Forest wood volume estimation method and device, terminal and readable storage medium

Abstract

The invention is suitable for the field of forest resource evaluation, and provides a wood volume estimation method and device, a terminal and a readable storage medium, and the method comprises the steps: obtaining the external form data and specific wave band spectrum data of each tree in a target region; the specific wave band spectral data corresponds to a specific wave band, and the specific wave band is associated with tree health information; determining health description information of each tree based on the spectral data of the specific wave band; based on the external form data and the health description information, determining a single wood volume of each tree; and based on the single wood volume of the plurality of trees in the target area, calculating the overall wood volume of the target area. According to the scheme, the overall wood volume of the target area is calculated through the external form data and the health description information of the trees in the target area, and the calculation accuracy of the wood volume is improved.

Description

Technical Field

[0001] The present application belongs to the field of forest resource assessment, and in particular relates to a forest timber volume estimation method, device, terminal and readable storage medium. Background Art

[0002] Timber volume is an important concept in forest resource management. It represents the total volume of all trees in a forest area and is an important basis for assessing the quantity and quality of forest resources, formulating forest management plans, and planning timber production and utilization. The prediction of timber volume plays a very important role in forest resource surveys, national strategic timber reserves, timber transactions, timber asset assessments, carbon sink assessments, financial insurance value assessments, and loan evaluations. In recent years, the methods for assessing timber volume have become increasingly diverse, but most methods are still limited to estimating using external morphological data such as plant height, breast diameter, and crown width, and the timber volume obtained in this way is less accurate. Summary of the invention

[0003] The embodiments of the present application provide a forest timber volume estimation method, device, terminal and readable storage medium to solve the problem of poor calculation accuracy of forest timber volume in the prior art.

[0004] A first aspect of an embodiment of the present application provides a method for estimating forest timber volume, comprising:

[0005] Acquire external morphological data and specific band spectral data of each tree in the target area; the specific band spectral data corresponds to a specific band, and the specific band is associated with tree health information;

[0006] Determining health description information of each of the trees based on the specific band spectral data;

[0007] Determining the individual wood volume of each of the trees based on the external morphological data and the health description information;

[0008] The overall forest volume of the target area is calculated based on the individual wood volumes of the plurality of trees in the target area.

[0009] A second aspect of an embodiment of the present application provides a forest timber volume estimation device, comprising:

[0010] An acquisition module, used to acquire external morphological data and specific band spectral data of each tree in the target area; the specific band spectral data corresponds to a specific band, and the specific band is associated with tree health information;

[0011] A first determination module, configured to determine health description information of each of the trees based on the spectral data of the specific band;

[0012] A second determination module, configured to determine the single wood volume of each of the trees based on the external morphological data and the health description information;

[0013] A calculation module is used to calculate the overall forest volume of the target area based on the single wood volume of multiple trees in the target area.

[0014] A third aspect of an embodiment of the present application provides a terminal, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the method described in the first aspect when executing the computer program.

[0015] A fourth aspect of an embodiment of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method described in the first aspect are implemented.

[0016] A fifth aspect of the present application provides a computer program product, which, when executed on a terminal, enables the terminal to execute the steps of the method described in the first aspect.

[0017] As can be seen from the above, the present application not only obtains the external morphological data of each tree in the target area, but also obtains the specific band spectral data of each tree. By introducing the specific band spectral data, the errors and limitations caused by single data are reduced, and the accuracy and reliability of subsequent timber volume calculations are improved. Among them, the specific band spectral data corresponds to a specific band, and the specific band is associated with the tree health information. In this way, based on the specific band spectral data, the health description information of each tree is determined to achieve an accurate assessment of the health status of the trees. Then, based on the external morphological data and the health description information, the single wood volume of each tree is determined. This process takes into account the health status of the trees, so that the accuracy of the timber volume of each tree is improved. On the premise that the accuracy of the single wood volume is greatly improved, the accuracy of the overall forest volume of the target area calculated based on the single wood volume of multiple trees in the target area is also improved accordingly. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0019] Figure 1 is a flow chart of a forest timber volume estimation method provided in an embodiment of the present application;

[0020] Figure 2is a training flow chart of a tree health classification model provided in an embodiment of the present application;

[0021] Figure 3 is a structural diagram of a forest timber volume estimation device provided in an embodiment of the present application;

[0022] Figure 4 It is a structural diagram of a terminal provided in an embodiment of the present application. DETAILED DESCRIPTION

[0023] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present application. However, it should be clear to those skilled in the art that the present application may also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present application.

[0024] It should be understood that when used in this specification and the appended claims, the term "comprising" indicates the presence of described features, integers, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or groups thereof.

[0025] It should also be understood that the terms used in this application specification are only for the purpose of describing specific embodiments and are not intended to limit the application. As used in this application specification and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include plural forms.

[0026] It should be further understood that the term “and / or” used in the specification and appended claims refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0027] As used in this specification and the appended claims, the term "if" may be interpreted as "when" or "upon" or "in response to determining" or "in response to detecting," depending on the context. Similarly, the phrases "if it is determined" or "if [described condition or event] is detected" may be interpreted as meaning "upon determination" or "in response to determining" or "upon detection of [described condition or event]" or "in response to detecting [described condition or event]," depending on the context.

[0028] In a specific implementation, the terminal described in the embodiments of the present application includes, but is not limited to, other portable devices such as mobile phones, laptop computers, or tablet computers with touch-sensitive surfaces (e.g., touch screen displays and / or touch pads). It should also be understood that in some embodiments, the device is not a portable communication device, but a desktop computer with a touch-sensitive surface (e.g., touch screen displays and / or touch pads).

[0029] In the following discussion, a terminal including a display and a touch-sensitive surface is described. However, it should be understood that the terminal may include one or more other physical user interface devices such as a physical keyboard, mouse and / or joystick.

[0030] The terminal supports various applications, such as one or more of the following: a drawing application, a presentation application, a word processing application, a website creation application, a disk burning application, a spreadsheet application, a game application, a telephone application, a video conferencing application, an email application, an instant messaging application, a workout support application, a photo management application, a digital camera application, a digital camcorder application, a web browsing application, a digital music player application, and / or a digital video player application.

[0031] Various applications that can be executed on the terminal can use at least one common physical user interface device such as a touch-sensitive surface. One or more functions of the touch-sensitive surface and corresponding information displayed on the terminal can be adjusted and / or changed between applications and / or within corresponding applications. In this way, the common physical architecture of the terminal (e.g., the touch-sensitive surface) can support various applications with user interfaces that are intuitive and transparent to the user.

[0032] It should be understood that the size of the serial numbers of the steps in this embodiment does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present application.

[0033] In order to illustrate the technical solution described in this application, a specific embodiment is provided below for illustration.

[0034] See also Figure 1 , Figure 1 This is a flow chart of a forest timber volume estimation method provided in an embodiment of the present application.

[0035] like Figure 1 As shown, a method for estimating forest timber volume comprises the following steps:

[0036] Step 101, obtaining external morphological data and specific band spectral data of each tree in the target area; the specific band spectral data corresponds to a specific band, and the specific band is associated with tree health information.

[0037] The target area is the forest area where the timber volume is to be estimated. Trees include but are not limited to eucalyptus, pine, poplar, fir, oak, birch and maple. The external morphological data include plant height, crown width and crown volume. The specific band is a specific wavelength range that is closely related to the health of trees. The specific band spectral data refers to the spectral images corresponding to these specific wavelength ranges, or the quantitative information such as frequency distribution and reflectivity calculated on these bands. The specific band spectral data is preprocessed data.

[0038] During their growth, trees are often attacked by pests and diseases, especially boring pests. Their impact on forest timber volume cannot be ignored. This step introduces specific band spectral data related to the health status of trees to determine the health information of trees, which helps to improve the accuracy of timber volume calculation.

[0039] In some embodiments, the obtaining of external morphological data and specific band spectral data of each tree in the target area includes: constructing a target feature map of the target area having tree appearance information and tree spectral information; performing feature segmentation on the target feature map according to the trees to obtain single tree feature maps corresponding to the individual trees, each of the single tree feature maps containing single tree appearance information and single tree spectral information of the individual trees; determining the plant height, crown width and crown volume of the tree as the external morphological data based on the single tree appearance information in the single tree feature map; performing spectral data extraction on the single tree spectral information of the single tree feature map according to the specific band to obtain the specific band spectral data.

[0040] Before constructing the target feature map of the target area, multi-dimensional information required for the composition is collected based on ground equipment and / or UAVs.

[0041] In some embodiments, the multi-dimensional information is a visible light image and a hyperspectral image of the target area, or the multi-dimensional information is a visible light image, a hyperspectral image and lidar data of the target area.

[0042] Visible light cameras are used to collect visible light images carrying location information, hyperspectral cameras are used to collect hyperspectral images carrying location information, and lidar equipment is used to collect lidar data carrying location information.

[0043] In some embodiments, the target area generally occupies a large area. In order to improve the quality of spatial information, multi-dimensional information collection of the target area is usually achieved through multiple collections of small areas, and multiple visible light images, hyperspectral images and lidar data are collected.

[0044] In some embodiments, visible light images carrying the geographic location and hyperspectral images carrying the geographic location of the target area are collected with the help of an integrated device equipped with a visible light camera and a hyperspectral camera, or, visible light images carrying the geographic location, hyperspectral images carrying the geographic location and lidar data carrying the geographic location of the target area are collected with the help of an integrated device equipped with a visible light camera, a hyperspectral camera and a lidar device, so as to realize synchronous collection of information, reduce the number of repeated monitoring and information collection, improve monitoring efficiency, reduce data errors, and improve subsequent mapping efficiency.

[0045] Preprocess the multidimensional information to remove noise from the information. According to the location information carried by each image, determine the repeated area between it and the adjacent images, crop the image, and remove the repeated area in the image. After that, according to the location information, realize image stitching to obtain the complete visible light image and complete hyperspectral image of the target area respectively. According to the preprocessed LiDAR data, calculate the elevation values ​​of each location in the target area and construct the elevation image of the target area.

[0046] In the case where the multidimensional data is a visible light image and a hyperspectral image, image position registration is achieved based on the position information carried by the complete image to ensure that the visible light image and the hyperspectral image can correspond to each other in space. The tree appearance information corresponding to the complete visible light image at the same position and the tree spectral information corresponding to the complete hyperspectral image are fused to obtain a target feature map with tree appearance information and tree spectral information of the target area. In this case, the target feature map is a two-dimensional image. Among them, the fusion can be achieved using methods such as pixel-based fusion, feature-based fusion, or decision-based fusion.

[0047] In the case where the multidimensional data is visible light images, hyperspectral images and lidar data, a two-dimensional image with tree appearance information and tree spectral information of the target area is constructed based on the complete visible light image and the complete hyperspectral image. This process is not repeated here. Three-dimensional reconstruction is performed based on the two-dimensional image and the elevation image to obtain a three-dimensional point cloud of the target area. At this time, the target feature map is a three-dimensional point cloud. Among them, the introduction of lidar technology helps to obtain information about the obscured parts, so as to build a more accurate target feature map based on visible light images, hyperspectral images and lidar data.

[0048] By building a map, the appearance of the trees can be matched to the spectrum, ensuring that no trees are missed. At the same time, the tree characteristic information of each tree is not missed, thereby improving the accuracy of forest volume estimation.

[0049] After obtaining the target feature map, a single tree segmentation method is performed by using binarization processing, deep learning method, threshold segmentation method or segmentation method based on point cloud data, etc., to obtain a single tree feature map of each tree having the appearance information and spectral information of the tree. Accordingly, a two-dimensional image or a three-dimensional point cloud image of the single tree feature map.

[0050] The single-tree segmentation operation enables each tree to have its own tree portrait, and can accurately and quickly obtain the tree's external morphological data and specific band spectral data.

[0051] In some embodiments, appearance data is extracted from the appearance information of individual trees in the individual tree feature map, and plant height, crown width and crown volume are directly calculated.

[0052] When the single tree feature map is a two-dimensional image, image processing algorithms such as edge detection and contour extraction are used to identify the bottom and top of the tree, and calculate the distance between them to obtain the tree's plant height. Image processing algorithms are also used to identify the crown boundary, and calculate its projected crown width to obtain the tree's crown width. Two-dimensional images cannot directly provide three-dimensional volume information. In this case, the crown volume is estimated based on the plant height and crown width. The crown volume estimation method will not be described in detail in this application.

[0053] When the single tree feature map is a three-dimensional point cloud image, the image processing algorithm is used to identify the bottom and top of the tree and calculate the distance between them to obtain the tree height. In the three-dimensional point cloud image, the boundary points of the crown can be clearly identified, and the crown width can be obtained by calculating the convex hull or the diameter of the minimum enclosing circle formed by these boundary points. The three-dimensional point cloud image provides the three-dimensional shape information of the tree, so the crown area can be directly obtained by calculating the volume occupied by the point cloud inside the crown.

[0054] In some embodiments, the plant height, crown width and crown area are the corrected plant height, corrected crown width and corrected crown area, that is, the plant height, crown width and crown area calculated above are corrected based on the plant height correction coefficient, crown width correction coefficient and crown area correction coefficient to obtain the corrected plant height, corrected crown width and corrected crown area.

[0055] Among them, the plant height correction coefficient, crown width correction coefficient and crown volume correction coefficient are all calculated based on the image measurement data and the actual measurement data. i , 1≤i≤n, i and n are positive integers, and the measured plant height is h ti ,

[0056] These correction coefficients are intended to reduce systematic errors and improve the data accuracy of external morphological data, and thus significantly improve the accuracy and reliability of forest timber volume estimation results when using these corrected external morphological data to determine forest timber volume. At the same time, the specific band is used as an extraction basis to extract specific band spectral data corresponding to the specific band from the single tree spectral information of the single tree characteristic map, which is used to determine the health description information of the tree.

[0057] At the same time, the specific band is used as an extraction basis to extract specific band spectral data corresponding to the specific band from the single tree spectral information of the single tree image, so as to determine the health description information of the tree.

[0058] Step 102: determining health description information of each of the trees based on the spectral data of the specific wavelength band.

[0059] The health description information may be comprehensive tree health status assessment information expressed in the form of health grade or health value.

[0060] The specific band spectral data is an optical mapping of the tree's health status. Through the specific band spectral data of the tree, the health description information used to evaluate the internal and external health status of the tree can be determined.

[0061] Through spectral analysis technology, the health status of trees can be accurately assessed. Non-contact monitoring of the spectrum reduces interference and damage to trees, while also improving the efficiency and safety of monitoring. In addition, the health monitoring and assessment method based on spectral data realizes automated and intelligent monitoring and assessment, greatly improving management efficiency.

[0062] In addition, the health description information can be displayed to forestry workers, who can use the information to discover health problems with the trees and take corresponding preventive and treatment measures to prevent the problems from further deteriorating. They can also use the information to determine whether to harvest the trees at the current time. In other words, in addition to determining the tree volume, the health description information can also realize health warnings and management decisions.

[0063] In some embodiments, the health description information is a health level, the specific band spectral data is a plurality of specific band spectral images, and the health description information of each of the trees is determined based on the specific band spectral data, including: obtaining a correlation between each of the specific bands and the health level; determining a stacking order of a plurality of the specific band spectral images based on the correlation; stacking a plurality of the specific band spectral images based on the stacking order to obtain a multi-channel image, wherein the multi-channel image contains an arrangement and combination relationship between spectral images of different bands in the stacking order, and different arrangement and combination relationships reflect different health levels; inputting the multi-channel image into a tree health grading model to obtain the health level of the tree.

[0064] The correlation degree is the correlation degree between the band and the final evaluation result, i.e., the health description information, when evaluating the health status of trees, and is a predetermined value. Each specific band corresponds to a correlation degree, and each specific band corresponds to a specific band spectral image, that is, each specific band spectral image corresponds to a correlation degree.

[0065] Before stacking, in order to ensure that the data of each channel is consistent in terms of spatial resolution, pixel size, etc., the multiple specific band spectral images are first normalized. Then, the stacking order of the multiple specific band spectral images is determined according to the correlation degree.

[0066] In some embodiments, the stacking order is positively correlated with the magnitude of the correlation. The greater the correlation, the earlier it is stacked. Stacking in descending order of correlation can put the spectral data with a large correlation in the front, so that these spectral data with a greater impact on the health assessment results receive more attention during data processing and analysis, which helps to highlight the key features in tree health assessment.

[0067] Multiple specific band spectral images are stacked in order of their corresponding correlation degrees to generate a multi-channel image. Each pixel in the multi-channel image contains the spectral information of the position under multiple specific bands. The multi-channel image corresponds to the permutation and combination relationship between spectral images of different bands in the stacking order, and the health level is determined based on the permutation and combination relationship.

[0068] By combining these channels, the contrast of the image can be enhanced, making the health of the trees more obvious. External factors such as temperature and light may interfere with the image of a single channel. Through the joint analysis of multi-channel images, the correlation between different channels can be used to reduce the impact of these interference factors and improve the accuracy and credibility of health description information. In addition, during the training process of the tree health grading model, the use of multi-channel images helps the model learn key features faster, thereby improving the accuracy and generalization ability of the model.

[0069] The stacked multi-channel images contain detailed information about the trees in multiple key spectral bands, which together reflect the health of the trees. The multi-channel images are then input into a pre-trained tree health grading model, which can extract features from the input images and classify the health level of the trees based on these features. Ultimately, the model outputs a health level representing the health of the trees, which can be used for subsequent timber volume calculations.

[0070] The tree health classification model used above is a pre-trained model. Figure 2 As shown, Figure 2 The present invention provides a flow chart of a tree health classification model training process, wherein the tree health classification model training process comprises the following steps:

[0071] Step 201, based on the deviation value between the theoretical data and the measured data of each sample tree in the sample area, the loss rate of the sample tree is calculated, and the tree health level corresponding to the loss rate is determined as the sample health level of the sample tree, and the loss rate is positively correlated with the size of the deviation value;

[0072] Step 202, selecting the spectral data of the specific band used to present the health status of the tree from the sample band spectrum as the sample specific band spectral data of the sample tree;

[0073] Step 203: Perform model training based on the sample health grade and the sample specific band spectral data to obtain the tree health grading model.

[0074] In some embodiments, before step 201, a sample feature map having tree appearance information and tree spectral information is pre-constructed in the sample area. The construction process refers to the construction process of the aforementioned target feature map, which will not be described in detail here. Afterwards, the sample feature map is segmented to obtain a sample single tree feature map of the selected sample tree. The sample single tree feature map has the sample tree appearance information and the sample tree spectral information. Based on the sample tree appearance information in the sample single tree feature map, the sample theoretical appearance morphological data such as the sample theoretical plant height, sample theoretical crown width and sample theoretical crown volume of the sample tree are determined. The segmentation process and the sample theoretical appearance morphological data refer to the segmentation process of the aforementioned target feature map and the determination process of the external morphological data, respectively, which will not be described in detail here.

[0075] The sample tree characteristic map carries location information, and the sample trees can be found in the forest area based on the location information, thus realizing field measurement of some data.

[0076] In some embodiments, after finding the sample trees, the measured crown width and measured crown volume of the sample trees are measured. The crown layer is cut down and removed, and the trees are measured again to obtain the measured plant height, wood density, measured mass, measured timber volume, and bark diameters at multiple set heights of the sample trees. The bark diameters at multiple set heights may be bark diameters at 0.1H, 0.2H, 0.3H, 0.4H, 0.5H, 0.6H, 0.7H, 0.8H, and 0.9H relative heights of the trunks, which are only examples, and the specific set heights are set as needed.

[0077] In some embodiments, the plant height correction coefficient, crown width correction coefficient and crown volume correction coefficient are determined respectively according to the sample theoretical plant height, sample theoretical crown width and sample theoretical crown volume and the measured plant height, measured crown width and measured crown volume.

[0078] In some embodiments, the theoretical data is the theoretical volume of the sample data, and the measured data is the measured volume of the sample trees. The difference between the theoretical volume and the measured volume is the volume deviation value. The ratio of the volume deviation value to the measured volume is calculated to obtain the loss rate of the sample trees.

[0079] In one embodiment, the process of determining the theoretical volume of the sample trees is as follows: obtaining the bark diameters of each of the sample trees at multiple set heights; and calculating the theoretical volume of the sample trees based on the bark diameters at the multiple set heights and the differentiated quadrature method.

[0080] In some embodiments, the theoretical volume of the sample trees may also be calculated by central section area method, average section area method or Dancing Lue algorithm, and any data required may be measured during the calculation process.

[0081] In some embodiments, the water content of trees affects the volume of trees. Considering the water content, the theoretical data is determined as the theoretical mass of the sample trees, and the measured data is determined as the measured mass of the sample trees. Theoretical mass of the sample trees = theoretical volume × wood density. The difference between the theoretical mass and the measured mass is the mass deviation value. The ratio of the mass deviation value to the measured mass is calculated to obtain the loss rate of the sample trees.

[0082] When the measured volume or mass remains unchanged, the greater the deviation value, the greater the loss rate.

[0083] In some embodiments, a comparison table of loss rate and tree health level is pre-built. As shown in Table 1, Table 1 is a corresponding relationship table of loss rate L and tree health level. Table 1 is only an example, and the specific corresponding relationship depends on the actual situation.

[0084] Table 1 Correspondence between loss rate L and tree health level

[0085] Loss rate L Tree Health Grade L≤5％ Level 1 5％<L≤10％ Level 2 10％<L≤15％ Level 3 L>15％ Level 4

[0086] In the above embodiment, the sample health level of the sample tree is determined by looking up a table, that is, the tree health level corresponding to the loss rate is determined as the sample health level of the sample tree.

[0087] A large number of sample trees are selected in the sample area to ensure that each tree health level has corresponding tree spectral information.

[0088] In some embodiments, the tree spectral information may be visible light spectral information of 490nm-780nm and near infrared spectral information of 780nm-1100nm.

[0089] In some embodiments, if the loss rate calculated based on the selected sample trees cannot cover the tree health levels, for example, there are loss rates of L≤5%, 5%<L≤10%, and 10%<L≤15%, and there is a lack of loss rates greater than 15%, it is necessary to expand the sampling range and supplement the data. For example, when it is known which area has relatively serious pest and disease problems, trees are selected from this area as supplementary data sources to obtain sample trees with a loss rate greater than 15% and the corresponding tree spectral information. In this way, it is ensured that there is corresponding tree spectral information for each tree health level.

[0090] In some embodiments, a calculation formula for the loss rate and the tree health level is pre-constructed, and the loss rate of the sample trees is substituted into the calculation formula to calculate the sample health level of the sample trees.

[0091] In some embodiments, step 202 specifically includes: extracting the spectral data of each band corresponding to different tree health levels from the sample band spectra; based on the spectral data of each band corresponding to different tree health levels, determining the numerical change information of the spectral data of each band in the case where the tree health level changes from the first level to the second level; based on the numerical change information, determining the correlation degree between each band and the tree health level; based on the correlation degree, selecting a target band from the sample bands as the specific band; and selecting the sample specific band spectral data from the sample band spectra according to the specific band.

[0092] Each tree health level corresponds to tree spectral information, that is, the sample band spectra. The spectral data of each band corresponding to the corresponding tree health level is extracted from the sample band spectra, such as reflectivity, emissivity, etc.

[0093] In one embodiment, the sample band spectra are spectra vulnerable to pests and diseases. For example, the content of chlorophyll is vulnerable to pests and diseases, and its corresponding spectrum is regarded as a spectrum vulnerable to pests and diseases.

[0094] Based on the spectral data of each band corresponding to different tree health levels, determine the numerical change information of the spectral data of each band in the case where the tree health level changes from the first level to the second level, such as when the tree health level changes from level 1 to level 2. Calculate the change correlation degree between the level change amount and the spectral change information to obtain the correlation degree between each band and the tree health level.

[0095] In one embodiment, the correlation degrees are sorted from large to small, and the correlation degrees are successively accumulated in descending order. When the sum of the accumulated correlation degrees is greater than a set accumulation threshold, such as 90%, the target band corresponding to the correlation degree participating in the accumulation calculation is used as the specific band, and the correlation degree of the target band is greater than that of the remaining bands.

[0096] In one embodiment, the relevance of each band is compared with a set relevance threshold, and the band corresponding to the relevance greater than or equal to the set relevance threshold is regarded as a target band, that is, a specific band.

[0097] Based on the correlation, the most important bands relative to the tree health level are selected, and those bands that are irrelevant or weakly correlated with the health level are excluded, thereby reducing data redundancy. This can not only reduce the cost of data processing and storage, but also improve the speed and efficiency of data processing.

[0098] After determining the specific band, the sample specific band spectrum data corresponding to the specific band is selected from the sample band spectrum. That is, the specific spectrum is also used to determine the specific band spectrum data in the target area.

[0099] In some embodiments, the stacking order of multiple sample specific band spectral images is determined according to the correlation between each specific band and the health level, and based on the stacking order, the multiple sample specific band spectral images are stacked to obtain a sample multi-channel image corresponding to each sample health level.

[0100] In some embodiments, based on the sample health level and the corresponding sample multi-channel image, a health grading training set and a health grading test set are constructed, and a deep learning method such as a convolutional neural network is used to perform model training based on the health grading training set and the health grading test set to obtain the tree health grading model. The tree health grading model outputs the health level of the tree according to the input multi-channel image.

[0101] In some embodiments, a health grading training set and a health grading test set are constructed directly based on the sample health grade and the corresponding sample specific band spectral data, and a deep learning method such as a convolutional neural network is used to perform model training based on the health grading training set and the health grading test set to obtain the tree health grading model. Accordingly, the tree health grading model outputs the tree health grade according to the input specific band spectral data.

[0102] The construction of a tree health grading model helps to speed up the assessment of health levels, thereby improving the speed and accuracy of timber volume calculation. At the same time, the tree health grading model can determine the health level of trees based on spectral information, saving the manpower cost used to determine the health status of trees.

[0103] Step 103: determining the individual wood volume of each of the trees based on the external morphological data and the health description information.

[0104] The external morphological data such as plant height, crown width and crown volume and health description information such as health grade of the tree are input into the volume prediction model, and the corresponding volume of the tree is output by the volume prediction model.

[0105] In some embodiments, the volume of a single tree is determined based on a volume prediction model, and the training process of the volume prediction model includes: obtaining sample external morphological data of each of the sample trees; calculating the corrected volume of the sample trees based on the loss rate and theoretical volume of each of the sample trees; and performing model training based on the sample external morphological data, the sample health level and the corrected volume to obtain the volume prediction model.

[0106] The sample external morphological data include the measured plant height, measured crown width and measured crown volume of the sample trees.

[0107] In some embodiments, the loss rate is calculated based on the theoretical mass and the measured mass, taking into account the water content. The theoretical volume is corrected using the loss rate to obtain a corrected volume, i.e., corrected volume = theoretical volume × (1-loss rate). The theoretical volume is obtained based on the skin diameter at multiple set heights and the differentiated area method.

[0108] In yet another embodiment, the measured volume is directly used as the corrected volume.

[0109] Based on the sample external morphological data, sample health grade and corrected timber volume, the timber volume training set and timber volume test set are constructed, and the model training is carried out using machine learning methods such as Lasso Regression (Least Absolute Shrinkage and Selection Operator, LASSO) method, ElasticNet method, Ridge Regression (Ridge), Gradient Boosting method or eXtreme Gradient Boosting (XGBoost) method to obtain the timber volume prediction model. The timber volume prediction model can estimate the single timber volume of the tree according to the input tree external morphological data and health grade.

[0110] It should be noted that if the health description information is not the health level but health description information in other forms of expression, it can be replaced accordingly during the training process of the timber volume prediction model.

[0111] This step combines external morphological data and health description information to more comprehensively evaluate the individual tree volume. Individual tree volume is an important data for forest timber volume calculation and can improve the accuracy of forest timber volume calculation.

[0112] Step 104: Calculate the overall forest volume of the target area based on the individual wood volumes of the plurality of trees in the target area.

[0113] The overall forest volume is the sum of all individual trees in the target area. The individual trees in the target area are summed to obtain the overall forest volume in the target area.

[0114] Based on external morphological data and health description information, the individual timber volume of each tree is accurately predicted. Correspondingly, the accuracy of the overall forest timber volume calculated based on the greatly improved individual timber volume is also improved accordingly.

[0115] In some embodiments, the actual timber volume of the forest area can be tracked and counted, and compared with the overall timber volume to determine the overall timber volume correction factor to further improve the accuracy of the overall timber volume.

[0116] This application achieves high-accuracy estimation of forest timber volume based on external morphological data and health description information, which solves the defect of traditional productivity assessment methods that only estimate based on external morphological data, fills the gap in forest timber volume estimation regarding the losses caused by pests and diseases during the growth process of trees and the estimation of effective timber volume, and has stronger scientificity, trustworthiness, objectivity and repeatability.

[0117] The timber volume estimated based on this application is more accurate, which will help play an important role in industries such as forest resource surveys, strategic timber reserves, timber trading, carbon sink assessment, forest asset assessment, financial insurance value assessment, and loan evaluation.

[0118] In the embodiment of the present application, both the external morphological data of each tree in the target area and the specific band spectral data of each tree are obtained. By introducing the specific band spectral data, the errors and limitations caused by single data are reduced, and the accuracy and reliability of subsequent timber volume calculations are improved. Among them, the specific band spectral data corresponds to a specific band, and the specific band is associated with the tree health information. In this way, based on the specific band spectral data, the health description information of each tree is determined to achieve an accurate assessment of the health status of the trees. Then, based on the external morphological data and the health description information, the single wood volume of each tree is determined. This process takes into account the health status of the trees, so that the accuracy of the timber volume of each tree is improved. Under the premise that the accuracy of the single wood volume is greatly improved, the accuracy of the overall forest volume of the target area calculated based on the single wood volume of multiple trees in the target area is also improved accordingly.

[0119] See also Figure 3 , Figure 3 This is a structural diagram of a forest timber volume estimation device provided in an embodiment of the present application. For ease of explanation, only the parts related to the embodiment of the present application are shown.

[0120] The forest timber volume estimation device 300 includes: an acquisition module 301 , a first determination module 302 , a second determination module 303 , and a calculation module 304 .

[0121] The acquisition module is used to acquire the external morphological data and specific band spectral data of each tree in the target area; the specific band spectral data corresponds to a specific band, and the specific band is associated with tree health information.

[0122] The first determination module is used to determine the health description information of each of the trees based on the spectral data of the specific band.

[0123] The second determination module is used to determine the individual wood volume of each of the trees based on the external morphological data and the health description information.

[0124] A calculation module is used to calculate the overall forest volume of the target area based on the single wood volume of multiple trees in the target area.

[0125] In some embodiments, the acquisition module is specifically used to:

[0126] Construct a target feature map with tree appearance information and tree spectrum information in the target area;

[0127] Performing feature segmentation on the target feature map according to the trees to obtain single-tree feature maps corresponding to the individual trees, each of the single-tree feature maps containing single-tree appearance information and single-tree spectrum information of the individual trees;

[0128] Based on the single tree appearance information in the single tree feature map, determining the plant height, crown width and crown volume of the tree as the external morphological data;

[0129] According to the specific band, spectral data is extracted from the single tree spectral information of the single tree characteristic image to obtain the spectral data of the specific band.

[0130] In some embodiments, the health description information is a health level, the specific band spectral data is a plurality of specific band spectral images, and the first determination module is specifically used to:

[0131] Obtaining the correlation between each of the specific wavebands and the health level;

[0132] Based on the correlation degree, determining a stacking order of the plurality of spectral images of the specific wavelength band;

[0133] Based on the stacking sequence, a plurality of the spectral images of the specific bands are stacked to obtain a multi-channel image, wherein the multi-channel image contains the arrangement and combination relationship between the spectral images of different bands in the stacking sequence, and different arrangement and combination relationships reflect different health levels;

[0134] The multi-channel image is input into a tree health grading model to obtain the health grade of the tree.

[0135] In some embodiments, the apparatus further comprises a model training module for:

[0136] Based on the deviation value between the theoretical data and the measured data of each sample tree in the sample area, the loss rate of the sample tree is calculated, and the tree health level corresponding to the loss rate is determined as the sample health level of the sample tree, and the loss rate is positively correlated with the size of the deviation value;

[0137] Selecting the spectrum data of the specific band used to present the health status of the tree from the sample band spectrum as the sample specific band spectrum data of the sample tree;

[0138] Model training is performed based on the sample health grade and the sample specific band spectral data to obtain the tree health grading model.

[0139] Extracting spectral data of each band corresponding to different tree health levels from the sample band spectrum;

[0140] Based on the spectral data of each band corresponding to different tree health levels, determine the value change information of the spectral data of each band when the tree health level changes from the first level to the second level;

[0141] Based on the value change information, determining the correlation between each band and the tree health level;

[0142] Selecting a target band from the sample bands as the specific band based on the correlation degree;

[0143] According to the specific waveband, the sample specific waveband spectrum data is selected from the sample waveband spectrum.

[0144] In some embodiments, the single wood volume is determined based on a wood volume prediction model, and the model training module is further used to:

[0145] Acquiring sample external morphological data of each of the sample trees;

[0146] Calculating the corrected volume of the sample trees based on the loss rate and theoretical volume of each of the sample trees;

[0147] Model training is performed based on the sample external morphological data, the sample health level and the corrected timber volume to obtain the timber volume prediction model.

[0148] In some embodiments, the computing module is further configured to:

[0149] Obtaining the bark diameter of each of the sample trees at a plurality of set heights;

[0150] The theoretical volume of the sample tree is calculated based on the bark diameters at the plurality of set heights and the differentiated quadrature method.

[0151] The device for estimating the forest volume provided in the embodiment of the present application can implement each process of the embodiment of the above-mentioned method for estimating the forest volume, and can achieve the same technical effect. To avoid repetition, it will not be described again here.

[0152] Figure 4 4 is a structural diagram of a terminal provided in an embodiment of the present application. As shown in the figure, the terminal 4 of this embodiment includes: at least one processor 40 ( Figure 4 Only one is shown in the figure), a memory 41, and a computer program 42 stored in the memory 41 and executable on the at least one processor 40, wherein the processor 40 implements the steps of any of the above-mentioned method embodiments when executing the computer program 42.

[0153] The terminal 4 may be a computing device such as a desktop computer, a notebook, a PDA, or a cloud server. The terminal 4 may include, but is not limited to, a processor 40 and a memory 41. Those skilled in the art will appreciate that Figure 4 It is only an example of terminal 4 and does not constitute a limitation on terminal 4. It may include more or fewer components than shown in the figure, or a combination of certain components, or different components. For example, the terminal may also include input and output devices, network access devices, buses, etc.

[0154] The processor 40 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc.

[0155] The memory 41 may be an internal storage unit of the terminal 4, such as a hard disk or memory of the terminal 4. The memory 41 may also be an external storage device of the terminal 4, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the terminal 4. Further, the memory 41 may also include both an internal storage unit and an external storage device of the terminal 4. The memory 41 is used to store the computer program and other programs and data required by the terminal. The memory 41 may also be used to temporarily store data that has been output or is to be output.

[0156] The technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0157] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0158] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0159] In the embodiments provided in the present application, it should be understood that the disclosed devices / terminals and methods can be implemented in other ways. For example, the device / terminal embodiments described above are only schematic. For example, the division of the modules or units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0160] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0161] In addition, each functional unit in each embodiment of the present application 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.

[0162] If the integrated module / unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present application implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium. It should be noted that the content contained in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media do not include electric carrier signals and telecommunication signals.

[0163] The present application implements all or part of the processes in the above-mentioned embodiment method, and may also be implemented through a computer program product. When the computer program product runs on a terminal, the terminal can implement the steps in the above-mentioned method embodiments when executing the computer program product.

[0164] The embodiments described above are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A method for estimating forest timber volume, characterized in that: include: Acquire external morphological data and specific band spectral data of each tree in the target area; the specific band spectral data corresponds to a specific band, and the specific band is associated with tree health information; Determining health description information of each of the trees based on the specific band spectral data; Determining the individual wood volume of each of the trees based on the external morphological data and the health description information; The overall forest volume of the target area is calculated based on the individual wood volumes of the plurality of trees in the target area.

2. The method according to claim 1, characterized in that The method of obtaining the external morphological data and specific band spectral data of each tree in the target area includes: Construct a target feature map with tree appearance information and tree spectrum information in the target area; Performing feature segmentation on the target feature map according to the trees to obtain single-tree feature maps corresponding to the individual trees, each of the single-tree feature maps containing single-tree appearance information and single-tree spectrum information of the individual trees; Based on the single tree appearance information in the single tree feature map, determining the plant height, crown width and crown volume of the tree as the external morphological data; According to the specific band, spectral data is extracted from the single tree spectral information of the single tree characteristic image to obtain the spectral data of the specific band.

3. The method according to claim 2, characterized in that The health description information is a health grade, the specific band spectral data is a plurality of specific band spectral images, and the health description information of each tree is determined based on the specific band spectral data, including: Obtaining the correlation between each of the specific wavebands and the health level; Based on the correlation degree, determining a stacking order of the plurality of spectral images of the specific wavelength band; Based on the stacking sequence, a plurality of the spectral images of the specific bands are stacked to obtain a multi-channel image, wherein the multi-channel image contains the arrangement and combination relationship between the spectral images of different bands in the stacking sequence, and different arrangement and combination relationships reflect different health levels; The multi-channel image is input into a tree health grading model to obtain the health grade of the tree.

4. The method according to claim 3, characterized in that The training process of the tree health classification model includes: Based on the deviation value between the theoretical data and the measured data of each sample tree in the sample area, the loss rate of the sample tree is calculated, and the tree health level corresponding to the loss rate is determined as the sample health level of the sample tree, and the loss rate is positively correlated with the size of the deviation value; Selecting the spectrum data of the specific band used to present the health status of the tree from the sample band spectrum as the sample specific band spectrum data of the sample tree; Model training is performed based on the sample health grade and the sample specific band spectral data to obtain the tree health grading model.

5. The method according to claim 4, characterized in that The step of selecting the spectral data of the specific band used to present the health status of the tree from the sample band spectrum as the sample specific band spectral data of the sample tree comprises: Extracting spectral data of each band corresponding to different tree health levels from the sample band spectrum; Based on the spectral data of each band corresponding to different tree health levels, determine the value change information of the spectral data of each band when the tree health level changes from the first level to the second level; Based on the value change information, determining the correlation between each band and the tree health level; Selecting a target band from the sample bands as the specific band based on the correlation degree; According to the specific waveband, the sample specific waveband spectrum data is selected from the sample waveband spectrum.

6. The method according to claim 4, characterized in that The single wood volume is determined based on a wood volume prediction model, and the training process of the wood volume prediction model includes: Acquiring sample external morphological data of each of the sample trees; Calculating the corrected volume of the sample trees based on the loss rate and theoretical volume of each of the sample trees; Model training is performed based on the sample external morphological data, the sample health level and the corrected timber volume to obtain the timber volume prediction model.

7. The method according to claim 6, characterized in that The method further comprises: Obtaining the bark diameter of each of the sample trees at a plurality of set heights; The theoretical volume of the sample tree is calculated based on the bark diameters at the plurality of set heights and the differentiated quadrature method.

8. A device for estimating forest timber volume, characterized in that: include: An acquisition module, used to acquire external morphological data and specific band spectral data of each tree in the target area; the specific band spectral data corresponds to a specific band, and the specific band is associated with tree health information; A first determination module, configured to determine health description information of each of the trees based on the spectral data of the specific band; A second determination module, configured to determine the single wood volume of each of the trees based on the external morphological data and the health description information; A calculation module is used to calculate the overall forest volume of the target area based on the single wood volume of multiple trees in the target area.

9. A terminal comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images