboring area;

[0072] Based on the filled image and the development and construction land area, edge detection is performed to obtain the irregular image edge, and the irregular image edge is corrected to obtain an edge-corrected image;

[0073] Binarization is performed on the edge-corrected image to obtain a binarized image;

[0074] The size of the connected area in the binarized image is calculated, and the connected area is filtered according to the size of the connected area to obtain the development and construction map spots.

[0075] Preferably, the development and construction map patches are processed according to the boundary constraints to obtain the final development and construction status range, which specifically includes the following steps:

[0076] Conduct an intersection analysis on boundary constraints and development and construction map patches to obtain the final development and construction status scope.

[0077] The present invention also provides a remote sensing monitoring system for the current status of development and construction land for linear engineering projects based on boundary constraints, comprising:

[0078] The first acquisition module is used to obtain project approval information;

[0079] A boundary constraint condition generation module is used to generate boundary constraint conditions based on project approval information;

[0080] The second acquisition module is used to acquire remote sensing images;

[0081] The interpretation module is used to input the remote sensing image into the interpretation model for interpretation to obtain the development and construction land area;

[0082] A post-processing module is used to post-process the development and construction land area to obtain the development and construction map spots;

[0083] The intersection screening module is used to process the development and construction map patches according to the boundary constraints to obtain the final development and construction status scope.

[0084] In order to better illustrate the technical effect of the present invention, the present invention provides the following specific embodiments to illustrate the above technical process:

[0085] Embodiment 1: A remote sensing monitoring method for the current status of development and construction land of a linear engineering project based on boundary constraints, such as Figure 1 As shown, the following steps are included:

[0086] Step 1: Define boundary constraints

[0087] In order to more accurately extract the scope of highway and railway development and construction land and avoid interference factors outside the boundary (shadow, lighting, etc.), the approval scope of highway and railway linear engineering projects is used as a boundary constraint condition, which can greatly improve the efficiency of image segmentation and the stability of extracted elements. This boundary constraint condition is applied to the entire process of extracting the development and construction scope of major construction projects - highway and railway linear projects.

[0088] Step 2: Establish the characteristics of land for development and construction of highway and railway linear engineering projects

[0089] According to the land use monitoring results of linear projects such as highways and railways, typical highways and railway project roads are selected, and sample features are established based on functional zoning, including roadbed, bridges, interchanges, slopes, etc., and a sample training set is constructed based on this feature.

[0090] Step 3: Construction of a sample dataset of land for development and construction of highway and railway linear engineering projects under boundary constraints

[0091] Based on the sample characteristics, sample annotation tools are used to perform sample annotation around the project approval boundary impact range to implement sample annotation constraints and obtain the sample annotated image.

[0092] This paper proposes an active learning method for generating remote sensing image samples of linear construction projects, and uses a combination of manual annotation and model autonomous annotation to create a sample dataset:

[0093] First, we manually collected some sample data of land used for development and construction of highway and railway linear engineering projects, and trained the initial data set model. We used the model to evaluate the unlabeled data and labeled the unlabeled data. Then, we manually checked the labeled data in the model, corrected the mislabeled data, and added the corrected data to the original data set. We repeated the above operations and gradually iterated the data set. We sliced ​​the highway and railway construction land samples into 512x512 slices to construct the final sample data set.

[0094] Step 4: Training of interpretation model for highway and railway development and construction land

[0095] Based on Transformer, the interpretation model training is carried out under the PyTorch framework: the interpretation of highway and railway development and construction land is carried out, and the model parameter configuration is appropriately adjusted according to the training results, such as learning rate, number of training rounds, etc., to improve the accuracy of identifying development and construction land for highway and railway linear engineering projects.

[0096] During the training process, the number of training iterations was set to 100 epochs, the initial learning rate parameter was set to 0.001, cosine annealing was used to reduce the learning rate, cross entropy was used to calculate the loss, and the adm optimizer was used for model optimization. The model was evaluated based on the two evaluation indicators of accuracy and recall. The optimal model for training was selected based on the comprehensive evaluation results of the indicators.

[0097] Accuracy: The proportion of all correct predictions (positive and negative) to the total. The calculation formula is as follows:

[0098] Accuracy=(TP+TN) / (TP+TN+FP+FN)

[0099] Recall: The ratio of correctly predicted positive results to all actual positive results. The calculation formula is as follows:

[0100] Recall=TP / (TP+FN)

[0101] TP (True Positive): A correct positive example, an instance is a positive class and is also judged to be a positive class;

[0102] FN (False Negative): False negative examples, missed reports, which are positive examples but are judged as false examples.

[0103] FP (False Positive): False positive example, false alarm, which is false class but judged as positive class;

[0104] TN (True Negative): A correct counterexample, an instance is a false class and is also judged to be a false class.

[0105] Step 5: Image interpretation under boundary constraints

[0106] According to the approval constraint range of the given highway and railway linear engineering project, the invalid range of the image is filtered and the minimum enclosing rectangle is cropped. The cropped image is then input into the trained model. The interpretation program reads the image to be analyzed, and performs inference and interpretation to extract the development and construction land of the highway and railway linear engineering project, thereby realizing the calculation of the scope constraint.

[0107] For example: when interpreting the image of the major highway and railway linear construction project in Hongta District, it is necessary to input the approval scope of the major highway and railway linear construction project in Hongta District as the image analysis and calculation scope to achieve effective image range cropping.

[0108] Step 6: Post-processing of the algorithm for extracting the results of the interpretation model

[0109] The results extracted by the interpretation model are finally processed by post-processing algorithms such as small target filtering, short-distance disconnection repair, and edge smoothing. The implementation principle of filtering small targets and filling holes in the result post-processing is mainly based on image processing technology, involving operations such as expansion, corrosion, opening and closing operations.

[0110] The specific steps are as follows:

[0111] Identify small objects or holes: Identify small objects or holes that need to be filled through image analysis. This can be done by finding continuous pixel areas with the same color value.

[0112] Morphological operations: Use structural elements to dilate and erode images to fill or remove small objects. Dilation can add adjacent background points to an object, while erosion does the opposite, removing small objects or disconnecting two connected objects.

[0113] Edge detection: After filling the holes, edge detection may be required to correct the edge irregularities caused by the filling process. This can be done by calculating the difference between the original image and the filled image.

[0114] Cleaning small objects: In the binarized image, small objects can be identified and removed by calculating the size of the connected regions. This is usually done by using a label matrix and counting the number of pixels for each label.

[0115] Iterative optimization: For complex images, it may be necessary to iterate the above steps multiple times to ensure that all small objects are removed and holes are properly filled.

[0116] Step 7: Lock the extraction range under boundary constraints

[0117] The smoothed and other post-processed image patches are again subjected to spatial intersection screening using the project approval scope boundaries to implement result screening constraints, lock in the current scope of development and construction of the final highway and railway linear engineering project, and implement extraction result screening constraints.

[0118] For example: after extracting the construction scope of major highway and railway linear projects in Hongta District, the program obtains this scope and the project approval scope, outputs the results after intersection analysis, and the final extraction results fall within the project scope of Hongta District.

[0119] In the several embodiments provided by the present invention, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of the modules, modules or units is only a logical function division, and there may be other division methods in actual implementation, for example, multiple units, modules or components can be combined or integrated into another device, or some features can be ignored or not executed.

[0120] The units may or may not be physically separated, and the components displayed as units may be one physical unit or multiple physical units, that is, they may be located in one place or distributed in multiple different places. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0121] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0122] In particular, according to the embodiments disclosed in the present invention, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the embodiments disclosed in the present invention include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through the communication part, and / or installed from a removable medium. When the computer program is executed by the central processing unit (CPU), the above-mentioned functions defined in the method of the present invention are executed. It should be noted that the above-mentioned computer-readable medium of the present invention can be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium can be, for example, but not limited to, a system, device or device of an electrical, magnetic, optical, electromagnetic, infrared segment, or semiconductor, or any combination of the above.

[0123] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present invention. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs the specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0124] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions within the technical scope disclosed by the present invention should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A remote sensing monitoring method for the current status of development and construction land for linear engineering projects based on boundary constraints, characterized in that: The following steps are involved: Obtain project approval information; Generate boundary constraints based on project approval information; Acquisition of remote sensing images; Input the remote sensing image into the interpretation model for interpretation to obtain the development and construction land area; Finding a continuous pixel area with the same color value from the development and construction land area to obtain a processing target; the processing target is a small target or a hole; Use the structural element to dilate or erode the processing target to obtain a filled image; The expansion operation adds the pixel points of the neighboring area in the development and construction land area to the processing target, and the erosion operation adds the pixel points of the processing target in the development and construction land area to the neighboring area; Perform edge detection on the filled image and the development and construction land area to obtain irregular image edges, and correct the irregular image edges to obtain an edge-corrected image; Binarization is performed on the edge-corrected image to obtain a binarized image; Calculate the size of the connected area in the binarized image, and filter the connected area according to the size of the connected area to obtain the development and construction map spots; According to the boundary constraints, the development and construction map patches are processed to obtain the final development and construction status scope.

2. The method for remote sensing monitoring of the current status of development and construction land for linear engineering projects based on boundary constraints according to claim 1 is characterized in that: Before the remote sensing image is input into the interpretation model for interpretation, the following steps are also included: Based on boundary constraints, the remote sensing images are filtered for invalid ranges and cropped with the minimum bounding rectangle.

3. The remote sensing monitoring method for the current status of linear engineering project development and construction land based on boundary constraints according to claim 2 is characterized in that: The training method of the interpretation model comprises the following steps: The specific steps include: Construct a sample dataset; Based on Transformer, the interpretation model training is carried out under the PyTorch framework.

4. The method for remote sensing monitoring of the current status of linear engineering project development and construction land based on boundary constraints according to claim 3 is characterized in that: Constructing a sample data set includes the following steps: Determine the land use type based on the project approval information; the land use type includes typical roads and railway project roads; According to the land use type, sample features are established to obtain the corresponding remote sensing images with completed sample features and uncompleted remote sensing images; The historical remote sensing images with labeled sample features are input into the data set model for training to obtain a trained data set model; The unlabeled remote sensing images are input into the trained dataset model for labeling to obtain the training sample dataset.

5. The remote sensing monitoring method for the current status of linear engineering project development and construction land based on boundary constraints according to claim 4 is characterized by: The sample features include roadbed, bridges, interchanges and slopes.

6. The method for remote sensing monitoring of the current status of development and construction land for linear engineering projects based on boundary constraints according to claim 5 is characterized by: The training iteration number of the interpretation model is set to 100 epochs, the initial learning rate parameter is set to 0.001, cosine annealing is used to reduce the learning rate, cross entropy is used to calculate the loss, and the adm optimizer is used for model optimization.

7. The method for remote sensing monitoring of the current status of linear engineering project development and construction land based on boundary constraints according to claim 1 is characterized in that: According to the boundary constraints, the development and construction map patches are processed to obtain the final development and construction status scope, which specifically includes the following steps: Conduct an intersection analysis on boundary constraints and development and construction map patches to obtain the final development and construction status scope.

8. A remote sensing monitoring system for the current status of linear engineering project development and construction land based on boundary constraints, used to implement the remote sensing monitoring method for the current status of linear engineering project development and construction land based on boundary constraints as described in any one of claims 1 to 7, characterized in that: include: The first acquisition module is used to obtain project approval information; A boundary constraint condition generation module is used to generate boundary constraint conditions based on project approval information; The second acquisition module is used to acquire remote sensing images; The interpretation module is used to input the remote sensing image into the interpretation model for interpretation to obtain the development and construction land area; A post-processing module is used to post-process the development and construction land area to obtain the development and construction map spots; The intersection screening module is used to process the development and construction map patches according to the boundary constraints to obtain the final development and construction status scope.

Images