Methods, apparatus and electronic equipment for estimating the location and height of individual trees

Abstract

This application discloses a method, apparatus, and electronic device for estimating the location and height of individual trees. The method includes: processing digital aerial photography data to obtain a canopy height model and normalized point cloud data; setting tree height parameters; segmenting the canopy height model using a marker-controlled watershed algorithm based on the tree height parameters to obtain a first segmentation result; setting horizontal spacing parameters between individual trees; segmenting the normalized point cloud data using a region growing method based on the horizontal spacing parameters to obtain a second segmentation result; merging the first and second segmentation results to obtain a set of segmentation results; and optimizing the different segmentation results in the set of segmentation results to obtain a target segmentation result. By analyzing and processing digital aerial photography data, the accuracy and reliability of estimating the location and height of individual trees are improved, and the method is applicable to estimating the location and height of individual trees in sparse forest stands of different regions and tree species.

Description

Technical Field

[0001] This application relates to the field of forestry remote sensing technology, and in particular to methods, apparatus and electronic equipment for estimating the location and height of individual trees. Background Technology

[0002] Forests, as the most important terrestrial vegetation resource, play a vital role in improving and maintaining regional ecological environments, and also significantly contribute to regulating global and regional climate change and bioenergy consumption. Among related technologies, traditional on-site measurement methods are often time-consuming, labor-intensive, and highly dangerous, and it is difficult to conduct continuous sampling surveys of large areas of forest cover, resulting in poor timeliness and applicability in reflecting dynamic changes in forest ecosystems. Satellite remote sensing technology is often limited by satellite revisit cycles, resolution, and the influence of cloud cover, rainfall, and other natural factors, making timely and rapid responses difficult. LiDAR technology is often too costly to widely implement in the short term.

[0003] Therefore, improving the accuracy and reliability of the estimation results of individual tree positions and tree heights has become an urgent problem to be solved. Summary of the Invention

[0004] This application aims to at least partially address one of the technical problems in the related art.

[0005] Therefore, the first objective of this application is to propose a method for estimating the location and height of a single tree, in order to solve the technical problem in the prior art that the location and height of a single tree cannot be estimated accurately and reliably.

[0006] To achieve the above objectives, a first aspect of this application provides a method for estimating the location and height of a single tree. The method includes: processing digital aerial photography data to obtain a canopy height model and normalized point cloud data; setting tree height parameters; segmenting the canopy height model using a marker-controlled watershed algorithm based on the tree height parameters to obtain a first segmentation result; setting horizontal spacing parameters between individual trees; segmenting the normalized point cloud data using a region growing method based on the horizontal spacing parameters to obtain a second segmentation result; merging the first segmentation result and the second segmentation result to obtain a set of segmentation results; and optimizing the different segmentation results in the set of segmentation results to obtain a target segmentation result.

[0007] In addition, the method for estimating the location and height of a single tree according to the above embodiments of this application may also have the following additional technical features:

[0008] According to one embodiment of this application, before processing the digital aerial photography data, the process includes: performing trajectory calculation and quality inspection on the digital aerial photography data.

[0009] According to one embodiment of this application, optimizing different segmentation results in the segmentation result set to obtain a target segmentation result includes: obtaining a first average spacing of dense stands and a second average spacing of sparse stands; and optimizing different segmentation results in the segmentation result set based on the first average spacing and the second average spacing to obtain a target segmentation result.

[0010] According to one embodiment of this application, optimizing different segmentation results in the segmentation result set based on the first average spacing and the second average spacing to obtain a target segmentation result includes: for the dense forest stand, in response to the horizontal spacing between the individual trees in the segmentation result set being less than the first average spacing, retaining the segmentation result with the largest tree height value in the segmentation result set; for the sparse forest stand, in response to the horizontal spacing between the individual trees in the segmentation result set being less than the second average spacing, retaining the segmentation result with the largest tree height value in the segmentation result set; until only one individual tree corresponds to the segmentation result at the same location or within a specified range in the dense forest stand or the sparse forest stand, then taking the current segmentation result as the target segmentation result.

[0011] According to one embodiment of this application, the method further includes: acquiring measured data of the location of a single tree and measured data of the height of a single tree; acquiring the single tree identification accuracy of the location of the single tree based on the target segmentation result and the measured data of the location of the single tree; and acquiring the estimation accuracy of the height of the single tree based on the target segmentation result and the measured data of the height of the single tree.

[0012] According to one embodiment of this application, obtaining the single-tree recognition accuracy of the single-tree position based on the target segmentation result and the measured data of the single-tree position includes: obtaining an evaluation index of the single-tree recognition accuracy based on the target segmentation result and the measured data of the single-tree position; and obtaining the single-tree recognition accuracy of the single-tree position based on the evaluation index.

[0013] According to one embodiment of this application, the step of obtaining the estimation accuracy of the single tree height based on the target segmentation result and the measured data of the single tree height includes: obtaining the coefficient of determination and root mean square error of the target segmentation result and the single tree height based on the target segmentation result and the measured data of the single tree height; and obtaining the estimation accuracy of the single tree height based on the coefficient of determination and the root mean square error.

[0014] To achieve the above objectives, a second aspect of this application provides an apparatus for estimating the location and height of a single tree. The apparatus includes: a first acquisition module for processing digital aerial photography data to acquire a canopy height model and normalized point cloud data; a second acquisition module for setting tree height parameters, segmenting the canopy height model using a marker-controlled watershed algorithm based on the tree height parameters to acquire a first segmentation result, and setting horizontal spacing parameters between individual trees, segmenting the normalized point cloud data using a region growing method based on the horizontal spacing parameters to acquire a second segmentation result; a merging module for merging the first segmentation result and the second segmentation result to acquire a set of segmentation results; and an optimization module for optimizing different segmentation results in the set of segmentation results to acquire a target segmentation result.

[0015] In addition, the device for estimating the position and height of a single tree according to the above embodiments of this application may also have the following additional technical features:

[0016] According to one embodiment of this application, the first acquisition module is further configured to: perform trajectory calculation and quality inspection on the digital aerial photography data.

[0017] According to one embodiment of this application, the optimization module is further configured to: obtain a first average spacing of dense stands and a second average spacing of sparse stands; and optimize different segmentation results in the segmentation result set based on the first average spacing and the second average spacing to obtain a target segmentation result.

[0018] According to one embodiment of this application, the optimization module is further configured to: for the dense forest stand, in response to the horizontal spacing between the individual trees in the segmentation result set being less than the first average spacing, retain the segmentation result with the largest tree height value in the segmentation result set; for the sparse forest stand, in response to the horizontal spacing between the individual trees in the segmentation result set being less than the second average spacing, retain the segmentation result with the largest tree height value in the segmentation result set; until only one individual tree corresponds to the segmentation result at the same location or within a specified range in the dense forest stand or the sparse forest stand, then take the current segmentation result as the target segmentation result.

[0019] According to one embodiment of this application, the apparatus is further configured to: acquire measured data of the location of a single tree and measured data of the height of a single tree; acquire the single tree identification accuracy of the location of the single tree based on the target segmentation result and the measured data of the location of the single tree; and acquire the estimation accuracy of the height of the single tree based on the target segmentation result and the measured data of the height of the single tree.

[0020] According to one embodiment of this application, the apparatus is further configured to: obtain an evaluation index of the single-tree recognition accuracy based on the target segmentation result and the measured data of the single-tree position; and obtain the single-tree recognition accuracy of the single-tree position based on the evaluation index.

[0021] According to one embodiment of this application, the apparatus is further configured to: obtain the coefficient of determination and root mean square error of the target segmentation result and the single tree height based on the target segmentation result and the measured data of the single tree height; and obtain the estimation accuracy of the single tree height based on the coefficient of determination and the root mean square error.

[0022] To achieve the above objectives, a third aspect of this application provides an electronic device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements the method for estimating the location and height of a single tree as described in any one of the first aspects of this application.

[0023] To achieve the above objectives, a fourth aspect of this application provides a non-transitory computer-readable storage medium storing computer instructions for causing a computer to execute a method for estimating the position and height of a single tree as described in any one of the first aspects of this application. Attached Figure Description

[0024] Figure 1 This is a schematic diagram of a method for estimating the location and height of a single tree according to an embodiment of this application.

[0025] Figure 2 This is a schematic diagram of the result of processing digital aerial photography data as disclosed in this application.

[0026] Figure 3 This is a schematic diagram illustrating the measurement of single-tree spacing and determination of seed points based on NPPC point cloud data in this application.

[0027] Figure 4 This is a schematic diagram of a method for estimating the location and height of a single tree, as disclosed in another embodiment of this application.

[0028] Figure 5 This is a schematic diagram of a method for estimating the location and height of a single tree, as disclosed in another embodiment of this application.

[0029] Figure 6 This is a schematic diagram of a method for estimating the location and height of a single tree, as disclosed in another embodiment of this application.

[0030] Figure 7 This is a schematic diagram of a method for estimating the location and height of a single tree, as disclosed in another embodiment of this application.

[0031] Figure 8 This is a schematic diagram of the structure of a device for estimating the position and height of a single tree as disclosed in one embodiment of this application.

[0032] Figure 9 This is a schematic diagram of the structure of an electronic device disclosed in an embodiment of this application. Detailed Implementation

[0033] To better understand the above technical solutions, exemplary embodiments of this disclosure will be described in more detail below with reference to the accompanying drawings. While exemplary embodiments of this disclosure are shown in the drawings, it should be understood that this disclosure can be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of this disclosure to those skilled in the art.

[0034] The following description, with reference to the accompanying drawings, illustrates a method and apparatus for estimating the location and height of a single tree according to an embodiment of this application.

[0035] Figure 1 This is a flowchart illustrating a method for estimating the location and height of a single tree according to an embodiment of this application.

[0036] like Figure 1 As shown in the embodiments of this application, the method for estimating the location and height of a single tree specifically includes the following steps:

[0037] S101. Process the digital aerial photography data to obtain the canopy height model and normalized point cloud data.

[0038] It should be noted that, before processing digital aerial photography data, in order to ensure the quality of the aerial photography data, track calculation can be performed on the digital aerial photography data, that is, the calculation of six parameters: latitude and longitude coordinates, relative flight altitude, roll angle, pitch angle, and yaw angle. In addition, quality inspection of digital aerial photography data can be performed, that is, quality inspection of digital aerial photography data such as forward overlap, lateral overlap, image sharpness, cloud coverage, shadow content, exposure, and total number of single images.

[0039] Optionally, digital surface model (DSM), digital elevation model (DEM), canopy height model (CHM), and normalized photography point cloud (NPPC) data can be obtained by using structure from motion (SFM) algorithm and regional network adjustment techniques, through feature point matching and point cloud densification.

[0040] It should be noted that the SfM algorithm is the process of estimating the three-dimensional structure of a scene from digital aerial photography data. In order to determine the relative position and attitude of each digital aerial photography data in a specified ground coordinate system, a regional network can be constructed based on the relative position and attitude after relative orientation of the images. The regional network results are then subjected to absolute orientation and self-calibrated bundle adjustment to obtain the image exterior orientation elements, object point three-dimensional coordinates, and adjustment files in the specified ground coordinate system. After feature point matching, dense point cloud data is generated through aerial triangulation.

[0041] It should be noted that for dense point cloud data, such as Figure 2 As shown in (a), a series of processes are performed. First, point cloud noise reduction is performed. Then, Kriging interpolation is applied to the noise-filtered DSM dense point cloud to obtain a DSM raster image with a defined spatial resolution, as shown in (a). Figure 2 As shown in (b), ground points and non-ground points are then classified. Ground points are interpolated using a triangulated irregular network (TIN) to obtain a DEM with the same spatial resolution as the DSM, as shown below. Figure 2 As shown in (c), DEM focuses on expressing topography and landforms, while DSM adds the expression of elevation information of surface structures on the basis of DEM. DEM is used to normalize DSM raster image and dense point cloud to obtain the corresponding canopy height model CHM and normalized point cloud data NPPC. Normalization is to subtract the topographic height information from the point cloud with absolute height generated after aerial triangulation to obtain the relative height of the point cloud, thereby eliminating the influence of topography on canopy height, i.e. CHM = DSM – DEM.

[0042] S102. Set tree height parameters. Based on the tree height parameters, use the marker-controlled watershed algorithm to segment the canopy height model to obtain the first segmentation result. Set the horizontal spacing parameters between individual trees. Based on the horizontal spacing parameters, use the region growing method to segment the normalized point cloud data to obtain the second segmentation result.

[0043] It should be noted that the Canopy Height Model (CHM) is a rasterized representation of the relative height of the forest canopy. It is used to reflect the vertical height variation and horizontal distribution of the forest canopy and can provide a lot of information, such as the number of individual trees, their location, tree height, and canopy width.

[0044] The label-controlled watershed algorithm improves segmentation accuracy by adding internal and external labels.

[0045] It should be noted that when using the marker-controlled watershed algorithm for single-tree segmentation, after the water basin within the range of a single tree is determined, parameters such as the treetop, tree height, and position of the single tree can be calculated using the CHM grayscale value and position within the water basin. The maximum value of the CHM grid within the water basin is designated as the vertex of the detected single tree, the CHM grayscale value is the tree height of the single tree, and the X and Y values ​​of the treetop are the position of the single tree.

[0046] For example, a segmentation function can be calculated, and the image can be preprocessed using gradient magnitude images. A filtering algorithm is used to filter the CHM horizontally and vertically. The filtered image will show larger values ​​at boundaries and smaller values ​​where there are no boundaries. Foreground (internal) markers are calculated; these internal markers are binary images used to mark possible individual tree locations. Reconstruction is mainly achieved through morphological opening and closing operations. The opening operation involves erosion followed by dilation, which shrinks the image and expands it, filtering out protrusions smaller than the structuring element and separating slender overlaps. The closing operation involves dilation followed by erosion, filling gaps or pores smaller than the structuring element. Shortened intervals serve as connections. Then, background (external) markers are calculated. The external markers are binary images used to remove non-canopy areas. In the external marker image, pixels with a value of 1 represent canopy areas, and pixels with a value of 0 represent non-canopy areas. The segmentation function is then modified. Before the watershed transform, the segmented image needs to be inverted. Combining the internal and external markers, the segmentation function is improved. Then, the segmented image is reconstructed, and finally, the watershed transform is performed. The calculation is based on the segmentation function, and the foreground markers, background markers, and segmentation object boundaries are superimposed onto the initial image.

[0047] It should be noted that Normalized Point Cloud Data (NPPC) is a discrete point representation of the relative height of the forest canopy. To reduce the oversegmentation of individual trees in CHM, the region growing method is used to segment the normalized point cloud into individual trees.

[0048] It should be noted that the region growing method gathers pixels or regions with similar or identical properties to form the region to be segmented. The selected seed point is used as the starting point of the growth target. According to the pre-specified growth rules and attributes, the consistency of neighboring pixels is checked. If the neighboring pixels meet the consistency criteria, these neighboring pixels with similar attributes and growth rules to each seed point are merged into the region of the corresponding seed point. If they do not meet the criteria, they are discarded, thereby completing the segmentation.

[0049] Optionally, seed point extraction can be performed. The local maximum value method of dynamic window is used to detect the position of the top of a single tree as the starting seed point. The maximum dynamic detection window is set to twice the average crown width. The highest point in the study plot is extracted first using the dynamic window. This point is regarded as the top of the single tree and is used as the starting seed point for point cloud classification. The point cloud is then classified. Since there is a gap between the tree tops, the gap between the tree tops in sparser stands is larger and the gap between the tree tops in denser stands is smaller. Through actual measurement and inspection, the gap between the tree tops in sparser and denser stands is given respectively, as shown in Figure (3). Point A is the starting seed point in the area. By setting the horizontal gap threshold, starting from the tree top seed point, the relative distance between the tree top and nearby points is determined to classify the point cloud and determine the growth of target tree A. Due to the influence of crown width, the gap between trees decreases from top to bottom, especially in denser stands. In particular, for those overlapping trees, point cloud classification is more difficult. Therefore, the point clouds are sorted from high to low relative height, and detection is performed starting from the highest point. Point clouds with a spacing greater than a given distance threshold are excluded from the target tree of the seed point, while point clouds with a spacing less than the given distance threshold are classified into the target tree of the corresponding seed point.

[0050] Furthermore, based on the tree height parameter, the canopy height model can be segmented using the marker-controlled watershed algorithm to obtain the first segmentation result, and based on the horizontal spacing parameter, the normalized point cloud data can be segmented using the region growing method to obtain the second segmentation result.

[0051] It should be noted that this application does not impose any restrictions on the setting of tree height parameters and horizontal spacing parameters, which can be set according to the actual situation.

[0052] Optionally, there is no limit to the number of tree height and horizontal spacing parameters that can be set.

[0053] For example, you can set three tree height parameters (h-1), h, and (h+1) and three horizontal spacing parameters (d-0.5)m, dm, and (d+0.5).

[0054] S103. Merge the first segmentation result with the second segmentation result to obtain a set of segmentation results.

[0055] In this embodiment of the application, after obtaining the first segmentation result and the second segmentation result, the first segmentation result and the second segmentation result can be merged to obtain a segmentation result set.

[0056] S104. Optimize the different segmentation results in the segmentation result set to obtain the target segmentation result.

[0057] It should be noted that since there are different segmentation results in the segmentation result set, there may be cases where the same position or the vicinity is segmented multiple times. Therefore, it is necessary to optimize the different segmentation results in the segmentation result set in order to obtain the target segmentation result.

[0058] Optionally, optimization conditions can be set, and different segmentation results in the segmentation result set can be optimized according to the optimization conditions to obtain the target segmentation result.

[0059] It should be noted that after obtaining the target segmentation results, the estimated results of the individual tree position and individual tree height can be obtained from the target segmentation results.

[0060] The method for estimating the location and height of individual trees provided in this application processes digital aerial photography data to obtain a canopy height model and normalized point cloud data. Tree height parameters are set, and based on these parameters, the canopy height model is segmented using a marker-controlled watershed algorithm to obtain a first segmentation result. Horizontal spacing parameters between individual trees are then set, and based on these parameters, the normalized point cloud data is segmented using a region growing method to obtain a second segmentation result. The first and second segmentation results are merged to obtain a set of segmentation results. The different segmentation results in this set are then optimized to obtain the target segmentation result. Therefore, this application improves the accuracy and reliability of estimating the location and height of individual trees by analyzing and processing digital aerial photography data. Furthermore, this method is applicable to estimating the location and height of individual trees in sparse forest stands of different regions and tree species.

[0061] As one possible implementation, such as Figure 4 As shown, based on the above embodiments, the specific process of optimizing different segmentation results in the segmentation result set to obtain the target segmentation result in step S103 includes the following steps:

[0062] S401. Obtain the first average spacing of dense stands and the second average spacing of sparse stands.

[0063] It should be noted that since the spacing between dense and sparse stands may be different, the first average spacing of dense stands and the second average spacing of sparse stands can be obtained.

[0064] S402. Optimize the different segmentation results in the segmentation result set according to the first average spacing and the second average spacing to obtain the target segmentation result.

[0065] Optionally, for dense stands, in response to the horizontal spacing between individual trees in the segmentation result set being less than the first average spacing, the segmentation result with the largest tree height value in the segmentation result set is retained; for sparse stands, in response to the horizontal spacing between individual trees in the segmentation result set being less than the second average spacing, the segmentation result with the largest tree height value in the segmentation result set is retained, until only one individual tree corresponds to the segmentation result at the same location or within a specified range in either dense or sparse stands, then the current segmentation result is taken as the target segmentation result.

[0066] The method for estimating the location and height of individual trees provided in this application obtains the first average spacing of dense stands and the second average spacing of sparse stands, and optimizes different segmentation results in the segmentation result set based on the first and second average spacings to obtain the target segmentation result. Therefore, this application improves the accuracy and reliability of estimating the location and height of individual trees by optimizing different segmentation results in the segmentation result set.

[0067] Furthermore, after obtaining the target segmentation results, the single-tree recognition accuracy of the single-tree position can be obtained based on the target segmentation results and the measured data of the single-tree position, and the estimation accuracy of the single-tree height can be obtained based on the target segmentation results and the measured data of the tree height.

[0068] As one possible implementation, such as Figure 5 As shown, based on the above embodiments, the specific process of obtaining the single-tree identification accuracy of the single-tree position according to the target segmentation results and the measured data of the single-tree position, and obtaining the estimation accuracy of the single-tree height according to the target segmentation results and the measured data of the tree height, includes the following steps:

[0069] S501. Obtain the measured data of the location of a single tree and the measured data of the height of a single tree.

[0070] Optionally, a measurement area can be set, and the location and height of individual trees within the measurement area can be measured to obtain measured data of individual tree location and height.

[0071] S502. Based on the target segmentation results and the measured data of the individual tree positions, obtain the individual tree recognition accuracy of the individual tree position, and based on the target segmentation results and the measured data of the individual tree height, obtain the estimation accuracy of the individual tree height.

[0072] As one possible implementation, such as Figure 6As shown, based on the above embodiments, the specific process of obtaining the single-tree recognition accuracy at the single-tree location according to the target segmentation results and the measured data of the single-tree location includes the following steps:

[0073] S601. Based on the target segmentation results and the measured data of the single tree position, obtain the evaluation index of the single tree recognition accuracy.

[0074] It should be noted that if a tree is correctly segmented, it is called a correct segmentation, denoted by True Positive (TP); if a tree is actually a single tree but is not detected, it is called undersegmentation or missed detection, denoted by False Negative (FN); if a tree does not exist at all but is detected as a single tree, it is called oversegmentation or overdetection, denoted by False Positive (FP).

[0075] Optionally, evaluation indicators such as completeness r, precision p, and segmentation quality q can be obtained based on the target segmentation results and the measured data of individual tree positions.

[0076] The completeness rate can be calculated using the formula. Accuracy Segmentation quality

[0077] S602. Based on the evaluation indicators, obtain the single-tree identification accuracy of the single-tree location.

[0078] In this embodiment of the application, after obtaining the evaluation index, the single-tree identification accuracy of the single-tree position can be obtained based on the evaluation index.

[0079] As one possible implementation, such as Figure 7 As shown, based on the above embodiments, the specific process of obtaining the estimation accuracy of a single tree height according to the target segmentation results and the measured data of tree height includes the following steps:

[0080] S701. Based on the target segmentation results and the measured data of individual tree height, obtain the coefficient of determination and root mean square error of the target segmentation results and individual tree height.

[0081] Alternatively, the coefficient of determination and root mean square error of the target segmentation result and the height of a single tree can be obtained using the following formula:

[0082]

[0083] Where n is the number of individual trees, y i It is the measured value of the individual tree height of the i-th tree; It is y i The average value of y i ' is the estimated value of the individual tree height of the i-th tree.

[0084] S702. Based on the coefficient of determination and root mean square error, obtain the estimation accuracy of the height of a single tree.

[0085] In this embodiment of the application, after obtaining the coefficient of determination and the root mean square error, the estimation accuracy of the height of a single tree can be obtained based on the coefficient of determination and the root mean square error.

[0086] The method for estimating the location and height of a single tree provided in this application obtains measured data of the location and height of the single tree, and based on the target segmentation results and the measured data of the location, obtains the single tree identification accuracy of the location, and based on the target segmentation results and the measured data of the height, obtains the estimation accuracy of the single tree height. Therefore, this application improves the accuracy and reliability of estimating the location and height of a single tree by obtaining the single tree identification accuracy of the location and the estimation accuracy of the single tree height using measured data and target segmentation results.

[0087] Figure 8 This is a schematic diagram of the structure of a device for estimating the position and height of a single tree according to an embodiment of this application.

[0088] like Figure 8 As shown, the device 100 for estimating the location and height of a single tree includes: a first acquisition module 11, a second acquisition module 12, a merging module 13, and an optimization module 14. Among them,

[0089] The first acquisition module 11 is used to process digital aerial photography data to obtain canopy height model and normalized point cloud data;

[0090] The second acquisition module 12 is used to set tree height parameters, and according to the tree height parameters, to segment the canopy height model using the marker-controlled watershed algorithm to obtain a first segmentation result. It also sets horizontal spacing parameters between individual trees, and according to the horizontal spacing parameters, to segment the normalized point cloud data using the region growing method to obtain a second segmentation result.

[0091] The merging module 13 is used to merge the first segmentation result and the second segmentation result to obtain a set of segmentation results;

[0092] The optimization module 14 is used to optimize different segmentation results in the segmentation result set to obtain the target segmentation result.

[0093] According to one embodiment of this application, the first acquisition module 11 is further configured to: perform trajectory calculation and quality inspection on the digital aerial photography data.

[0094] According to one embodiment of this application, the optimization module 14 is further configured to: obtain a first average spacing of dense stands and a second average spacing of sparse stands; and optimize different segmentation results in the segmentation result set based on the first average spacing and the second average spacing to obtain a target segmentation result.

[0095] According to one embodiment of this application, the optimization module 14 is further configured to: for the dense forest stand, in response to the horizontal spacing between the individual trees in the segmentation result set being less than the first average spacing, retain the segmentation result with the largest tree height value in the segmentation result set; for the sparse forest stand, in response to the horizontal spacing between the individual trees in the segmentation result set being less than the second average spacing, retain the segmentation result with the largest tree height value in the segmentation result set; until only one individual tree corresponds to the segmentation result at the same location or within a specified range in the dense forest stand or the sparse forest stand, then take the current segmentation result as the target segmentation result.

[0096] According to one embodiment of this application, the apparatus 100 is further configured to: acquire measured data of the location of a single tree and measured data of the height of a single tree; acquire the single tree identification accuracy of the location of the single tree based on the target segmentation result and the measured data of the location of the single tree; and acquire the estimation accuracy of the height of the single tree based on the target segmentation result and the measured data of the height of the single tree.

[0097] According to one embodiment of this application, the apparatus 100 is further configured to: obtain an evaluation index of the single-tree recognition accuracy based on the target segmentation result and the measured data of the single-tree position; and obtain the single-tree recognition accuracy of the single-tree position based on the evaluation index.

[0098] According to one embodiment of this application, the apparatus 100 is further configured to: obtain the coefficient of determination and root mean square error of the target segmentation result and the single tree height based on the target segmentation result and the measured data of the single tree height; and obtain the estimation accuracy of the single tree height based on the coefficient of determination and the root mean square error.

[0099] This application provides a device for estimating the location and height of individual trees. It processes digital aerial photography data to obtain a canopy height model and normalized point cloud data. Tree height parameters are set, and the canopy height model is segmented using a marker-controlled watershed algorithm based on these parameters to obtain a first segmentation result. Horizontal spacing parameters between individual trees are then set, and the normalized point cloud data is segmented using a region growing method based on these parameters to obtain a second segmentation result. The first and second segmentation results are merged to obtain a set of segmentation results. Different segmentation results within this set are then optimized to obtain a target segmentation result. Therefore, this application improves the accuracy and reliability of estimating the location and height of individual trees by analyzing and processing digital aerial photography data. Furthermore, this method is applicable to estimating the location and height of individual trees in sparse forest stands of different regions and tree species.

[0100] To implement the above embodiments, this application also proposes an electronic device 2000, such as... Figure 9 As shown, it includes a memory 210, a processor 220, and a computer program stored in the memory 210 and capable of running on the processor 220. When the processor executes the program, it implements the aforementioned method for estimating the position and height of a single tree.

[0101] To implement the above embodiments, this application also proposes a non-transitory computer-readable storage medium storing computer instructions for causing a computer to execute the aforementioned method for estimating the position and height of a single tree.

[0102] In the description of this application, it should be understood that the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", "clockwise", "counterclockwise", "axial", "radial", "circumferential", etc., indicate the orientation or positional relationship shown in the accompanying drawings, and are only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this application.

[0103] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this application, "multiple" means two or more, unless otherwise explicitly specified.

[0104] In this application, unless otherwise expressly specified and limited, the terms "installation," "connection," "linking," and "fixing," etc., should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. Those skilled in the art can understand the specific meaning of the above terms in this application according to the specific circumstances.

[0105] In this application, unless otherwise expressly specified and limited, "above" or "below" the second feature can mean that the first feature is in direct contact with the second feature, or that the first feature is in indirect contact with the second feature through an intermediate medium. Furthermore, "above," "on top of," and "over" the second feature can mean that the first feature is directly above or diagonally above the second feature, or simply that the first feature is at a higher horizontal level than the second feature. "Below," "below," and "under" the second feature can mean that the first feature is directly below or diagonally below the second feature, or simply that the first feature is at a lower horizontal level than the second feature.

[0106] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0107] Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of this application.

Claims

1. A method for estimating the location and height of a single tree, characterized in that, include: Digital aerial photography data is processed to obtain canopy height models and normalized point cloud data; Set tree height parameters, and use the label-controlled watershed algorithm to segment the canopy height model according to the tree height parameters to obtain a first segmentation result. Set horizontal spacing parameters between individual trees, and use the region growing method to segment the normalized point cloud data according to the horizontal spacing parameters to obtain a second segmentation result. The first segmentation result is merged with the second segmentation result to obtain a set of segmentation results; The different segmentation results in the segmentation result set are optimized to obtain the target segmentation result; The step of optimizing different segmentation results in the segmentation result set to obtain the target segmentation result includes: Obtain the first average spacing of dense stands and the second average spacing of sparse stands; Based on the first average spacing and the second average spacing, the different segmentation results in the segmentation result set are optimized to obtain the target segmentation result; The step of optimizing different segmentation results in the segmentation result set based on the first average spacing and the second average spacing to obtain the target segmentation result includes: For the dense forest stand, in response to the horizontal spacing between the individual trees in the segmentation result set being less than the first average spacing, the segmentation result with the largest tree height value in the segmentation result set is retained; For the sparse forest stand, in response to the horizontal spacing between the individual trees in the segmentation result set being less than the second average spacing, the segmentation result with the largest tree height value in the segmentation result set is retained; When there is only one tree corresponding to the segmentation result at the same location or within a specified range in the dense forest stand or the sparse forest stand, the current segmentation result is taken as the target segmentation result.

2. The method according to claim 1, characterized in that, Before processing the digital aerial photography data, the following steps are included: The digital aerial photography data is subjected to trajectory calculation and quality inspection.

3. The method according to claim 1, characterized in that, The method further includes: Obtain measured data on the location and height of individual trees; Based on the target segmentation results and the measured data of the individual tree positions, the individual tree identification accuracy of the individual tree positions is obtained, and based on the target segmentation results and the measured data of the individual tree height, the estimation accuracy of the individual tree height is obtained.

4. The method according to claim 3, characterized in that, The step of obtaining the single-tree recognition accuracy at the single-tree location based on the target segmentation result and the measured data of the single-tree location includes: Based on the target segmentation results and the measured data of the single tree position, an evaluation index for the single tree recognition accuracy is obtained; Based on the evaluation indicators, the single-tree identification accuracy at the single-tree location is obtained.

5. The method according to claim 3, characterized in that, The step of obtaining the estimation accuracy of the single tree height based on the target segmentation result and the measured data of the single tree height includes: Based on the target segmentation results and the measured data of the individual tree height, the determination coefficient and root mean square error of the target segmentation results and the individual tree height are obtained. The estimation accuracy of the individual tree height is obtained based on the coefficient of determination and the root mean square error.

6. A device for estimating the position and height of a single tree, characterized in that, The method described by any one of claims 1-5 includes: The first acquisition module is used to process digital aerial photography data to obtain canopy height models and normalized point cloud data; The second acquisition module is used to set tree height parameters, segment the canopy height model using the marker-controlled watershed algorithm based on the tree height parameters to obtain a first segmentation result, and set horizontal spacing parameters between individual trees, segmenting the normalized point cloud data using the region growing method based on the horizontal spacing parameters to obtain a second segmentation result. The merging module is used to merge the first segmentation result and the second segmentation result to obtain a set of segmentation results; The optimization module is used to optimize different segmentation results in the segmentation result set to obtain the target segmentation result.

7. An electronic device, characterized in that, Including memory and processor; The processor reads executable program code stored in the memory to run a program corresponding to the executable program code, so as to implement the method as described in any one of claims 1-5.

Images