JP6570039B2 - Forest resource information calculation method and forest resource information calculation device - Google Patents

Description

  The present invention relates to a forest resource information calculation method and a forest resource information calculation device.

  Information on forest resources (referred to as forest resource information) is important basic information in forest management. Conventionally, forest resource information is based on ground surveys where a researcher sets up a small survey area at a location that appears to be a standard forest and measures the position, diameter, and height of trees in the survey area. A method for estimating the entire forest resource information by multiplying the result by the management area is widely used.

  However, since forests are not uniform, there is a problem in accuracy in the method of multiplying the measurement result of a small area by the management area. In order to improve the estimation accuracy of forest resource information, it is conceivable to increase the number of on-site survey locations. However, increasing the number of on-site survey locations requires a great deal of labor and cost.

  On the other hand, in recent years, forest resource information is calculated in units of trees (single trees) based on laser measurement data (three-dimensional point cloud data) obtained by irradiating a forest area to be investigated from the sky. Things are also done.

Japanese Patent No. 4946072 Japanese Patent No. 4279894 Japanese Patent No. 5507418

  Patent Documents 1 and 2 extract tree vertices (referred to as tree vertices) based on laser measurement data, and Patent Document 3 extracts tree positions (referred to as tree positions) based on laser measurement data. To do.

  However, in the inventions described in Patent Documents 1 to 3, each of the trees (referred to as survey target trees) to be surveyed (referred to as survey target trees) in the forest area to be surveyed (referred to as survey target forest area). The crown corresponding to the single tree is not extracted with high accuracy.

  For this reason, when attempting to extract tree vertices using the inventions described in Patent Documents 1 and 2, the tree vertices may be excessively extracted, and the tree vertices of each tree cannot be extracted correctly. In addition, when trying to extract a tree position using the invention described in Patent Document 3, tree vertices may be excessively extracted as in the inventions described in Patent Documents 1 and 2. Cannot be extracted correctly. This is because, in the inventions described in Patent Documents 1 to 3, as described above, the tree crown corresponding to each tree is not extracted with high accuracy, and a plurality of tree vertices are included in one tree crown. Is due to the existence of. In this specification, “tree canopy” refers to a portion composed of branches and leaves of each tree when the trees constituting the forest are viewed from above.

  Furthermore, when obtaining the volume of each tree as one of the forest resource information, it is common to obtain from the two-variable volume formula of the chest height diameter and the tree height of each tree, but the inventions described in Patent Documents 1 to 3 In the case where the tree vertices may be excessively extracted as described above, the tree height of each tree cannot be obtained with high accuracy. In the inventions described in Patent Documents 1 to 3, Chest height diameter is not required. For this reason, even if the tree height of each tree is calculated, if the volume is calculated using only the tree height, there are trees with the same tree height but different chest height diameters due to differences in the number of standing trees and sparse density of the canopy. Since there are many, the volume cannot be obtained with high accuracy.

  As described above, in Patent Documents 1 to 3, since the crown of each tree is not extracted with high accuracy, the forest resource information such as the tree height, the chest height diameter, and the volume of material obtained from the tree height and the chest height diameter are not obtained. There is a problem that cannot be calculated with high accuracy.

  Therefore, the present invention has been made to solve the above-described problems, and forest resource information calculation that can calculate forest resource information with high accuracy by extracting the crown corresponding to each tree of the survey target tree with high accuracy. It is an object to provide a method and a forest resource calculation information determination device.

  [1] The forest resource information calculation method of the present invention creates forest resource information relating to the forest area to be surveyed based on laser measurement data obtained by irradiating the area including the forest area to be surveyed with laser light from above. A forest resource information calculation method, a survey target forest area image data creation processing step of creating survey target forest area image data based on the laser measurement data and geographical information, and the survey target forest area image data A crown height image data creation processing step for creating crown height image data representing the crown height of a tree existing in the survey target forest area, and a noise process for removing noise included in the crown height image data A canopy height image data noise processing step for creating noise-processed canopy height image data, and vegetation existing in the forest area to be investigated A precise crown image data creation processing step for creating a precise crown image data capable of extracting a crown corresponding to each tree to be surveyed, based on the noise-processed crown height image data; and A precise canopy information creation processing step for creating canopy information on the canopy corresponding to each tree from the precise canopy image data, and a forest resource information calculation process for creating forest resource information on the forest area to be investigated based on the precise canopy information And the noise included in the crown height image data includes laser measurement noise included in the laser measurement data, lower vegetation existing below the investigation target tree, and the laser light. Canopy transmission pulse generated by passing through the canopy and reaching the surface, and the chief protruding locally from the canopy And the crown height image data noise processing step performs the noise processing using mesh data of each mesh obtained by meshing the crown height image data, and the crown height image data noise processing step is performed. The high image data noise processing step includes a laser measurement noise processing step for removing the laser measurement noise, a lower vegetation processing step for creating a lower vegetation processing image data obtained by removing the lower vegetation from the crown height image data, Averaging processing image data is obtained by performing averaging processing of the crown height of each tree using an averaging filter that is set based on the average crown diameter of each tree for the lower vegetation processing image data. An averaging process step to be created, and a difference for taking a difference between the averaged image data and the lower vegetation processed image data Obtained by the processing step, the canopy transmission pulse processing step of creating the canopy transmission pulse processing image data from which the canopy transmission pulse is removed based on the difference result obtained by the difference processing step, and the difference processing step. And a crown contour processing step of creating crown contour processing image data that removes the length branches and leaves and improves the unclear portion of the crown contour based on the difference result.

  As described above, in the forest resource information calculation method of the present invention, the survey target forest area image data is created based on the laser measurement data obtained by irradiating the region including the survey target forest area from the sky with the laser beam, Creating canopy height image data based on the survey target forest area image data, and applying noise processing to remove noise from the canopy height image data, creating noise-treated canopy height image data, Precision tree image data is created based on the noise-processed tree crown height image data. The precise canopy image data created in this way becomes precise canopy image data that can extract each canopy corresponding to each tree to be investigated with high accuracy. Therefore, the precise canopy information created based on the precise canopy image data becomes highly accurate canopy information corresponding to each tree of the survey target tree, and based on the precise canopy information, information on forest resources (forest resource information ) Can be obtained highly accurate forest resource information (see, for example, FIGS. 18 to 20). As a result, highly accurate forest resource information with low measurement errors and high objectivity can be provided in the entire forest area to be surveyed, in each subgroup, and in any range, and can be used for forest management such as thinning and other forest management planning.

  [2] In the forest resource information calculation method of the present invention, meshed digital surface model data and digital elevation model data are created based on the forest area image data to be surveyed, and the digital surface model data in each mesh Preferably, the crown height image data is created by taking a difference between the digital elevation model data and the digital elevation model data.

  As a result, the crown height can be calculated with high accuracy. In this case, the digital surface model data refers to data representing the altitude of a surface layer such as a tree, and the digital elevation model data refers to data representing the altitude of the ground surface. Therefore, the crown height (the height of the crown from the ground surface) can be obtained by taking the difference between the digital surface model data and the digital elevation model data.

  [3] In the forest resource information calculation method of the present invention, in the lower vegetation processing step, a maximum crown height of the crown heights of the lower vegetation is input, and a mesh having a crown height less than or equal to the maximum crown height is input. It is preferable that the pixel value is zero.

  Thereby, the lower vegetation can be reliably removed as noise, and the lower vegetation can be displayed in black in the crown height image displayed on the display.

  [4] In the forest resource information calculation method of the present invention, the averaging processing step sets each mesh of the crown height image data as a processing target mesh, superimposes the averaging filter on the processing target mesh, It is preferable to sequentially perform the process of setting the average value of the crown height of each mesh included in the averaging filter as the crown height of the process target mesh in each process target mesh.

  By performing such averaging processing, the canopy transmission pulse is recognized as a higher crown height due to the influence of the surrounding canopy, etc. And is recognized as a lower crown height. For this reason, if the difference with the original image data (in this case, the lower vegetation processed image data from which the lower vegetation has been removed) is taken, the difference value becomes larger than the crown portion. Can be recognized reliably. Thereby, the process which removes a canopy permeation | transmission pulse and length branch leaves as noise can be performed reliably.

  [5] In the forest resource information calculation method of the present invention, the difference processing step includes subtracting the lower vegetation processing image data from the averaged processing image data to obtain a difference between the averaged processing image data and the lower vegetation processing image data. It is preferable to perform a difference process for obtaining a difference value on each mesh.

  In this way, by subtracting the lower vegetation processing image data from the averaged processing image data and obtaining the difference value between the averaged processing image data and the lower vegetation processing image data (original image data), it corresponds to the canopy transmission pulse. The difference value obtained in the mesh and the difference value obtained in the mesh corresponding to the long branch leaf can be obtained as a larger value than the difference value obtained in the mesh of the crown portion.

  [6] In the forest resource information calculation method of the present invention, the canopy transmission pulse processing step uses the difference value of each mesh obtained by the difference process as a horizontal axis, and the number of meshes corresponding to the difference value as a vertical axis. The histogram is created using the difference value of zero (0) as an average value as an axis, and in the histogram, the difference value of each mesh obtained by the difference processing is defined as D, and is on the plus side of the average value. When the standard deviation is SD, the standard deviation on the minus side of the average value is -SD, and the mesh to be processed at this time is the processing target mesh, the difference value D obtained in the processing target mesh is D If> SD, the averaged image data corresponding to the processing target mesh is used as the processing target mesh data, and the processing target message When the difference value D obtained in step S is D ≦ SD, the canopy transmission pulse processed image data is generated using the lower vegetation processed image data corresponding to the processing target mesh as the processing target mesh data. It is preferable to do.

  Using the histogram created in this way, it is determined whether or not the difference value D obtained in the processing target mesh satisfies D> SD. If D> SD, the processing target mesh When the corresponding averaged image data is the data of the processing target mesh and the difference value D obtained in the processing target mesh is D ≦ SD, the lower vegetation processing corresponding to the processing target mesh By generating the canopy transmission pulse processed image data using the image data as the processing target mesh data, the generated canopy transmission pulse processed image data is obtained by removing the canopy transmission pulse.

  [7] In the forest resource information calculation method of the present invention, the crown contour processing step corresponds to the processing target mesh when the difference value D obtained in the processing target mesh is D <−SD. The canopy transmission pulse processed image data is smoothed to create a smoothed image, and when the difference value D obtained in the processing target mesh is D ≧ −SD, the processing target mesh It is preferable that the crown contour processed image data is generated using the corresponding canopy transmission pulse processed image data as data of the processing target mesh.

  Also in this case, when the difference value D obtained in the processing target mesh using the above histogram is D <−SD, the processing target mesh is smoothed and obtained in the processing target mesh. When the difference value D is D ≧ −SD, the generated crown is generated by generating crown contour processed image data using the canopy transmission pulse processed image data corresponding to the processing target mesh as data of the processing target mesh. In the contour processing image data, the length branches and leaves are removed, and the contour of the crown is smooth.

  [8] In the forest resource information calculation method of the present invention, it is preferable that the precise canopy image data creation step creates the precise canopy image data by utilizing the fact that the edge of the crown is darker than the crown portion.

  Thereby, the crown corresponding to each tree of the investigation object tree can be extracted with high accuracy. Note that the creation of precise crown image data using the fact that the edge of the crown is darker than the crown can be automatically performed using the individual tree detection method (Individual tree detection method). Can be extracted.

  [9] In the forest resource information calculation method of the present invention, the precise canopy information creation processing step corresponds to at least the label number of the canopy corresponding to each tree and the respective trees using the precise canopy image data. It is preferable to calculate the crown position of the canopy, the crown diameter of the crown corresponding to each tree, and the crown area of the crown corresponding to each tree as the precise crown information.

  As a result, high-accuracy precise canopy information can be obtained as information on the canopy corresponding to each tree in the survey target tree, and information on forest resources (forest resource information) is calculated based on the precise canopy information. Highly accurate forest resource information can be obtained.

  [10] In the forest resource information calculation method of the present invention, the tree height calculation processing step for calculating the tree height of each tree, and the diameter of the human breast height in each tree is calculated as the breast height diameter of each tree. Chest height diameter calculation processing step, calculation step of calculating the total volume of each tree, calculation of the tree height of each tree, the chest height diameter of each tree and the volume of each tree and the accurate crown information creation processing step It is preferable to include a forest resource information aggregation processing step for performing a process of aggregating the precise tree crown information.

  By performing such processing, the tree height, breast height diameter, and volume of each surveyed tree can be calculated, and calculated tree height, breast height diameter, volume, and precise crown information creation step. Each tree with precise crown information (such as the label number of the crown corresponding to each tree, the crown position of the crown corresponding to each tree, the crown area of the crown corresponding to each tree, the crown diameter of the crown corresponding to each tree, etc.) Resource information (referred to as single tree resource information). Then, the aggregated single tree resource information can be registered in the attribute database, and the single tree resource information registered in the attribute database can be displayed on the display as a list.

  [11] In the forest resource information calculation method of the present invention, the tree height of each tree is the maximum tree height in each canopy among the tree heights obtained by the canopy height image data creation step. It is preferable that

  Thereby, the tree height corresponding to each canopy can be obtained, and thereby the height (tree height) of each tree to be investigated can be calculated with high accuracy.

  [12] In the forest resource information calculation method of the present invention, the breast height diameter of each tree is calculated by a multiple regression equation from at least one of the crown area and crown diameter of the tree and the tree height of the tree. Is preferred.

  The breast height diameter has a strong correlation with the crown area, crown diameter, and tree height. For this reason, the breast height diameter can be calculated by a multiple regression equation from at least one of the crown area and the crown diameter and the tree height. In addition, the breast height diameter of unmeasured trees can be calculated by interpolation. This makes it possible to calculate the breast height diameter of all the survey target trees existing in the survey target forest area. In addition, when calculating multiple regression equations, it is possible to determine multiple regression equation variables by selecting about ten standard trees in the field survey for each tree type and measuring the crown area, crown diameter and tree height of the selected standard tree. preferable.

  [13] In the forest resource information calculation method of the present invention, the forest resource information calculation processing step in the forest resource information calculation processing step includes the number of the trees to be investigated and the crown diameter calculated for each tree. Average value, maximum and minimum values of crown diameters calculated for each tree, average value of crown area calculated for each tree, crown area calculated for each tree The maximum value and the minimum value, the average value of the breast height diameter calculated for each tree, the maximum value and the minimum value of the breast height diameter calculated for each tree, and for each tree Average value of calculated tree height, maximum value and minimum value of tree height calculated for each tree, average value of volume calculated for each tree, calculated for each tree The maximum volume, and With further has a processing to calculate the small value, it may further include a process for calculating the number and timber volume per unit area of the surveyed trees.

  The calculation result obtained is registered in the attribute database as forest resource summary information indicating the outline of the forest resource, and the forest resource summary information registered in the attribute database is displayed on the display as a list. Therefore, it is possible to get an overview of the forest resources in the forest area that corresponds to the survey.

[14] The forest resource information calculation apparatus of the present invention creates forest resource information relating to the forest area to be surveyed based on laser measurement data obtained by irradiating a region including the forest area to be surveyed from above with laser light. A forest resource information calculation device for creating a survey target forest area image data based on the laser measurement data and geographical information, and a survey target forest area image data A crown height image data creation processing unit for creating crown height image data representing a crown height of a tree existing in the forest area to be investigated, and noise processing for removing noise included in the crown height image data The canopy height image data noise processing unit that creates noise-processed canopy height image data, and the vegetation existing in the survey target forest area is the survey target. Precision crown image data creation processing unit that creates precise crown image data capable of extracting a crown corresponding to each tree to be investigated based on the noise-processed crown height image data, and the accurate crown image data A precision crown information creation processing unit that creates crown information about the crown corresponding to each tree; and a forest resource information calculation processing unit that creates forest resource information about the forest area to be investigated based on the precise crown information. In addition, the noise included in the crown height image data includes laser measurement noise included in the laser measurement data, lower vegetation existing below the investigation target tree, and the laser light transmitted through the crown. And canopy transmission pulses generated by reaching the surface of the earth, and prominent branches and crown contours protruding locally from the crown The canopy high image data noise processing unit performs the noise processing using mesh data of each mesh obtained by meshing the canopy high image data, and the canopy high image data noise processing unit includes the laser measurement. A laser measurement noise processing unit that removes noise, a lower layer vegetation processing unit that creates lower vegetation processing image data obtained by removing the lower vegetation from the crown height image data, and the lower vegetation processing image data, each tree An averaging processing unit that creates an averaged image data by performing an average processing of the crown height of each tree using an averaging filter set based on an average crown diameter, and the averaging processing Based on the difference processing unit that takes the difference between the image data and the lower vegetation processing image data, and the difference result obtained by the difference processing unit, A canopy transmission pulse processing unit that creates canopy transmission pulse processing image data from which a canopy transmission pulse has been removed, and based on the difference result obtained by the difference processing unit, the crown branch leaves are removed and the crown contour unclear part And a crown contour processing unit for creating image data with improved crown contour processing.
It is characterized by that.

  [15] In the forest resource information calculation apparatus according to the present invention, meshed digital surface model data and digital elevation model data are created based on the surveyed forest area image data, and the digital surface layer model data is generated in each mesh. Preferably, the crown height image data is created by taking a difference between the digital elevation model data and the digital elevation model data.

  [16] In the forest resource information calculation apparatus according to the present invention, the lower vegetation processing unit inputs a maximum crown height among the crown heights of the lower vegetation, and a mesh having a crown height less than or equal to the maximum crown height. It is preferable that the pixel value is zero.

  [17] In the forest resource information calculation apparatus of the present invention, the averaging processing unit sets each mesh of the crown height image data as a processing target mesh, superimposes the averaging filter on the processing target mesh, It is preferable to sequentially perform the process of setting the average value of the crown height of each mesh included in the averaging filter as the crown height of the process target mesh in each process target mesh.

  [18] In the forest resource information calculation device according to the present invention, the difference processing device subtracts the lower vegetation processing image data from the averaged processing image data to obtain a difference between the averaged processing image data and the lower vegetation processing image data. It is preferable to perform a difference process for obtaining a difference value on each mesh.

  [19] In the forest resource information calculation apparatus of the present invention, the canopy transmission pulse processing unit uses the difference value of each mesh obtained by the difference process as a horizontal axis, and the number of meshes corresponding to the difference value as a vertical axis. The histogram is created using the difference value of zero (0) as an average value as an axis, and in the histogram, the difference value of each mesh obtained by the difference processing is defined as D, and is on the plus side of the average value. When the standard deviation is SD, the standard deviation on the minus side of the average value is -SD, and the mesh to be processed at this time is the processing target mesh, the difference value D obtained in the processing target mesh is D If> SD, the averaged image data corresponding to the processing target mesh is used as data of the processing target mesh, and the processing target mesh When the difference value D obtained in the above is D ≦ SD, the canopy transmission pulse processed image data is generated using the lower vegetation processed image data corresponding to the processing target mesh as the processing target mesh data. It is preferable to do.

  [20] In the forest resource information calculation apparatus according to the present invention, the crown contour processing unit corresponds to the processing target mesh when the difference value D obtained in the processing target mesh is D <−SD. The canopy transmission pulse processed image data is smoothed to create a smoothed image, and when the difference value D obtained in the processing target mesh is D ≧ −SD, the processing target mesh It is preferable that the crown contour processed image data is generated using the corresponding canopy transmission pulse processed image data as data of the processing target mesh.

  [21] In the forest resource information calculation apparatus of the present invention, it is preferable that the precise canopy image data creation unit creates the precise canopy image data by utilizing the fact that the edge of the canopy is darker than the crown portion.

  [22] In the forest resource information calculation apparatus according to the present invention, the precise canopy information creation unit uses the precise canopy image data to correspond to at least the label number of the canopy corresponding to each tree and each tree. It is preferable that the crown position of the crown, the crown diameter of the crown corresponding to each tree, and the crown area of the crown corresponding to each tree are calculated as the precise crown information.

  [23] In the forest resource information calculation apparatus of the present invention, the forest resource information calculation unit includes a tree height calculation processing unit that calculates the tree height of each tree, and a diameter of a human breast height in each tree. Chest height diameter calculation processing unit for calculating the chest height diameter of each tree, a material totalization processing unit for calculating the volume of each tree, the tree height of each tree, the chest height diameter of each tree, and the volume of each tree and the precision It is preferable to have a forest resource information aggregation processing unit that performs a process of aggregating the precise crown information calculated by the tree crown information creation processing unit.

  [24] In the forest resource information calculation apparatus of the present invention, the tree height of each tree is the maximum tree height in each tree crown among the tree heights determined by the tree height image data creation unit. It is preferable that

  [25] In the forest resource information calculation apparatus of the present invention, the breast height diameter of each tree is calculated by a multiple regression equation from at least one of the crown area and crown diameter of the tree and the tree height of the tree. Is preferred.

  [26] In the forest resource information calculation apparatus of the present invention, the forest resource information totalization processing unit in the forest resource information calculation processing unit is configured to calculate the number of trees to be examined and the crown diameter calculated for each tree. Average value, maximum and minimum values of crown diameters calculated for each tree, average value of crown area calculated for each tree, crown area calculated for each tree The maximum value and the minimum value, the average value of the breast height diameter calculated for each tree, the maximum value and the minimum value of the breast height diameter calculated for each tree, and for each tree Average value of calculated tree height, maximum value and minimum value of tree height calculated for each tree, average value of volume calculated for each tree, calculated for each tree Calculate maximum and minimum values of volume That process with further having, further it is preferable to perform processing to calculate the number and timber volume per unit area of the surveyed trees.

  In addition, in the forest resource information calculation apparatus of the present invention described in [14] to [26], the same effect as that obtained by the corresponding forest resource information calculation method of the present invention described in [1] to [13] is obtained. The effect is obtained.

It is a flowchart shown in order to demonstrate the forest resource information calculation method which concerns on embodiment. It is a figure which shows the laser measurement image displayed on the display. It is a figure which shows the forest area image of the investigation object displayed on the display. It is an enlarged view of the crown height image displayed on the display. It is a schematic diagram for demonstrating the noise contained in tree crown high image data. It is a flowchart shown in order to demonstrate a canopy high image data noise process. It is a figure which shows the laser measurement noise process image displayed on the display. It is a figure which shows the lower layer vegetation process image displayed on the display. It is a figure which shows the averaging process image displayed on the display. It is a figure which shows the difference process image displayed on the display. It is a figure which shows the histogram produced based on the difference result obtained by the difference process. It is a figure which shows the canopy transmission pulse processing image displayed on the display. It is a figure which shows the smoothing process image displayed on the display. It is a figure which shows the canopy outline process image displayed on the display. It is a figure which shows the precision tree crown image displayed on the display. It is a figure which shows the label number assignment | providing image by which the label number was attached | subjected to each tree crown displayed on the display. It is a figure which shows the tree height calculation image displayed on the display. It is a figure which shows the list of single tree resource information displayed on the display. It is a figure which shows the list of the forest resource summary information displayed on the display. It is a figure which shows an example of the simulation of the thinning plan displayed on the display. It is a figure shown in order to demonstrate the forest resource information calculation apparatus 1 which concerns on embodiment.

  FIG. 1 is a flowchart for explaining the forest resource information calculation method according to the embodiment. The forest resource information calculation method according to the embodiment is a forest resource that creates forest resource information related to a survey target forest area based on laser measurement data obtained by irradiating a region including the target forest area from the sky with laser light. This is an information calculation method. In the embodiment, it is assumed that the drone is used to irradiate a region including the forest area to be surveyed with laser light.

  As shown in FIG. 1, the processing performed in the forest resource information calculation method according to the embodiment is a data input process for inputting laser measurement data (three-dimensional point cloud data) to a data input unit of an information processing apparatus such as a personal computer. (Step S10), survey target forest area image data processing (step S20) for creating survey target forest area image data based on the laser measurement data and the geographical information obtained from the geographic information system (GIS) 100; Canopy height image data creation processing (step S30) for creating crown height image data representing the crown height of a tree existing in the survey target forest area based on the survey target forest area image data, and noise included in the crown height image data Noise image processing to remove noise and create noise-processed crown height image data Processing (step S40), and accurate canopy image data that can extract a crown corresponding to each tree of the survey target tree to be surveyed among the vegetation existing in the survey target forest area, canal height image data that has been subjected to noise processing Precision canopy image data creation processing (step S50) created based on the above, and precise canopy information creation processing (step S60) for creating canopy information regarding the canopy corresponding to each tree (each tree of the survey target tree) from the precise canopy image data ) And forest resource information calculation processing (step S70) for creating forest resource information regarding the forest area to be investigated based on the precise tree crown information. In addition, when the “survey target tree” is described as a single tree, it may be simply abbreviated as “each tree”.

  In the forest resource information calculation method according to the embodiment, the survey target trees are planted trees that are high trees such as larch, red pine, cedar, cypress, etc., and the survey target forest areas are these It is assumed that the forest formed by the trees is a forest area having a predetermined area.

  Here, the noise included in the canopy height image data includes laser measurement noise included in the laser measurement data, understory vegetation such as shrubs that exist below the surveyed tree, and laser light transmitted through the canopy. Canopy transmission pulse generated by reaching the surface of the earth, canopy branches and leaves that protrude locally from the canopy, and canopy contours where the outline of each canopy is unclear due to overlapping of multiple crowns ing. Each of these noises will be described in detail later.

  Further, the canopy height image data noise processing (step S40) performs noise processing as mesh data of each mesh obtained by meshing the canopy height image data. And in the said crown high image data noise process (step S40), the laser measurement noise process (step S41) which removes a laser measurement noise, and the lower vegetation process image data which removed the lower vegetation from the tree height image data are produced. For the lower vegetation process (step S42) and the lower vegetation process image data, the crown height of each tree is averaged using an averaging filter set based on the average crown diameter of each tree. The averaging process (step S43) for creating the averaged image data, the difference process for taking the difference between the averaged image data and the lower vegetation process image data (step S44), and the difference obtained by the difference process Based on the results, canopy transmission pulse processing (step data) that creates canopy transmission pulse processing image data from which canopy transmission pulses have been removed 45) and crown contour processing (step S46) for creating crown contour processing image data that removes the length branch leaves and improves the crown contour unclear portion based on the difference result obtained by the difference processing. It is. Specific processing contents of these processes (steps S41 to S46) will be described later.

  The precise canopy information creation process (step S60) uses the precise canopy image data, at least the label number as the serial number of the canopy corresponding to each tree, the canopy position of the canopy corresponding to each tree, and the canopy corresponding to each tree. Canopy diameter and the crown area of the canopy corresponding to each tree are calculated as precise canopy information. It is also possible to calculate the crown shape of the crown corresponding to each tree as the precise crown information.

  Further, in the forest resource information calculation process (step S70), a tree height calculation process (step S71) for calculating the tree height of each tree, and the diameter of the breast height of a human in each tree is calculated as the breast height diameter of each tree. Chest height diameter calculation processing (step S72), volume resource calculation processing for calculating the volume of each tree (step S73), tree height of each tree, chest height diameter of each tree, volume of each tree, and precise crown information creation processing ( Forest resource information totaling processing (step S74) for performing processing for totaling the precise tree crown information calculated in step S60) is included. Specific processes of these processes (steps S71 to S74) will be described later.

  In addition, the forest resource information calculation process (step S74) in the forest resource information calculation process (step S70) is calculated for the number of survey target trees, the average value of the crown diameter calculated for each tree, and each tree. The maximum and minimum values of the crown diameter, the average value of the crown area calculated for each tree, the maximum and minimum values of the crown area calculated for each tree, and for each tree Average breast height diameter calculated, average breast height diameter calculated for each tree, maximum and minimum breast height diameters calculated for each tree, calculated for each tree Average tree height, maximum and minimum tree heights calculated for each tree, average volume calculated for each tree, maximum volume calculated for each tree Further processing to calculate values and minimum values Together, further comprising a process for calculating the number and timber volume per unit area of the surveyed trees.

  By the way, the above-mentioned “crown diameter” represents the horizontal extent of the crown when viewed in plan, and the “chest height diameter” represents the thickness of the trunk of the tree. Although the crown and the trunk of the tree are not actually a perfect circle, it is not strictly a “diameter”. However, in this specification, the “crown” represents the size of the horizontal spread of the tree. “Crown diameter” is written, and “crown diameter” is written as the thickness of the trunk.

  Hereinafter, the forest resource information calculation method according to the embodiment will be described in detail along the processing procedure.

  In the forest resource information calculation method according to the embodiment, as shown in FIG. 1, laser measurement data is input to a data input unit of an information processing apparatus such as a personal computer (step S10), and obtained by laser measurement data on a display. The obtained image (referred to as laser measurement image) is displayed.

  FIG. 2 is a diagram showing a laser measurement image displayed on the display. In the laser measurement image shown in FIG. 2, the portion displayed in white indicates the crown of the investigation target tree (referred to as larch). In addition, when performing each process in each subsequent step and each component, the forest area is meshed from, for example, several centimeters to about 1 meter with respect to the laser measurement data, and the data obtained in each mesh Processing is performed using (mesh data). If the mesh size is too large, there may be a problem in the accuracy of the calculated forest resource information. Conversely, if the mesh size is too small, the amount of data to be processed increases. Therefore, it is preferable to set the mesh size to an optimum size in consideration of the area of the forest area to be surveyed, the state of the forest in the forest area to be surveyed, the accuracy of the calculated forest resource information, and the like.

  The input laser measurement data includes geographical information obtained from the geographic information system (GIS) 100 (in this case, forest area boundary data registered in a forest management ledger etc. maintained in prefectures, etc.) ) Is created (step S20), and the survey target forest area image corresponding to the survey target forest area image data is displayed on the display.

  FIG. 3 is a diagram showing the forest area image to be surveyed displayed on the display. The survey target forest area image shown in FIG. 3 is an image corresponding to the survey target forest area image data, and can be created by superimposing the survey target forest area extracted from the GIS 100 on the laser measurement data.

  Subsequently, a canopy height image data creation process (step S30) is performed. The canopy height image data creation process (step S30) is based on the survey target forest area image data, and canopy height image data representing the crown height (crown height from the ground surface) of a tree existing in the survey target forest area. create. Specifically, digital surface model data meshed based on the survey target forest area image data created by the survey target forest area image data creation process (step S20) (data representing the altitude of the surface layer such as trees) And digital elevation model data (data representing the elevation of the ground surface), and by taking the difference between the digital surface model data and the digital elevation model data for each mesh, the crown height of each mesh (the height of the crown from the ground surface) ) Is created. And the crown height image corresponding to the said crown height image data created in this way is displayed on the display.

  FIG. 4 is an enlarged view of the crown height image displayed on the display. The gradation from black to white in the crown height image shown in FIG. 4 is a change in height (for example, 0.0 meters (black) to 33.1 meters) with respect to the reference when the ground surface is used as the reference (0.0 meters). (White)). Note that the canopy height image data corresponding to the canopy height image shown in FIG. 4 includes various noises (described later), and a plurality of tree canopies are overlapped, so that it is difficult to extract the canopy corresponding to each tree. It has become a thing.

  Here, as described above, examples of various noises included in the crown height image data include laser measurement noise, crown transmission pulse, crown contour indistinct portion, canopy branches in the crown, and lower vegetation. These noises are noises that become obstacles when the crowns corresponding to the respective trees are extracted with high accuracy, and are indicated by symbols a to e in the crown term image shown in FIG. Hereinafter, laser measurement noise a, crown transmission pulse b, crown contour indistinct portion, long branch leaf, undergrowth vegetation, laser measurement noise a, crown transmission pulse b, crown contour unclear portion c, crown branch leaf It may be expressed as d, lower vegetation e. In addition, the laser measurement noise a, the canopy transmission pulse b, the canopy contour unclear portion c, the long foliage d in the canopy, and the lower vegetation e are shown as one representative portion in FIG. In reality, however, each of them exists in many places.

  FIG. 5 is a schematic diagram for explaining noise included in the crown height image data. FIG. 5 shows a schematic view of a forest standing on the ground surface L when the human is standing in the horizontal direction. The laser measurement noise a, the canopy transmission pulse b, and the canopy contour included in the canopy height image data. The obscured part c, the long foliage d in the crown, and the lower vegetation e are illustrated.

  The laser measurement noise a, the canopy transmission pulse b, the canopy contour unclear portion c, the long foliage d in the canopy, and the lower vegetation e will be described in detail with reference to FIGS.

  The laser measurement noise a is, for example, a case where the pixel value of the laser measurement data becomes zero or less due to a sensor measurement error or the like, and is indicated by a white dot in FIG.

  The canopy transmission pulse b is information generated by the laser beam passing through the canopy and reaching the ground surface through a space where the space existing in the canopy or a layer of branches and leaves is thin (for example, a place indicated by symbol b in FIGS. 4 and 5). In the crown height image shown in FIG. 4, it is indicated by black dots.

  The crown contour unclear portion c is formed by overlapping the crowns of a plurality of trees (for example, a portion indicated by reference sign c in FIGS. 4 and 5), and distinguishing the crown corresponding to each tree with an unclear outline as a whole. It is an area that is difficult to attach.

  The chief branches and leaves d are portions where branches protrude from the tree crown (for example, locations indicated by reference sign d in FIGS. 4 and 5). Such chief branches and leaves project in the horizontal direction in the crown height image shown in FIG. It is expressed as a local convex part.

  The lower vegetation e is a herb and a shrub (for example, a part indicated by a symbol e in FIGS. 4 and 5) existing below the investigation target tree.

  The laser measurement noise a, the canopy transmission pulse b, the canopy contour unclear portion c, the long foliage d, and the lower vegetation e described in FIGS. 4 and 5 are treated as noise. In the canopy high image data noise processing S40, Perform noise processing to remove noise.

  The canopy height image data noise process (step S40) is performed on the canopy height image data created in the canopy height image data creation process (step S30), and is a pre-process of the process for creating the precise canopy image data. Do. The crown height image data noise processing (step S40) will be described.

  FIG. 6 is a flowchart for explaining the crown height image data noise processing. The process surrounded by the broken line frame in FIG. 6 is the crown term image data noise process (step S40) in FIG.

  As shown in FIG. 6, first, laser measurement noise processing (step S41) is performed on the crown height image data created by the crown height image data creation processing (step S30), and the laser measurement noise processing is performed. The processed laser measurement noise image is displayed on the display.

  FIG. 7 is a diagram showing a laser measurement noise processing image displayed on the display. The laser measurement noise process is a process for correcting the laser measurement noise pixel having a negative value (pixel value equal to or less than zero) to zero. By performing such correction, the crown height image shown in FIG. The laser measurement noise present in the

  Subsequently, a lower vegetation process for creating lower vegetation processing image data from which the lower vegetation has been removed is performed on the laser measurement noise processed image data subjected to the laser measurement noise process (step S42). Then, a lower vegetation processing image corresponding to the lower vegetation processing image data subjected to the lower vegetation processing is displayed on the display.

  FIG. 8 is a diagram showing a lower vegetation processing image displayed on the display. In the lower vegetation processing, the maximum crown height among the crown heights of the lower vegetation is input, and the pixel value of the mesh having the crown height equal to or lower than the maximum crown height is set to zero. By performing such lower vegetation removal processing, lower vegetation processing image data from which the lower vegetation has been removed is created, and a lower vegetation processing image corresponding to the lower vegetation processing image data is displayed on the display. In FIG. 8, the area | region shown by the black which exists in the circumference | surroundings of each tree crown (area | region represented by gradation of white to gray) is an area | region from which the lower layer vegetation was removed. In FIG. 8, a large number of black spots existing in each canopy are the canopy transmission pulse b. At this stage, the processing for removing the canopy transmission pulse b has not been performed, so the canopy transmission pulse exists as noise. doing.

  Returning to FIG. 6, the averaging process (step S43) is subsequently performed on the lower vegetation processing image data. The averaging process (step S43) is performed by superimposing averaging filters on each mesh and setting the average crown height obtained by the averaging filter as the crown height of the mesh. To do.

  The averaging process will be described in a simplified manner. For example, if the averaging filter is a filter corresponding to an area of 5 meters on one side in a forest area, and if the size of one mesh is 1 meter on one side, An averaging filter is overlapped so that a mesh (referred to as a processing target mesh) is at the center, and a process in which the average value of the crown height of each mesh included in the averaging filter is set as the crown height of the processing target mesh, It carries out sequentially in the processing target mesh. For example, the crown heights of the respective meshes (for example, a total of 25 meshes of 5 mesh × 5 mesh) included in the range superimposed on the averaging filter (referred to as the averaging processing range) are integrated, A value obtained by dividing the total value by 25 is set as an average value of the averaging processing range, and the average value is set as a crown height of the processing target mesh. Such processing is performed on all meshes included in the crown height image data. Here, in order to simplify the description of the averaging process, the mesh size has been described with one side being 1 meter, but in the embodiment, as described above, the mesh size is from several centimeters per side. It is about 1 meter.

  The averaging filter is set based on the average canopy diameter of the survey target tree existing in the target forest area. For example, if the target trees in the target forest area are mainly larch, the averaging filter is about 7 meters on a side, and if the target tree in the target forest area is mainly red pines, the average The averaging filter is set to have a side of about 10 meters, and if the survey target tree existing in the survey target forest area is mainly cedar, the averaging filter is set to have a side of about 5 meters.

  FIG. 9 is a diagram showing an averaged image displayed on the display. Note that the averaged image shown in FIG. 9 is an averaged image obtained by averaging the lower vegetation processed image data created in the lower vegetation process (step S42).

  Although it is difficult to identify in FIG. 9, by performing the averaging process on the lower vegetation removal processing image data, the places where the original height is higher than the surroundings (for example, the apex portion of the tree crown and the length branches) It is affected by the lower part of the periphery of the apex and is expressed lower than the original height, and conversely, the part where the original height is lower than the surroundings (for example, a canopy transmission pulse) has a high crown around it. The image is expressed higher than the original height by being influenced by the part.

  Subsequently, a difference process (step S44) is performed. This difference process is a process of subtracting, in each mesh, the lower vegetation processing image data subjected to the lower vegetation process by the lower vegetation process (step S42) from the averaged image data averaged by the averaging process (step S43). For this reason, in the mesh of the canopy transmission pulse b originally having zero or a value close to zero, as described above, since the average value is a high value, it was obtained in the mesh of the canopy transmission pulse b. The difference value is larger in the positive direction than the difference value obtained in the tree crown portion. On the other hand, as described above, the average value of the mesh of the puppet branch d is low, so the difference value obtained in the mesh of the puppet branch d is the difference value obtained in the crown portion. It becomes larger in the negative direction compared to.

  FIG. 10 is a diagram showing a difference processed image displayed on the display. The difference processed image shown in FIG. 10 is an image corresponding to the difference processed image data. In FIG. 10, a portion represented by white (for example, a portion indicated by an arrow b) is a mesh of a canopy transmission pulse b, and a portion represented by black (for example, a portion indicated by an arrow d) is a prosthetic branch leaf d. The mesh. In addition, although the mesh of the canopy transmission pulse b and the mesh of the length branch leaves d are shown only in one place in FIG. 10, there are actually many.

  FIG. 11 is a diagram illustrating a histogram created based on the difference result obtained by the difference process. The histogram shown in FIG. 11 is created with the difference value of each mesh obtained by the difference processing as the horizontal axis, the number of meshes corresponding to the difference value as the vertical axis, and the difference value of zero (0) as an average value. It is a histogram. In this histogram, the difference value of each mesh obtained by the difference process is D, the standard deviation on the plus side of the average value is SD, and the standard deviation on the minus side of the average value (the minus standard deviation). ) To -SD. Here, the number of data (number of meshes) is 55,229, the maximum value of the difference value D is 24.96 (meters), the minimum value of the difference value D is -14.61 (meters), and the average value of the difference values Is zero, the standard deviation SD is 3.53, and the negative standard deviation −SD is −3.53.

  In the histogram in FIG. 11, the mesh that belongs between the standard deviation (SD) and the negative standard deviation (−SD) around the average value 0 can be regarded as a mesh corresponding to a tree crown, and the standard deviation ( A mesh that deviates to the plus side from SD) can be regarded as noise such as a canopy transmission pulse b, and a mesh that deviates to the minus side from minus standard deviation (−SD) can be regarded as noise such as a prosthetic branch d. it can.

  Subsequently, a canopy transmission pulse process (step S45) is performed. This crown transmission pulse process (step S45) is a process surrounded by a one-dot chain line frame in FIG. Note that the mesh to be processed at this time is the processing target mesh.

  In the canopy transmission pulse processing (step S45), the difference value D obtained in the processing target mesh is compared with the standard deviation SD, that is, whether or not the difference value D obtained in the processing target mesh satisfies D> SD. When the determination process (step S451) is performed and the difference value obtained in the processing target mesh is D> SD, the averaged image data corresponding to the processing target mesh is obtained as data of the processing target mesh. (Step S452). On the other hand, when the difference value D obtained in the processing target mesh is not D> SD (when D ≦ SD), the lower vegetation processing image data corresponding to the processing target mesh is used as the data of the processing target mesh. (Step S453).

Specifically, when the difference value D obtained in the processing target mesh is D> SD, the processing target mesh is assumed to be a mesh in which a canopy transmission pulse exists and corresponds to the processing target mesh. The averaged image data to be processed is set as data of the processing target mesh. That is, the averaged image data of the mesh corresponding to the processing target mesh determined as D> SD is used as the processing target mesh data. On the other hand, if the difference value obtained in the processing target mesh is not D> SD (D ≦ SD), it is determined that the difference is not a crown transmission pulse, and the lower vegetation processing image data corresponding to the processing target mesh is The processing target mesh data is used (step S453).
Then, the data obtained in step S452 and the data obtained in step S453 are combined to generate canopy transmission pulse processed image data (step S454).

  By performing the processing of step S451 to step S454 for all the meshes, the canopy transmission pulse processed image data subjected to the canopy transmission pulse processing can be generated. Then, the canopy transmission pulse processed image corresponding to the generated canopy transmission pulse processed image data is displayed on the display.

  FIG. 12 is a diagram showing a canopy transmission pulse processed image displayed on the display. The canopy transmission pulse processed image shown in FIG. 12 is an image obtained by removing the canopy transmission pulse b (the black dot portion existing in the canopy in FIG. 8) from the lower vegetation removal image shown in FIG.

  Subsequently, a crown contour process is performed to remove the length branches and leaves d and to improve the unclear portion c of the crown contour (step S46). This tree crown contour process (step S46) is a process surrounded by a two-dot chain line frame in FIG. In this case as well, the mesh to be processed at the present time is the processing target mesh.

  That is, the canopy contour process compares the difference value D of each mesh with the standard deviation SD in the same manner as the canopy transmission pulse process. In the canopy contour process, the difference value D obtained in the processing target mesh is: Processing for determining whether or not D <−SD is performed (step S461). Here, when the difference value D obtained in the processing target mesh is D <−SD, the crown transmission pulse processing image data corresponding to the processing target mesh is smoothed (step S462), Smoothed image data is created. On the other hand, when the difference value D obtained in the processing target mesh is not D <−SD (when D ≧ −SD), the canopy transmission pulse processing image data corresponding to the processing target mesh is used as the processing target. The data is mesh data (step S463).

  Specifically, when the difference value D obtained in the processing target mesh is D <−SD, the smoothing process is performed on the processing target mesh (step S453) on the assumption that the mesh has the length branch leaf d. )I do. The smoothing process performed here is performed using a smoothing filter capable of smoothing the irregularities of the long branches and crowns. As a result, the irregular branches and crowns and the irregularities of the crown are removed. And the smoothing process image corresponding to the smoothing process image data produced | generated in this way is displayed on a display.

  FIG. 13 is a diagram illustrating a smoothed image displayed on the display. By performing the smoothing process on the corresponding mesh of the canopy transmission pulse processed image data generated in step S454, as shown in FIG. 13, the lengthy branches and leaves that become noise are deleted, and the unevenness of the canopy and the like are removed.

On the other hand, as a result of performing the process of determining whether or not the difference value D obtained in the processing target mesh satisfies D <−SD (step S461), the difference value D obtained in the processing target mesh is When D <−SD is not satisfied (when D ≧ −SD), it is determined that the tree is not an elongated branch and leaves, and the canopy transmission pulse processed image data corresponding to the processing target mesh is used as data of the processing target mesh (step S463). .
Then, the data obtained in step S462 and the data obtained in step S463 are combined to generate canopy transmission pulse processed image data (step S464).

  By performing the processes in steps S461 to S464 for all meshes, it is possible to generate the crown contour processing image data subjected to the crown contour processing. Then, the crown contour processing image corresponding to the tree crown contour processing image data generated in this way is displayed on the display.

  FIG. 14 is a diagram showing a crown contour processing image displayed on the display. According to the crown contour processing image shown in FIG. 14, an image in which the crown transmission pulse has been removed and the length branches and leaves are also removed, and the crown contour unclear portion (for example, the crown contour unclear portion c in FIG. 4). ) Is emphasized. As a result, the crown contour processing image shown in FIG. 14 is rounded as a whole with the convex portions of each crown reduced compared to the crown transmission pulse processing image shown in FIG.

  Thus, the canopy high image data noise processing is completed. Subsequently, precise crown image data creation processing is performed (step S50). This precise canopy image data creation process is a process for creating precise canopy image data so that the canopy corresponding to each tree can be extracted. The precise crown image data creation processing (step S50) generates precise crown image data based on the crown contour processing image data generated by the crown height image data noise processing (step S40).

  The precise canopy image data creation processing (step S50) is performed by utilizing the fact that the edge of the crown is darker than the crown portion, and the individual tree crown extraction individual described in the known document 1 shown below. -Using a tree detection method (Individual tree detection method), precise tree crown image data is created, and a precise tree crown image corresponding to the precise tree crown image data is displayed on the display.

  Known Document 1: J Hyypp, M Inkinen, “Detecting and matching attributes for single trees using laser scanner”, The photogrammetric journal of Finland 16 (2), 1999, 27-42.

  FIG. 15 is a diagram showing a precise tree crown image displayed on the display. In FIG. 15, the area displayed in white is a tree crown. By integrating the number of pixels in each crown area in the precise crown image shown in FIG. 15, the crown area and crown diameter can be obtained. In FIG. 15, the area shown in black is a non-crown area, and when viewed from above, the non-crown area is a portion where there is no tree of an upper-layer crown, and a lower layer where there is no genital or tall tree It is a part (gap) covered with vegetation.

  In this way, when the precise canopy image data is created by performing the precise canopy image data creation processing (step S50), the precise canopy information creation processing (step S60) is performed based on the precise canopy image data. Precise information on the tree crown, that is, precise tree crown information is created. As the precise crown information created in the precise crown information creation process (step S60), the label number (serial number) of the crown corresponding to each tree, the crown position (center position of the crown) of the crown corresponding to each tree, and each tree Canopy diameter, the crown area of the canopy corresponding to each tree, etc. In addition to this, the canopy shape of the canopy corresponding to each tree can be obtained as precise canopy information.

  FIG. 16 is a diagram showing a label number assignment image in which a label number is assigned to each tree crown displayed on the display. As shown in FIG. 16, label numbers such as 1, 2,... Are assigned to each tree crown by automatic labeling. With this label number, the number of canopies recognized as tree canopies (the number of canopies) in the survey target forest area can be determined, whereby the number of survey target trees in the survey target forest area can be determined.

  Subsequently, forest resource information calculation processing (step S70) is performed. In the forest resource information calculation process, information on forest resources is calculated based on the precise canopy information created by the precise canopy information creation unit 60. Examples of the information on the forest resources calculated in the forest resource information calculation process (step S70) include the tree height of each tree, the breast height diameter of each tree, and the volume of each tree.

  In order to calculate information about these forest resources, in the forest resource information calculation process (step S70), a tree height calculation process (step S71) for calculating the tree height of each tree and a breast height diameter calculation for calculating the breast height diameter of each tree. A process (step S72), a timber accumulation determination process (step S73) for calculating the volume of each tree resource, and a forest resource information totaling process (step S74) for totaling forest resource information are performed.

  Here, the tree height calculation processing (step S71) for calculating the tree height of each tree is based on the precise crown image data created by the precise crown image data creation unit 50, and the maximum value of the crown height image data corresponding to each crown. Is the height of the tree.

  FIG. 17 is a diagram showing a tree height calculation image displayed on the display. In FIG. 17, the tree height obtained at each tree crown is shown in meters up to the first decimal place. As shown in FIG. 17, one tree height can be obtained for one tree (tree) by displaying the tree height obtained in each tree canopy.

  The breast height diameter calculation process (step S72) can calculate the breast height diameter by a multiple regression equation from at least one of the crown area and the crown diameter and the tree height. In addition, the breast height diameter of unmeasured trees can be calculated by interpolation. This makes it possible to calculate the breast height diameter of all the survey target trees existing in the survey target forest area. In addition, when calculating multiple regression equations, it is possible to determine multiple regression equation variables by selecting about ten standard trees in the field survey for each tree type and measuring the crown area, crown diameter and tree height of the selected standard tree. preferable.

  In the material integration determination process (step S73), the material volume is obtained from a two-variable material volume expression of the breast height diameter and the tree height. It should be noted that a quick reference table in which the volume is calculated from the breast height diameter and the tree height is known, and the tree resource (the volume) can be determined from such a quick reference table. The quick reference table is described in the following documents (Known Documents 2 and 3).

Known Document 2: Forestry Agency Planning Section, Tachiki Trunk Volume Table East Japan, Japan Forestry Research Committee, 334 pages, 2003

  The forest resource information totaling process (step S74) totals the resource information of each tree such as the number of trees to be investigated, the crown diameter, the crown area, the breast height diameter, the tree height and the volume of each tree, that is, single tree resource information. . Then, the aggregated single tree resource information is registered in the attribute database. Single tree resource information registered in the attribute database can be displayed on the display as a list. The single tree resource information displayed as a list includes XY coordinates that represent the center position of the tree crown. The number of survey target trees can be obtained based on the label number obtained by the precise canopy information creation process (step S60).

  FIG. 18 is a diagram showing a list of single tree resource information displayed on the display. As shown in FIG. 18, as the single tree resource information of the forest tree to be surveyed, the crown number (label number) corresponding to each tree, the XY coordinates representing the center position of each crown, the crown diameter, the crown area, the chest height diameter (In FIG. 18, “DBH” is indicated.) The tree height and the volume are displayed in association with the number (label number) of the tree crown.

  In addition, the forest resource information aggregation process (step S74) includes the number of survey target trees, the average value of the crown diameter calculated for each tree, the maximum value of the crown diameters calculated for each tree, and Minimum value, average value of crown area calculated for each tree, maximum and minimum values of crown area calculated for each tree, average value of chest height diameter calculated for each tree, Average breast height diameter calculated for each tree, maximum and minimum breast height diameters calculated for each tree, average tree height calculated for each tree, for each tree The maximum and minimum values of the calculated tree height, the average value of the volume calculated for each tree, the maximum and minimum values of the volume calculated for each tree, The total volume is also calculated and the forest volume Calculate the amount of forest resources per hectare (ha) used in business (here, the number of trees and the volume of wood), and information indicating the outline of the forest resources based on the calculated results (forest resource summary information) Is also registered in the attribute database. The forest resource summary information registered in the attribute database can be displayed on the display as a list.

  FIG. 19 is a diagram showing a list of forest resource summary information displayed on the display. As shown in FIG. 19, the number of trees to be investigated, the average value of the crown diameter, the maximum and minimum values of the crown diameter, the average value of the crown area, the maximum and minimum values of the crown area, the average value of the breast height diameter, the breast height The average diameter, maximum and minimum breast height diameter, average tree height, maximum and minimum tree height, average volume, maximum and minimum volume, and total volume are displayed. The amount of forest resources per hectare (ha) used in forest management (here, the number of trees and the volume of wood) is displayed. By displaying such forest resource summary information on the display, it is possible to grasp the summary of forest resources in the forest area to be investigated.

  In this specification, the single tree resource information shown in FIG. 18 and the forest resource summary information shown in FIG. 19 are combined as forest resource information. By combining such forest resource information and existing forest survey book data, it can be used for forest management such as grasping the current forest situation and planning forest operations such as thinning. For example, when trying to thin a forest, it is useful when making a simulation of a thinning plan.

  FIG. 20 is a diagram illustrating an example of a thinning plan simulation displayed on the display. As shown in FIG. 20, the arrangement and thinning ratio of thinning trees can be determined on a desk, and the amount of cutting can be calculated. Here, a layout diagram of line thinning is shown, and FIG. 20 shows a case where a tree whose crown is shown in gray is a thinned tree.

  As described above, according to the forest resource information calculation method according to the embodiment, the forest area image data to be surveyed based on the laser measurement data obtained by irradiating the region including the forest area to be surveyed from above with the laser beam. The canopy height image data that has been subjected to noise processing is created by applying the noise processing to remove the noise from the canopy height image data. And the precise crown image data is created based on the noise-processed crown height image data.

  The precise canopy image data created in this way becomes precise canopy image data that can extract each canopy corresponding to each tree to be investigated with high accuracy. Therefore, the precise canopy information created based on the precise canopy image data becomes highly accurate canopy information corresponding to each tree of the survey target tree, and based on the precise canopy information, information on forest resources (forest resource information ) Can be obtained highly accurate forest resource information (see, for example, FIGS. 18 and 19). As a result, it is possible to provide highly accurate forest resource information with little measurement error and high objectivity in the entire forest area to be surveyed, for each subgroup, and in any range, so it can be used for forest management such as forest management planning such as thinning. .

FIG. 21 is a diagram for explaining the forest resource information calculation apparatus 1 according to the embodiment. The forest resource information calculation device 1 according to the embodiment is a device for performing the processing of each step shown in FIGS. 1 and 6, and a data input unit 10 having a function for performing the processing of step S10 in FIG. 1. And the investigation target forest area image data creation processing unit 20 having the function for performing the process of step S20 in FIG. 1, and the canopy height image data creation processing unit 30 having the function for performing the process of step S30 in FIG. A canopy high image data noise processing unit 40 having a function for performing the process of step S40 in FIG. 1, and a precise canopy image data creation processing unit 50 having a function of performing the process of step S50 in FIG. a precision crown information creation processing unit 60 has a function of performing the processing of step S60 in FIG. 1, the process of step S70 in FIG. 1 A forest resource information calculation processing unit 70 having a function for, a.

  Further, the crown height image data noise processing unit 40 includes a laser measurement noise processing unit 41 that performs the process of step S41 in FIG. 1, and a lower vegetation processing unit 42 that has a function for performing the process of step S42 in FIG. The averaging processing unit 43 having a function for performing the processing of step S43 in FIG. 1, the difference processing unit 44 having a function of performing the processing of step S44 in FIG. 1, and the processing of step S45 in FIG. A canopy transmission pulse processing unit 45 having a function for performing and a canopy contour processing unit 46 having a function for performing the processing of step S46 in FIG. 1 are included.

Further, the forest resource information calculation unit 70 includes a tree height calculation processing unit 71 having a function for performing the process of step S71 in FIG. 1, and a breast height diameter calculation processing unit having a function for performing the process of step S72 in FIG. , A material integration fixed processing unit 73 having a function for performing the process of step S73 in FIG. 1 and a forest resource information totaling processing unit 74 having a function for performing the process of step S74 in FIG. 1 are included. .

  Thus, since each component of the forest resource information calculation apparatus 1 according to the embodiment shown in FIG. 21 has a function of performing each step of the forest resource information calculation method according to the above-described embodiment, The forest resource information calculation apparatus 1 according to the embodiment can obtain the same effects as those obtained by the forest resource information calculation method according to the above-described embodiment. In addition, the description about the content of the process of the forest resource information calculation apparatus 1 which concerns on embodiment is abbreviate | omitted.

  In addition, the forest resource information calculation apparatus 1 according to the embodiment has the functions of the respective components (see FIG. 2) included in the forest resource information calculation apparatus 1 installed as a computer program. By giving predetermined data to the constituent elements, the functions of the respective constituent elements are executed on the software of the computer.

  The present invention is not limited to the above-described embodiment, and various modifications can be made without departing from the spirit of the present invention. For example, the following modifications are possible.

  (1) In the above embodiment, for example, the case where the investigation target tree is a planted tree such as larch, red pine, cedar, cypress, etc. is exemplified, but the investigation target tree is not limited to the above-described planted tree. For example, in the case where there is a forest in which beech and oak are grouped within a predetermined area, the beech and oak can be used as a survey target tree.

  (2) In the above embodiment, the case of irradiating laser measurement data using a drone is exemplified, but the present invention is not limited to a drone, and for example, laser measurement data can be irradiated using an aircraft. is there.

1 ... forest resources information calculating device, 10 ... data inputting unit, 2 0 ... surveyed forest area image data creation processing section, 30 ... crown height image data creation processing section, 40 ... crown High image data noise processing unit, 41... Laser measurement noise processing unit, 42... Lower vegetation processing unit, 43... Average processing unit, 44. Processing part 46 ... Tree crown contour processing part 50 ... Precision crown image data creation processing part 60 ... Precision tree crown information creation processing part 70 ... Forest resource information calculation processing part 71 ... Tree height calculation processing unit, 72... Chest height diameter calculation processing unit, 73... Material integration set processing unit, 74 .. forest resource information totaling processing unit, a .. laser measurement noise, b. , C: crown contour unclear part, d: long branch leaf, e ... Understory vegetation, S10 ... Laser measurement data input process, S20 ... Investigation target forest area image data creation process, S30 ... Crown height image data creation process, S40 ... Crown height image data noise process S41 ... Laser measurement noise processing, S42 ... Underlayer vegetation processing, S43 ... Averaging processing, S44 ... Difference processing, S45 ... Crown transmission pulse processing, S46 ... Crown crown processing, S50 ... Precise canopy image data creation, S60 ... Precision canopy information creation processing, S70 ... Forest resource information calculation processing, S71 ... Tree height calculation processing, S72 ... Chest height diameter calculation processing, S73 ...・ Material integration determination processing, S74... Forest resource information aggregation processing, S454... D> SD determination processing, S461... D <-SD determination processing, S462.

Claims (26)

A forest resource information calculation method for creating forest resource information relating to the forest area to be surveyed based on laser measurement data obtained by irradiating laser light to an area including the forest area to be surveyed from above,
Based on the laser measurement data and geographical information, a survey target forest area image data creation processing step of creating survey target forest area image data;
Based on the survey target forest area image data, canopy height image data creation processing step for creating crown height image data representing the crown height of a tree existing in the survey target forest area;
Canopy high image data noise processing step of creating noise processed canopy high image data by performing noise processing for removing noise included in the canopy high image data;
Based on the noise-processed crown height image data, precise crown image data that can extract a crown corresponding to each tree of the survey target tree among the vegetation existing in the survey target forest area is created. Precision crown image data creation processing step,
Precise crown information creation processing step for creating precise crown information about the crown corresponding to each tree from the precise crown image data;
Forest resource information calculation processing step for creating forest resource information related to the forest area to be investigated based on the precise tree crown information;
Have
The noise included in the canopy height image data includes laser measurement noise included in the laser measurement data, lower vegetation existing below the investigation target tree, and the laser beam transmitted through the tree canopy and the ground surface. And a canopy transmission pulse generated by reaching the canopy, and a prominent branch leaf and a crown outline unclear portion protruding locally from the canopy,
The canopy high image data noise processing step performs the noise processing using mesh data of each mesh obtained by meshing the canopy high image data, and the canopy high image data noise processing step includes:
A laser measurement noise processing step for removing the laser measurement noise;
A lower vegetation processing step for creating lower vegetation processing image data obtained by removing the lower vegetation from the crown height image data,
Averaging processing image data by averaging the crown height of each tree using an averaging filter that is set based on an average canopy diameter of each tree for the lower vegetation processing image data An averaging process step to create
A difference processing step for taking a difference between the averaged image data and the lower vegetation image data;
Based on the result of the difference obtained by the difference processing step, canopy transmission pulse processing step of creating the crown transmission pulse processing image data from which the canopy transmission pulse has been removed,
Based on the result of the difference obtained by the difference processing step, the crown contour processing step for creating the crown contour processing image data in which the length branch leaves are removed and the crown contour unclear portion is improved;
Forest resource information calculation method characterized by including
In the forest resource information calculation method according to claim 1,
The canopy height image data creation processing step creates meshed digital surface model data and digital elevation model data based on the forest area image data to be investigated, and the digital surface layer model data and the digital elevation in each mesh A forest resource information calculation method, wherein the crown height image data is created by taking a difference from model data. In the forest resource information calculation method according to claim 2,
Forest resource information characterized in that the lower vegetation processing step inputs a maximum crown height among the crown heights of the lower vegetation and sets a pixel value of a mesh having a crown height less than or equal to the maximum crown height to zero. Calculation method. In the forest resource information calculation method according to claim 3,
The averaging processing step sets each mesh of the crown height image data as a processing target mesh, superimposes the averaging filter on the processing target mesh, and averages the crown height of each mesh included in the averaging filter. Forest resource information calculation method characterized by sequentially performing processing for setting the crown height of the processing target mesh on each processing target mesh. In the forest resource information calculation method according to claim 4,
In the difference processing step, each mesh is subjected to a difference process for obtaining a difference value between the averaged image data and the lower vegetation image data by subtracting lower vegetation image data from the averaged image data. Forest resource information calculation method. In the forest resource information calculation method according to claim 5,
In the canopy transmission pulse processing step, the difference value of each mesh obtained by the difference processing is set on the horizontal axis, the number of meshes corresponding to the difference value is set on the vertical axis, and the difference value is zero (0) as an average value. In the histogram, the difference value of each mesh obtained by the difference processing is set as D, the standard deviation on the plus side of the average value is set as SD, and the difference value on the minus side of the average value is set in the histogram. When the standard deviation is -SD and the mesh to be processed at the current time is the processing target mesh,
When the difference value D obtained in the processing target mesh is D> SD, the averaged image data corresponding to the processing target mesh is used as data of the processing target mesh, and is obtained in the processing target mesh. When the obtained difference value D is D ≦ SD, generating the canopy transmission pulse processed image data using the lower vegetation processed image data corresponding to the processing target mesh as data of the processing target mesh. A characteristic forest resource information calculation method. In the forest resource information calculation method according to claim 6,
When the difference value D obtained in the processing target mesh is D <−SD, the crown contour processing step smoothes the crown transmission pulse processing image data corresponding to the processing target mesh. Create a smoothed image,
When the difference value D obtained in the processing target mesh is D ≧ −SD, the crown contour processing is performed using the crown transmission pulse processing image data corresponding to the processing target mesh as data of the processing target mesh. Forest resource information calculation method characterized by generating image data. In the forest resource information calculation method according to claim 7,
The method for calculating forest resource information, wherein the precise canopy image data creation processing step creates the precise canopy image data using the fact that the edge of the canopy is darker than the crown portion. In the forest resource information calculation method according to claim 8,
The precise canopy information creation processing step uses the precise canopy image data, and at least the label number of the canopy corresponding to each tree, the canopy position of the canopy corresponding to each tree, and the canopy of the canopy corresponding to each tree A forest resource information calculation method comprising calculating a crown diameter and a crown area of a crown corresponding to each tree as the precise crown information. In the forest resource information calculation method according to claim 9,
The forest resource information calculation processing step includes:
Tree height calculation processing step for calculating the tree height of each tree,
Chest height diameter calculation processing step for calculating the diameter of the breast height of each person as the breast height diameter of each tree in each tree;
A material integration determination processing step for calculating the volume of each tree,
Forest resource information aggregation processing step for performing a process of aggregating the tree height of each tree, the breast height diameter of each tree and the volume of each tree, and the precise crown information calculated in the precise crown information creation processing step;
Forest resource information calculation method characterized by having In the forest resource information calculation method according to claim 10,
The forest height information calculation method according to claim 1, wherein the height of each tree is a maximum height within each canopy among the heights obtained by the canopy height image data creation processing step.
In the forest resource information calculation method according to claim 10 or 11,
The method of calculating forest resource information, wherein the breast height diameter of each tree is calculated by a multiple regression equation from at least one of the crown area and the crown diameter of the tree and the tree height of the tree. In the forest resource information calculation method according to any one of claims 10 to 12,
The forest resource information aggregation processing step in the forest resource information calculation processing step includes:
The number of trees to be investigated, the average value of the crown diameter calculated for each tree, the maximum and minimum values of the crown diameters calculated for each tree, and calculated for each tree The average value of the crown area, the maximum and minimum values of the crown area calculated for each tree, the average value of the breast height diameter calculated for each tree, and calculated for each tree The maximum and minimum values of the breast height diameters, the average value of the tree heights calculated for each tree, the maximum and minimum values of the tree heights calculated for each tree, And a processing for calculating an average value of the calculated volume, a maximum value and a minimum value of the calculated volume for each tree, and the number and volume per unit area of the survey target tree. Having further processing to calculate Forest resources information calculation method and butterflies. A forest resource information calculation device that creates forest resource information on the forest area to be surveyed based on laser measurement data obtained by irradiating laser light to an area including the forest area to be surveyed from above,
Based on the laser measurement data and geographical information, a survey target forest area image data creation processing unit that creates survey target forest area image data;
Based on the survey target forest area image data, a crown height image data creation processing unit that creates crown height image data representing a crown height of a tree existing in the survey target forest area;
A canopy high image data noise processing unit that performs noise processing for removing noise included in the canopy high image data and creates noise-processed canopy high image data;
Based on the noise-processed crown height image data, precise crown image data that can extract a crown corresponding to each tree of the survey target tree among the vegetation existing in the survey target forest area is created. Precision crown image data creation processing unit,
A precise canopy information creation processing unit for creating precise canopy information about the canopy corresponding to each tree from the precise canopy image data;
A forest resource information calculation processing unit that creates forest resource information related to the forest area to be investigated based on the precise tree crown information;
Have
The noise included in the canopy height image data includes laser measurement noise included in the laser measurement data, lower vegetation existing below the investigation target tree, and the laser beam transmitted through the tree canopy and the ground surface. And a canopy transmission pulse generated by reaching the canopy, and a prominent branch leaf and a crown outline unclear portion protruding locally from the canopy,
The canopy high image data noise processing unit performs the noise processing using mesh data of each mesh obtained by meshing the canopy high image data, and the canopy high image data noise processing unit includes:
A laser measurement noise processing unit for removing the laser measurement noise;
A lower vegetation processing unit for creating lower vegetation processing image data obtained by removing the lower vegetation from the crown height image data;
Averaging processing image data by averaging the crown height of each tree using an averaging filter that is set based on an average canopy diameter of each tree for the lower vegetation processing image data An averaging processing unit for creating
A difference processing unit that takes a difference between the averaged image data and the lower vegetation processed image data;
Based on the result of the difference obtained by the difference processing unit, canopy transmission pulse processing unit for creating the canopy transmission pulse processing image data from which the canopy transmission pulse has been removed,
Based on the result of the difference obtained by the difference processing unit, the crown contour processing unit that creates the crown contour processing image data that removes the length branches and improves the crown contour unclear part,
Forest resource information calculation device characterized in that is included.
In the forest resource information calculation device according to claim 14,
The crown height image data creation processing unit creates meshed digital surface model data and digital elevation model data based on the forest area image data to be surveyed, and the digital surface layer model data and the digital elevation in each mesh A forest resource information calculation apparatus, characterized in that the crown height image data is created by taking a difference from model data. In the forest resource information calculation apparatus according to claim 14,
The lower vegetation processing unit inputs a maximum crown height among the crown heights of the lower vegetation, and sets a pixel value of a mesh having a crown height equal to or lower than the maximum crown height to zero. Calculation device. In the forest resource information calculation apparatus according to claim 16,
The averaging processing unit sets each mesh of the crown height image data as a processing target mesh, superimposes the averaging filter on the processing target mesh, and averages the crown height of each mesh included in the averaging filter. A forest resource information calculation device that sequentially performs the processing for setting the crown height of the processing target mesh on each processing target mesh. In the forest resource information calculation device according to claim 17,
The difference processing unit performs difference processing for each mesh to obtain a difference value between the averaged image data and lower vegetation image data by subtracting lower vegetation image data from the averaged image data. Forest resource information calculation device.
The forest resource information calculation apparatus according to claim 18,
The crown transmission pulse processing unit uses the difference value of each mesh obtained by the difference processing as the horizontal axis, the number of meshes corresponding to the difference value as the vertical axis, and the difference value of zero (0) as an average value In the histogram, the difference value of each mesh obtained by the difference processing is set as D, the standard deviation on the plus side of the average value is set as SD, and the difference value on the minus side of the average value is set in the histogram. When the standard deviation is -SD and the mesh to be processed at the current time is the processing target mesh,
When the difference value D obtained in the processing target mesh is D> SD, the averaged image data corresponding to the processing target mesh is used as data of the processing target mesh, and is obtained in the processing target mesh. When the obtained difference value D is D ≦ SD, generating the canopy transmission pulse processed image data using the lower vegetation processed image data corresponding to the processing target mesh as data of the processing target mesh. A featured forest resource information calculation device. In the forest resource information calculation device according to claim 19,
When the difference value D obtained in the processing target mesh is D <−SD, the tree crown contour processing unit smoothes the crown transmission pulse processing image data corresponding to the processing target mesh. Create a smoothed image,
When the difference value D obtained in the processing target mesh is D ≧ −SD, the crown contour processing is performed using the crown transmission pulse processing image data corresponding to the processing target mesh as data of the processing target mesh. An apparatus for calculating forest resource information characterized by generating image data. In the forest resource information calculation device according to claim 20,
The forest resource information calculation device, wherein the precise canopy image data creation processing unit creates the precise canopy image data using the fact that the edge of the canopy is darker than the crown portion.
In the forest resource information calculation device according to claim 21,
The precise crown information creation processing unit uses the precise crown image data, and at least the label number of the crown corresponding to each tree, the crown position of the crown corresponding to each tree, the crown position corresponding to each tree A forest resource information calculation device that calculates a crown diameter and a crown area of a crown corresponding to each tree as the precise crown information.
In the forest resource information calculation device according to claim 22,
The forest resource information calculation processing unit
A tree height calculation processing unit for calculating the tree height of each tree;
A breast height diameter calculation processing unit for calculating a diameter of a breast height of a human in each tree as a breast height diameter of each tree;
A material integration fixed processing unit for calculating the volume of each tree;
A forest resource information totaling processing unit for performing processing for totalizing the tree height, the chest height diameter of each tree and the volume of each tree and the precise crown information calculated by the precise crown information creation processing unit;
A forest resource information calculation device characterized by comprising:
The forest resource information calculation apparatus according to claim 23,
The forest height information calculation apparatus according to claim 1, wherein the height of each tree is a maximum tree height in each crown among the crown heights obtained by the crown height image data creation processing unit.
In the forest resource information calculation device according to claim 23 or 24,
A forest resource information calculation apparatus, wherein the breast height diameter of each tree is calculated by a multiple regression equation from at least one of the crown area and the crown diameter of the tree and the tree height of the tree. In the forest resource information calculation apparatus in any one of Claims 23-25,
The forest resource information aggregation processing unit in the forest resource information calculation processing unit,
The number of trees to be investigated, the average value of the crown diameter calculated for each tree, the maximum and minimum values of the crown diameters calculated for each tree, and calculated for each tree The average value of the crown area, the maximum and minimum values of the crown area calculated for each tree, the average value of the breast height diameter calculated for each tree, and calculated for each tree The maximum and minimum values of the breast height diameters, the average value of the tree heights calculated for each tree, the maximum and minimum values of the tree heights calculated for each tree, And a processing for calculating an average value of the calculated volume, a maximum value and a minimum value of the calculated volume for each tree, and the number and volume per unit area of the survey target tree. Specialized in further processing to calculate Forest and resource information calculating device.