National storage forest field data rechecking method

Abstract

The invention discloses a national storage forest field data rechecking method, and belongs to the technical field of national storage forest data rechecking, the national storage forest field data rechecking method comprises the steps of data preparation, data processing, data analysis, data rechecking and the like, and through high-precision unmanned aerial vehicle image acquisition, an advanced data analysis algorithm and a strict data rechecking process, the data of the national storage forest field data can be rechecked. And the data accuracy reaches a centimeter level. For example, in subcompartment boundary re-checking, deviation is controlled within an extremely small range, it is ensured that construction design is highly matched with the actual situation, reworking and resource waste caused by data errors are reduced, the resource utilization rate is increased, afforestation quality is guaranteed, and the cost per mu of a traditional construction method is reduced to 40 yuan from cost comparison. In the aspect of manpower, the unmanned aerial vehicle assists in reducing on-site exploration manpower demands; in terms of material and machinery cost, unmanned aerial vehicle related consumables and equipment depreciation maintenance cost are lower than traditional measurement equipment, and a large amount of funds are saved for projects.

Description

Technical Field

[0001] The present invention relates to the technical field of national forest reserve data review, and in particular to a national forest reserve field data review method. Background Art

[0002] In the national reserve forest construction project, field data review is crucial. The traditional manual review method has many drawbacks, is extremely inefficient, and consumes a lot of manpower and time. When faced with a large area of ​​reserve forest, if manual review is used, facing such a large area, the staff needs to conduct field surveys, measurements and records in small groups one by one. Not only is the workload huge, but it is also easily affected by factors such as terrain and weather, making it difficult to obtain data. Moreover, manual review cannot guarantee the comprehensiveness and accuracy of the data, and omissions and errors are prone to occur, which brings many uncertainties to subsequent construction. For example, it may cause the construction design to be inconsistent with the actual terrain, affecting the afforestation effect and resource allocation efficiency.

[0003] Traditional manual review mainly relies on staff to carry measuring tools (such as total stations, rangefinders, GPS handheld locators, etc.) to conduct field measurements and records in the reserve forest area. For the boundaries of small classes, staff determine the coordinates of boundary points on the spot, use measuring tools to measure distances and angles to demarcate the boundaries, and manually record them in paper forms or spreadsheets. The records include key information such as the boundary coordinates of the small class, tree species distribution, tree grade, average breast diameter, and average tree height. In the data collation and analysis stage, it mainly relies on manual summary and simple calculation of the recorded data, lacking efficient data processing and analysis methods. This method is inefficient and requires a lot of manpower and time costs in large-scale reserve forest projects. For example, in a project of similar scale, it may take weeks or even months to complete the field data review, and over time, factors such as staff fatigue may cause the data accuracy to decrease. Therefore, traditional manual review has the following disadvantages:

[0004] (1) Low efficiency: For example, in the Huaining County project mentioned above, manual review requires a lot of manpower and is easily affected by the environment, which takes a long time and slows down the overall progress of the project. In areas with complex terrain, workers have difficulty in moving, which further reduces work efficiency.

[0005] (2) Poor accuracy: Manual measurement and recording are prone to human errors, such as boundary point measurement deviation, data recording errors, etc. In addition, during the data aggregation and analysis process, manual calculation and processing are difficult to ensure data consistency and accuracy, which may lead to construction design based on erroneous data and affect afforestation quality. Summary of the invention

[0006] In view of the above problems in the prior art, the present invention provides a method for reviewing field data of national reserve forests, which solves the problems of low efficiency and poor accuracy in traditional manual review.

[0007] In order to achieve the above-mentioned invention object, the technical solution adopted by the present invention is as follows: a method for reviewing the field data of national reserve forests is provided, which comprises:

[0008] Step 1. Data preparation: Obtain satellite map data, drone aerial photography data and existing reserve forest field data of the project area; satellite map data includes satellite maps of the project area; drone aerial photography data includes drone aerial images of the project area; existing reserve forest field data includes field survey maps of the project area;

[0009] Step 2, data processing: Select road intersections and building corners as control points in drone aerial images and satellite maps, convert drone image coordinates into satellite map coordinates, align drone images by comparing control point position deviations, overlay the aligned drone aerial images onto the satellite map layer to obtain overlaid images, and divide sub-class boundaries and tree species distribution in the overlaid images based on existing reserve forest field data;

[0010] Step 3, data analysis: Use object-oriented classification algorithm combined with multi-scale segmentation technology and feature weighting strategy to select multiple representative sample areas in the superimposed images. The multiple sample areas cover a variety of tree species and non-vegetation types, and record sample information; use confusion matrix method to verify the classification accuracy of multiple sample areas, ensure that the overall vegetation and tree species in multiple sample areas are not lower than the preset ratio, and reclassify or manually correct the substandard areas; use boundary analysis tools to extract the sub-class boundaries in the superimposed images and compare and adjust them with the field survey map. When the error is large, manually correct it with drone aerial images, calculate forest growth indicators with field data, evaluate growth trends and influencing factors with time series analysis, and analyze health status through image spectrum and texture features; process topographic map data in satellite maps to generate slope and aspect thematic maps, identify steep slope areas, use buffer analysis and proximity analysis to calculate the distance and spatial distribution characteristics between sub-classes and surrounding features, and provide suggestions for afforestation design;

[0011] Step 4. Data review: Use edge detection algorithm to automatically extract sub-class boundaries and compare with Aowei map. The deviation of sub-class boundaries shall not exceed ±5 meters, and the overlap degree shall be more than 95% to be qualified; use multi-spectral or hyperspectral characteristics of drone images combined with machine learning algorithms to review tree species with an accuracy of not less than 90%, and conduct on-site inspections of representative areas to obtain analysis results; visually compare analysis results with field survey maps and attribute information; calculate the error rate of sub-class boundaries, tree species and growth conditions, and use Kappa coefficient or Jaccard similarity index to evaluate consistency, and verify areas with large differences, blurred boundaries and important ecological nodes.

[0012] Furthermore, in step 1, the method for obtaining satellite map data of the project area is: obtaining the latest high-definition satellite map data of the project area from the Aowei map platform, the latest high-definition satellite map data includes high-resolution satellite images with a resolution of not less than 0.5m in TIF or GeoTIFF format and vector topographic maps in Shapefile format, the vector topographic maps including contour lines, water systems and roads.

[0013] Furthermore, in step 1, the method for obtaining UAV aerial photography data of the project area is: select a UAV of model DJIMini2SE, and let the UAV fly according to a route drawn in advance, the UAV is equipped with a high-resolution camera, the high-resolution camera sets the exposure mode to automatically or manually adjust to a suitable exposure value, and enables the distortion correction function, sets the forward overlap to not less than 80%, and the lateral overlap to not less than 70%; the high-resolution camera obtains UAV aerial photography data of the project area.

[0014] Furthermore, in step 1, the method for obtaining the field data of the existing reserve forests in the project area is as follows: first, the completed field survey data is comprehensively summarized, and the summarized data is converted from the original format of Excel and CSV to the Shapefile format using the Data Interoperability Tool of ArcGIS to ensure that the fields correspond; then, the integrity of the summarized data is checked using GIS software to check whether the small and medium classes in the summarized data of the project are complete and whether the attribute information is complete, and whether the boundaries of the small classes are overlapped or missing through spatial analysis, and the closure of the small class boundaries and the filling of the attribute fields are verified using the topology checking tool, and the problematic data is traced and corrected in a timely manner.

[0015] Furthermore, in step 1, in areas within the project area where the height difference does not exceed 50 meters, the UAV's flight altitude is between 50 and 150 meters; in areas within the project area where the slope exceeds 30°, the UAV's flight altitude is greater than 100 meters; the UAV adjusts its flight direction according to the solar azimuth and season to reduce shadows, identifies tall trees and bushes, and sets a suitable flight altitude.

[0016] Furthermore, in step 1, before the UAV takes off, the route is drawn on the Aowei map using DJI Pilot software based on the CAD coordinates to ensure at least 5% overlap of flights inside and outside the squad boundary, 80% overlap in heading, 60% overlap in lateral direction, and at least three flat, open take-off and landing points without electromagnetic interference are selected.

[0017] Further, in step 3, the number of the plurality of regions is greater than 150, and the size of each region is 50 m×50 m;

[0018] Use ArcGIS "Boundary Analysis Tool" to extract class boundaries; use ArcGIS "KML" tool to process vector topographic maps to generate slope and aspect thematic maps.

[0019] Furthermore, in step 4, ArcGIS and QGIS statistical analysis modules were used to calculate the error rates of subclass boundaries, tree species, and growth conditions.

[0020] Compared with the existing traditional manual review, the beneficial effects of the present invention are:

[0021] 1. Improve accuracy: Through high-precision drone image acquisition, advanced data analysis algorithms and strict data review process, data accuracy reaches the centimeter level. For example, in the small class boundary review, the deviation is controlled within a very small range, ensuring that the construction design is highly consistent with the actual situation, reducing rework and resource waste caused by data errors, improving resource utilization, and ensuring afforestation quality.

[0022] 2. Reduce costs: From the perspective of cost comparison, the traditional method costs 60 yuan per mu, while this method reduces it to 40 yuan. In terms of manpower, drone assistance reduces the demand for manpower for field surveys; in terms of material and machinery costs, the depreciation and maintenance costs of drone-related consumables and equipment are lower than those of traditional measurement equipment; although other costs include drone operator training, the overall cost is still significantly reduced, saving a lot of money for the project.

[0023] 3. Optimize resource allocation: Based on accurate data collection and analysis, reasonable allocation of construction resources can be achieved. For example, the best operation route and machinery configuration plan can be determined according to the terrain and forest conditions of the small class, avoiding unreasonable allocation of resources in different areas and links, improving resource utilization efficiency, and reducing manpower and material costs.

[0024] 4. Enhance risk early warning capabilities: During the construction process, through continuous data monitoring and analysis, risks such as fire, pests and diseases can be prevented in advance. For example, by using real-time updates of satellite maps and drone images, abnormal vegetation areas can be discovered in a timely manner, and corresponding measures can be taken to reduce loss costs and ensure the ecological safety of reserve forests.

[0025] 5. Improve management efficiency: Visual management of the entire process, from data preparation to output, each link can be clearly controlled. Through the dynamic adjustment mechanism, the personnel and machinery layout can be optimized in real time according to the construction progress and feedback, reducing management costs, ensuring the efficient and orderly progress of the project, and accelerating the construction progress. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 This is a flow chart of a method for reviewing field data of national forest reserves. DETAILED DESCRIPTION

[0027] The specific implementation modes of the present invention are described below so that those skilled in the art can understand the present invention. However, it should be clear that the present invention is not limited to the scope of the specific implementation modes. For those of ordinary skill in the art, as long as various changes are within the spirit and scope of the present invention as defined and determined by the attached claims, these changes are obvious, and all inventions and creations utilizing the concept of the present invention are protected.

[0028] like Figure 1 As shown, the present invention provides a method for reviewing the field data of national reserve forests, which comprises:

[0029] Step 1. Data preparation: Obtain satellite map data, drone aerial photography data and existing reserve forest field data of the project area; satellite map data includes satellite maps of the project area; drone aerial photography data includes drone aerial images of the project area; existing reserve forest field data includes field survey maps of the project area;

[0030] Step 2, data processing: Select road intersections and building corners as control points in drone aerial images and satellite maps, convert drone image coordinates into satellite map coordinates, align drone images by comparing control point position deviations, overlay the aligned drone aerial images onto the satellite map layer to obtain overlaid images, and divide sub-class boundaries and tree species distribution in the overlaid images based on existing reserve forest field data;

[0031] Step 3, data analysis: an object-oriented classification algorithm combined with multi-scale segmentation technology and feature weighting strategy is used to select multiple representative sample areas in the superimposed images. The multiple sample areas cover a variety of tree species and non-vegetation types, and the sample information is recorded; the confusion matrix method is used to verify the classification accuracy of the multiple sample areas to ensure that the overall vegetation and tree species in the multiple sample areas are not lower than the preset ratios, and the substandard areas are reclassified or manually corrected; specifically, the ratio of the overall vegetation is greater than or equal to 85% of the ratio of the vegetation coverage rate in the drone aerial images and satellite maps to the vegetation coverage rate of the actual small class, and the ratio of the tree species is the ratio of the number of tree species in the drone aerial images and satellite maps to the number of tree species in the actual small class, which is not less than 90%.

[0032] Use boundary analysis tools to extract sub-class boundaries in the superimposed images and compare and adjust them with field survey maps. When errors are large, use drone aerial images for manual corrections. Use field data to calculate tree growth indicators, use time series analysis to evaluate growth trends and influencing factors, and analyze health conditions through image spectrum and texture features. Process topographic map data in satellite maps to generate slope and aspect thematic maps, identify steep slope areas, use buffer analysis and proximity analysis to calculate the distance and spatial distribution characteristics between sub-classes and surrounding features, and provide suggestions for afforestation design.

[0033] Step 4. Data review: Use edge detection algorithm to automatically extract sub-class boundaries and compare with Aowei map. The deviation of sub-class boundaries shall not exceed ±5 meters, and the overlap degree shall be more than 95% to be qualified; use multi-spectral or hyperspectral characteristics of drone images combined with machine learning algorithms to review tree species with an accuracy of not less than 90%, and conduct on-site inspections of representative areas to obtain analysis results; visually compare analysis results with field survey maps and attribute information; calculate the error rate of sub-class boundaries, tree species and growth conditions, and use Kappa coefficient or Jaccard similarity index to evaluate consistency, and verify areas with large differences, blurred boundaries and important ecological nodes.

[0034] Furthermore, in step 1, the method for obtaining satellite map data of the project area is: obtaining the latest high-definition satellite map data of the project area from the Aowei map platform, the latest high-definition satellite map data includes high-resolution satellite images with a resolution of not less than 0.5m in TIF or GeoTIFF format and vector topographic maps in Shapefile format, the vector topographic maps including contour lines, water systems and roads.

[0035] Furthermore, in step 1, the method for obtaining UAV aerial photography data of the project area is: select a UAV of model DJIMini2SE, and let the UAV fly according to a route drawn in advance, the UAV is equipped with a high-resolution camera, the high-resolution camera sets the exposure mode to automatically or manually adjust to a suitable exposure value, and enables the distortion correction function, sets the forward overlap to not less than 80%, and the lateral overlap to not less than 70%; the high-resolution camera obtains UAV aerial photography data of the project area.

[0036] Specifically, in areas within the project area where the height difference does not exceed 50 meters, the UAV's flight altitude is between 50 and 150 meters; in areas within the project area where the slope exceeds 30°, the UAV's flight altitude is greater than 100 meters; the UAV adjusts its flight direction according to the solar azimuth and season to reduce shadows, identifies areas with tall trees and bushes, and sets a suitable flight altitude.

[0037] Specifically, before the drone takes off, the route is drawn on the Aowei map using DJI Pilot software based on the CAD coordinates to ensure at least 5% overlap of flights inside and outside the squad boundary, 80% heading overlap, 60% lateral overlap, and select at least three flat, open take-off and landing points without electromagnetic interference.

[0038] The drone is equipped with a GPS signal tracker to prevent loss of control, and is set to automatically return home when the battery power drops below 20%. A backup communication module is installed to enhance signal stability, and the automatic landing procedure is activated when the signal is lost. When the drone is performing aerial photography operations, it needs to undergo pre-flight inspection, flight operations, and data recording.

[0039] Pre-flight inspection includes checking that all parts of the drone are not damaged, such as propellers, motors, gimbals, etc.; the high-resolution camera is free of scratches and dust and has the latest firmware; the drone battery is fully charged and the voltage is stable. It is best to prepare two spare batteries; the remote control and display screen are fully charged and the connection is stable. Import Aowei map data into DJI Pilot to plan the route, set camera parameters and data storage path, select an open and interference-free take-off point and confirm weather conditions.

[0040] The flight operation includes taking off slowly after unlocking the drone, keeping a steady climb, and flying steadily to the preset altitude. Entering the route flight mode automatically shoots, monitors the drone status and image quality, and initiates return in case of abnormal conditions. After completing the flight, reduce the throttle to land, ensure a safe landing, and recover the drone.

[0041] Data records include flight logs such as take-off and landing time, duration, altitude, speed, and abnormal conditions and camera parameter adjustment records. Export data to a computer and use software such as Pix4Dmapper to check image integrity, clarity, and overlap, remove problematic images, and name and store them as required.

[0042] Furthermore, in step 1, the method for obtaining the field data of the existing reserve forests in the project area is as follows: first, the completed field survey data is comprehensively summarized, and the summarized data is converted from the original format of Excel and CSV to the Shapefile format using the Data Interoperability Tool of ArcGIS to ensure that the fields correspond; then, the integrity of the summarized data is checked using GIS software to check whether the small and medium classes in the summarized data of the project are complete and whether the attribute information is complete, and whether the boundaries of the small classes are overlapped or missing through spatial analysis, and the closure of the small class boundaries and the filling of the attribute fields are verified using the topology checking tool, and the problematic data is traced and corrected in a timely manner.

[0043] Furthermore, in step 3, the number of the multiple regions is greater than 150, and the size of each region is 50m×50m. The subclass boundaries are extracted using ArcGIS “Boundary Analysis Tool”; and the vector topographic map is processed using ArcGIS “KML” tool to generate slope and aspect thematic maps.

[0044] Furthermore, in step 4, ArcGIS and QGIS statistical analysis modules were used to calculate the error rates of subclass boundaries, tree species, and growth conditions.

[0045] In summary, the field data review method of the national reserve forest in the present invention achieves data accuracy at the centimeter level through high-precision drone image acquisition, advanced data analysis algorithms and strict data review processes. For example, in the small-class boundary review, the deviation is controlled within a very small range to ensure that the construction design is highly consistent with the actual situation, reduce rework and resource waste caused by data errors, improve resource utilization, and ensure afforestation quality; in terms of manpower, drones assist in reducing the demand for manpower for field surveys; in terms of material and machinery costs, the depreciation and maintenance costs of drone-related consumables and equipment are lower than those of traditional measurement equipment; although other costs include drone operator training, the overall cost is still significantly reduced, saving a lot of money for the project and solving the problems of low efficiency and poor accuracy in traditional manual review.

Claims

1. A method for reviewing the field data of national reserve forests, characterized in that: include: Step 1: Data preparation: Obtain satellite map data, drone aerial data and field data of existing reserve forests in the project area; Satellite map data includes satellite maps of the project area; drone aerial photography data includes drone aerial images of the project area; existing reserve forest field data includes field survey maps of the project area; Step 2, data processing: Select road intersections and building corners as control points in drone aerial images and satellite maps, convert drone image coordinates into satellite map coordinates, align drone images by comparing control point position deviations, overlay the aligned drone aerial images onto the satellite map layer to obtain overlaid images, and divide sub-class boundaries and tree species distribution in the overlaid images based on existing reserve forest field data; Step 3, data analysis: Use object-oriented classification algorithm combined with multi-scale segmentation technology and feature weighting strategy to select multiple representative sample areas in the superimposed images. The multiple sample areas cover a variety of tree species and non-vegetation types, and record sample information; use confusion matrix method to verify the classification accuracy of multiple sample areas, ensure that the overall vegetation and tree species in multiple sample areas are not lower than the preset ratio, and reclassify or manually correct the substandard areas; use boundary analysis tools to extract the sub-class boundaries in the superimposed images and compare and adjust them with the field survey map. When the error is large, manually correct it with drone aerial images, calculate forest growth indicators with field data, evaluate growth trends and influencing factors with time series analysis, and analyze health status through image spectrum and texture features; process topographic map data in satellite maps to generate slope and aspect thematic maps, identify steep slope areas, use buffer analysis and proximity analysis to calculate the distance and spatial distribution characteristics between sub-classes and surrounding features, and provide suggestions for afforestation design; Step 4: Data review: Use edge detection algorithm to automatically extract sub-class boundaries and compare with Aowei map. The sub-class boundary deviation does not exceed ±5 meters, and the overlap degree is more than 95% to be judged as qualified; Use the multispectral or hyperspectral characteristics of drone images combined with machine learning algorithms to review tree species with an accuracy of no less than 90%, and conduct on-site inspections of representative areas to obtain analysis results; Visually compare the analysis results with field survey maps and attribute information; Calculate the error rates of subclass boundaries, tree species, and growth conditions, use the Kappa coefficient or Jaccard similarity index to assess consistency, and verify areas with large differences, blurred boundaries, and important ecological nodes.

2. The method for reviewing the field data of national reserve forests according to claim 1, characterized in that: In step 1, the method for obtaining satellite map data of the project area is: obtain the latest high-definition satellite map data of the project area from the Aowei map platform. The latest high-definition satellite map data includes high-resolution satellite images with a resolution of not less than 0.5m in TIF or GeoTIFF format and vector topographic maps in Shapefile format. The vector topographic maps include contour lines, water systems and roads.

3. The method for reviewing the field data of national reserve forests according to claim 1, characterized in that: In step 1, the method for obtaining drone aerial photography data of the project area is as follows: select a drone of model DJIMini2SE, and let the drone fly according to a route drawn in advance. The drone is equipped with a high-resolution camera. The exposure mode of the high-resolution camera is set to automatic or manually adjusted to a suitable exposure value, and the distortion correction function is enabled. The forward overlap is set to not less than 80%, and the lateral overlap is not less than 70%; the high-resolution camera obtains drone aerial photography data of the project area.

4. The method for reviewing the field data of national reserve forests according to claim 1, characterized in that: In step 1, the method for obtaining the field data of the existing reserve forests in the project area is as follows: first, the completed field survey data is comprehensively summarized, and the summarized data is converted from the original format of Excel and CSV to the Shapefile format using ArcGIS's Data Interoperability Tool to ensure field correspondence; then, the integrity of the summarized data is checked using GIS software to check whether the small and medium classes in the project's summarized data are complete and whether the attribute information is complete, and whether the boundaries of small classes are overlapped or missing through spatial analysis, and the closure of small class boundaries and the filling of attribute fields are verified using topology checking tools, and problematic data are promptly traced and corrected.

5. The method for reviewing the field data of national reserve forests according to claim 3 is characterized in that: In step 1, in areas within the project area where the height difference does not exceed 50 meters, the drone's flight altitude is between 50 and 150 meters; in areas within the project area where the slope exceeds 30°, the drone's flight altitude is greater than 100 meters; the drone adjusts its flight direction according to the solar azimuth and season to reduce shadows, identifies tall trees and bushes, and sets a suitable flight altitude.

6. The method for reviewing the field data of national reserve forests according to claim 5 is characterized in that: In step 1, before the drone takes off, use DJI Pilot software to draw the route on the Aowei map based on the CAD coordinates to ensure at least 5% overlap inside and outside the squad boundary, 80% overlap in heading, and 60% overlap in lateral direction. Select at least three flat, open take-off and landing points without electromagnetic interference.

7. The method for reviewing the field data of national reserve forests according to claim 2 is characterized in that: In step 3, the number of the plurality of regions is greater than 150, and the size of each region is 50 m × 50 m; Use ArcGIS "Boundary Analysis Tool" to extract class boundaries; use ArcGIS "KML" tool to process vector topographic maps to generate slope and aspect thematic maps.

8. The method for reviewing the field data of national reserve forests according to claim 1 is characterized in that: In step 4, ArcGIS and QGIS statistical analysis modules were used to calculate the error rates of subclass boundaries, tree species, and growth conditions.

Images