JP4152808B2 - Tree modeling program - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a tree modeling program .
[0002]
[Prior art]
The thing described in patent document 1 is known as what performs forest aspect analysis using the technique of aerial surveying. In this conventional example, the highest elevation point of the original laser data is selected as a tree representative point, and a tree extraction process is performed assuming that a tree trunk exists immediately below the tree representative point.
[0003]
[Patent Document 1]
Japanese Patent Laid-Open No. 11-23263
[Problems to be solved by the invention]
However, in the conventional example described above, there is a problem that it is necessary to know the range of the tree crown in advance when determining the tree representative point. That is, if the area for determining the highest altitude point that is a tree representative point does not accurately correspond to the crown, the highest altitude point of a plurality of adjacent trees is determined as one tree representative point, or There is a problem that a plurality of tree representative points are extracted from one tree and specified as one tree, and reliability is lowered.
[0005]
The present invention has been made to solve the above drawbacks, and an object of the present invention is to provide a tree modeling program capable of accurately modeling a tree .
[0006]
[Means for Solving the Problems]
According to the present invention, the object is
A contour line C is generated based on a stereo pair image obtained by imaging the tree group 1 from the sky or laser elevation data,
It is accomplished by a Turkey to extract independent tree 3 separated from trees group 1 as the modeling unit by the generated contour C.
[0007]
In order to model each tree from the tree group 1 including a plurality of trees, it is necessary to accurately extract the tree to be modeled from the tree group 1, and for that purpose, a stereo pair image obtained from the tree group 1, Alternatively, the contour line C is generated based on the laser elevation data (original data 2).
[0008]
The obtained contour line C is for the crown portion 3a of the tree, and the independent tree 3 is extracted by dividing it from the crown of another adjacent tree. In the present invention for identifying the independent tree 3 based on the contour line C, the determination of the independent tree 3 may be performed in consideration of the highest point (vertex) in the predetermined area and the arrangement of the contour line C surrounding the vertex. Therefore, it is possible to separate and extract the independent tree 3 from the tree group 1 with higher accuracy than in the conventional example in which the independent tree 3 is determined based only on the vertex information.
[0009]
For example, if the elevation difference between the contour lines C is sufficiently small, as shown in FIG. 2A, when the contour line C having three vertices T1, T2, T3 is obtained, the relationship between the vertices T1, T2 is If the area surrounded by the outermost contour line C1 including only the vertices T1 and T2 is equal to or less than the area assumed to be the area of the crown 3a of a normal tree, it is considered to belong to the same tree, and the contour line C1 is surrounded. If the area is too large for the same tree, it can be estimated as a separate tree.
[0010]
Similarly, when the area surrounded by the contour line C2 including the vertices T1, T2, and T3 is about the area of the crown of a normal tree, it is determined that these are the same tree with the top branching into three branches. it can.
[0011]
The generation of the contour line C and the separation of the independent tree 3 can be performed manually or by visual confirmation. However, it is possible to automate the whole process or a part of it and to automate the whole process. Is
A stereo pair image obtained by imaging the tree group 1 from above, or a contour line generating means 4 for generating a contour line C from the laser elevation data 2;
Based on the contour line C, the independent tree separating means 5 for separating the independent tree 3 using the vertex information and the area surrounded by the contour line C including the vertex.
You can use a tree modeling program to make the computer function as
[0012]
In this case, it is possible to adopt a well-known method for the contour line generating means 4, and for example, a method of directly connecting the altitude measurement points while performing curve interpolation such as spline interpolation or bezier interpolation can be used.
[0013]
Moreover, the program for functioning as the independent tree isolation | separation means 5 can be obtained by describing the determination method of the independent tree 3 mentioned above as a procedure, for example.
[0014]
As described above, according to the present invention, since a single tree can be accurately separated and extracted from a dense tree, the number of trees in a forest or the like can be accurately measured.
[0015]
Further, since the contour line C is data indicating the shape itself, it can be displayed as a three-dimensional shape that can be immediately viewed without requiring other additional data, as will be described later. For the purpose of estimation or the like, it can be processed as a statistic having shape characteristics without going through a visible state.
[0016]
In this case, even from only the planar arrangement of the contour line C, for example, the tree having the contour line C shown in FIG. 2B and the tree having the contour line C shown in FIG. Although the degree of tree species distinction is possible so that it can be easily known, it is possible to obtain more information by performing three-dimensional modeling of the crown portion 3a.
[0017]
That is, by performing three-dimensional modeling of the tree crown part 3a, it becomes possible to accumulate three-dimensional shape elements such as the volume of the tree crown part 3a as statistics. For example, the volume and area of the tree crown part 3a due to tree species Alternatively, if a table in which geometric characteristic values such as height are associated with the thickness of the trunk portion 3b is prepared, information about the trunk portion 3b of the specific tree, such as volume, can be immediately derived. .
[0018]
It should be noted that the contour line C and the trunk portion 3b can be associated with the crown portion 3a by paying attention to various elements on condition that a one-to-one correspondence can be statistically obtained.
[0019]
The output by the three-dimensional modeling of the tree crown portion 3a is not necessarily a three-dimensional display image. As described above, when estimating the volume of the trunk portion 3b from the shape characteristics of the tree crown portion 3a, the three-dimensional coordinates are used. It is sufficient if the value of interest in the system can be derived.
[0020]
As described above, according to the present invention, it is possible to accurately grasp the trunk portion 3b by giving the contour line C to the obtained crown portion 3a. In addition to this, it can be used to display a tree three-dimensionally.
[0021]
In this case, a contour pair generating means 4 for generating a contour line C of the tree crown portion 3a based on the stereo pair image obtained by imaging the tree from the sky or the laser elevation data 2,
Crown modeling means 6 for modeling the shape of the crown portion 3a based on the contour line generating means 4;
A tree modeling program is available to allow the computer to function. Further, in order to obtain the shape of the whole tree, a program in which a function as the trunk partial estimation means 7 is added to the above-described modeling program can be used.
[0022]
For the purpose of three-dimensional display of the tree, since the exact thickness of the trunk portion 3b is not necessarily required, it is sufficient for the trunk portion estimation means 7 to be able to depict the position, shape, etc. of the trunk portion 3b. For modeling the trunk portion 3b, after determining the trunk position with respect to the crown portion 3a, the trunk portion 3b is linearly extended from the trunk position to the ground surface, or representative points such as centroids at each contour surface Cf are sequentially connected. Is done.
[0023]
For the determination of the representative point on each contour plane Cf, MAT (Medial Axis Transform) which is an algorithm for obtaining a skeleton having a predetermined planar shape can be applied.
[0024]
DETAILED DESCRIPTION OF THE INVENTION
FIG. 1 shows an embodiment of the present invention. This embodiment is configured to analyze the number of trees to be vegetated in the forest, the volume, the tree species, and the like, and is based on the contour line generating means 4 for generating the contour line C from the original data 2 and the contour line C. Independent tree separating means 5 for separating the independent tree 3 from the densely packed part of the tree, a crown modeling means 6 for obtaining the shape of the trunk part 3b of the separated independent tree 3, and a trunk part estimating means for estimating the trunk part 3b from the crown shape 7.
[0025]
Each of these means is realized by running a predetermined computer program on the computer.
[0026]
The original data 2 is acquired from a flying object such as an aircraft flying over the target area, and a stereo pair image that can be viewed stereoscopically or laser elevation data obtained by laser sweeping the target area from the sky is used. .
[0027]
Contour lines C are extracted from the original data 2 by the contour line generating means 4. As shown in FIG. 3, the contour line C is subjected to smoothing processing between the contour surfaces Cf adjacent in the vertical direction (height direction) in addition to the smoothing processing by the Bezier interpolation method in the contour surface Cf. Optimization is achieved.
[0028]
Based on the contour line C obtained as described above, the tree to be processed is separated and extracted from the tree group 1 by the independent tree separating means 5. In the case where the contour line C extracted in a forest or the like where trees are densely includes a plurality of vertices T in a region partitioned by the same contour line C as shown in FIG. As described above, the separating unit 5 identifies the unity of the tree based on the area of the contour line C including the vertex T.
[0029]
When the above processing in the contour line generating means 4 and the independent tree separating means 5 is completed, for the number of trees and each independent tree 3, the vertex height of the tree, the radius of the crown, and the height of the crown are acquired as statistics, The tree height is acquired by referring to the altitude data of the ground surface and accumulated in the tree database.
[0030]
Next, each tree is modeled by the crown modeling means 6 and the trunk partial estimation means 7 and displayed three-dimensionally. The tree crown modeling means 6 models the surface layer of the tree crown part 3a with a NURBS (Non-Uniform Rational B-Spline) curved surface to give a three-dimensional shape of the tree crown part 3a.
[0031]
On the other hand, the trunk portion estimation means 7 estimates the thickness and trunk position of the trunk portion 3b, and the thickness of the trunk portion 3b is statistically determined from the radius or volume of the crown portion 3a. On the other hand, as the tree trunk position, the median value of the plane position of the representative point such as the centroid of each contour plane Cf is used, and the tree trunk portion 3b is a plane position given by the median value as shown in FIG. It is shown by a rod-like body that stands upright and reaches the upper end of the crown portion 3a.
[0032]
In the above description, for the sake of convenience of explanation, a case has been described in which a three-dimensional shape is obtained by modeling an independent tree 3 extracted and separated from a dense part of a tree such as a forest. In the case of a tree, it can be configured without the independent tree separating means 5.
[0033]
【The invention's effect】
As is clear from the above description, according to the present invention, a tree can be accurately modeled.
[Brief description of the drawings]
FIG. 1 is a block diagram illustrating the present invention.
FIGS. 2A and 2B are explanatory diagrams of the operation of the present invention, in which FIG. 2A shows the operation of the independent tree separating means, and FIGS. 2B and 2C show the identification state of tree species.
FIG. 3 is a diagram showing contour lines of a crown portion of an independent tree, where (a) shows a state before smoothing, and (b) shows a state after smoothing processing.
FIG. 4 is a diagram illustrating a state in which a tree is three-dimensionally modeled.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 1 Tree group 2 Original data 3 Independent tree 3a Crown part 3b Trunk part 4 Contour line generation means 5 Independent tree separation means 6 Crown modeling means 7 Trunk part estimation means C Contour line

Claims (3)

A stereo pair image obtained by capturing a group of trees from the sky, or a contour line generating means for generating contour lines from laser elevation data,
Based on the contour lines, independent tree separating means for separating independent trees using vertex information and the area surrounded by the contour line including the vertexes
Tree modeling program to make the computer function as. A stereo pair image obtained by imaging a tree group from the sky, or a contour line generating means for generating a contour line of a crown portion of the tree group based on laser elevation data,
Independent tree separation means for separating independent trees using vertex information and the area surrounded by the contour line including the vertex based on the contour line,
Crown modeling means for modeling the shape of the crown portion of an independent tree based on the contour line
Tree modeling program to make the computer function as. A stereo pair image obtained by imaging a tree group from the sky, or a contour line generating means for generating a contour line of a crown portion of the tree group based on laser elevation data,
Independent tree separation means for separating independent trees based on the contour lines using vertex information and the area enclosed by the contour lines including the vertices;
Crown modeling means for modeling the shape of the crown portion of an independent tree based on the contour lines;
Trunk partial estimation means to estimate the morphological elements of the trunk part from the geometric characteristics of the crown model output from the canopy modeling means
Tree modeling program to make the computer function as.

Images