JP4177634B2 - Laser data edge matching method - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to an edge matching method for laser data.
[0002]
[Prior art]
In recent years, attention has been paid to the fact that the position / elevation data of the ground surface can be directly acquired from the sky, and aviation laser surveying has been frequently used. This aerial laser survey is performed by sweeping a laser from the flying object along the predetermined flight course toward the ground, and when the survey target area is so wide that it is difficult to scan by one flight. The surveying target area is divided into a plurality of parts, and laser data acquisition work is performed.
[0003]
In this case, the laser data is acquired for each of the flight courses overlapped to some extent in order to prevent data leakage, but the data density in the overlap region is matched with the data density in other non-polymerization regions. In order to achieve this, it is necessary to adopt the one acquired in either one of the flight courses.
[0004]
Conventionally, this laser data selection (edge match) is performed by the following method. FIG. 13 shows the laser data (flight course data 1) processed by this conventional example. These two courses are adjacent and do not overlap each other (referred to as the first and third courses from the left). The second course in the center overlapping both the first and third courses is shown. In order to facilitate understanding of the drawing, the flight course data 1-2 of the second course is shown by hatching in FIG.
[0005]
When performing edge matching, first, as shown in FIGS. 13B and 13C, after obtaining the overlapping area, as shown in FIGS. 13D and 13E, the center line 5 of each overlapping area is obtained. Is calculated. FIG. 13B shows the overlapping area of the first course and the second course, and FIG. 13C shows the overlapping area of the second course and the third course, respectively.
[0006]
After obtaining the center line 5 in this way, as shown in FIGS. 13 (f), 13 (g), and 13 (h), the overlapping area is divided into two with the center line 5 as a boundary, and the laser in each divided area 3. As the data, the adjacent flight course data 1 is selected and used. In FIG. 13 (f), the overlapping area of the first flight course data 1-1 and the second flight course data 1-2 is divided by the center line, and the left side of the divided area 3, that is, adjacent to the first course. The state where the flight course data 1-1 of the first course is adopted is shown in the part. Similarly, FIG. 13 (h) shows the distribution of the adopted data in the second course and the third course, and FIG. 13 (g) shows a state in which the laser data of the second course is used in the remaining area 3.
[0007]
However, the conventional example described above has the following drawbacks. That is, in general, laser data in aerial surveys has a long optical path length of the measurement light, so the decrease in data accuracy due to an increase in swing angle is larger than when measuring close objects, and the data acquired from directly under the flying object The accuracy is the best, and the accuracy tends to decrease as going to the edge in the width direction.
[0008]
On the other hand, in the above conventional example, the overlapping area is simply divided into two and its attribution is determined. For example, when the course width is wide or narrow, the laser approaches the center of the course and has high accuracy. In some cases, data is discarded and laser data away from the center of the course is selected, and consideration is given to data accuracy, that is, quality superiority or inferiority of data.
[0009]
In addition, the method in the above conventional example is effective when the overlapping area extends along a flight course in a certain direction, but the flight course such as a river / road is not in a certain direction and overlaps the intersecting areas. In the case of the area, there is a problem that it cannot be applied and lacks versatility.
[0010]
Further, as a similar technique, Patent Document 1 discloses a technique in which images obtained by division are connected and combined in the vicinity of a boundary portion. In this conventional example, two acquired images in an overlapping region are disclosed. Since there is no qualitative superiority or inferiority, there is a problem that the quality evaluation is not considered and the data cannot be made uniform.
[0011]
[Patent Document 1]
Japanese Patent Laid-Open No. 11-155142
[Problems to be solved by the invention]
The present invention has been made in order to eliminate the above-mentioned drawbacks, and ensures the uniformity of data, and the edge matching of laser data with improved versatility by eliminating the restriction by the flight course when acquiring the laser data. The purpose is to provide a method.
[0013]
[Means for Solving the Problems]
In the field of aerial surveying, laser data acquisition is to acquire the height (elevation) and position data of a feature while sweeping the laser in a direction orthogonal to the flight direction of the aircraft. The reliability of the data is the highest, and as the laser swing angle increases, that is, the reliability decreases as the distance to the course edge increases.
[0014]
[Means for Solving the Problems]
As described above, the present invention mainly determines measurement data having uniform accuracy over the entire range of the measurement region to determine which overlap region data of the flight course data 1 is adopted in the overlapping acquisition region of the laser data. It was made paying attention to the point that it is very important above, and the reliability of laser data increases as it approaches the aircraft.
Overlap area data acquired from the same area by multiple flight course data acquired by an aircraft laser scanner that swings the laser in a direction orthogonal to the flight direction from the flying object on the predetermined flight course toward the ground Including an edge matching method of laser data to be used as laser data in an overlapping region by selecting any of the plurality of overlapping region data,
The programmed computer system uses the distance from the target point to the center 2 in the width direction defined as a continuous curve or straight line of each flight course as the laser data of the arbitrary target point in the overlap region. Is configured as a laser data edge matching method for performing processing for selecting overlap region data of the flight course to which the center 2 in the width direction belongs.
[0015]
Therefore, in the present invention, since the selection of the flight course data 1 in the overlapping region is selected based on the distance from the center 2 in the width direction, the higher accuracy is preferentially used, and the data is improved while improving the accuracy. Can be ensured.
[0016]
Further, even if the flight courses overlap in a crossing manner, the flight course data 1 can be selected, so that versatility is improved.
[0017]
The center 2 in the width direction of the flight course can be appropriately defined in consideration of the calculation speed and the required accuracy. For example, the flight course position data at the time of flight course data acquisition or zigzag is acquired. A linear or multi-order regression curve from the center point of each laser scanning line 4 between the scanning direction change points of the flight course data 1 can be used.
[0018]
Further, in addition to defining the center 2 in the width direction as a continuous curve (straight line), it is possible to define it as a center point sequence of each laser scanning line 4 between the scanning direction change points of the flight course data 1. In this way, by defining the center 2 in the width direction as a set of discrete values based on the actual measurement values, it is possible to eliminate the statistical error associated with the regression processing, so that the selection accuracy of the overlap region data is higher. Become.
[0019]
DETAILED DESCRIPTION OF THE INVENTION
FIG. 1 shows a flowchart of a laser data edge matching method. In order to facilitate understanding, the processing procedure of the present invention will be described below with reference to the drawings. However, in actuality, processing is performed only by calculation on laser data without requiring illustration as illustrated. . Further, the following method can be automatically processed by a computer system in which a program in which the following procedure is described operates.
[0020]
FIG. 2 shows a case where laser data is acquired along three parallel flight courses, as in the conventional example described above. For convenience of explanation, in the following, when the first, second, and third flight course data 1-1 to 1-3 are sequentially shown from the left in the figure, the first and third flight course data 1-1 and 1-3 are mutually connected. There is no overlap, and the second flight course data 1-2 overlaps the first and third courses with a predetermined width. In order to clarify the overlapping portion, the first to third flight course data 1-1 to 1-3 are sequentially hatched in FIGS. 2B to 2D. In the present specification, for example, a course such as 1-1 for the flight course data 1 corresponding to the first course, 1-2 for the flight course data 1 corresponding to the second course, and so on. The number related elements are numbered with the course number.
[0021]
When performing such edge matching processing of laser data, first, a center point sequence (width direction center 2) of each flight course data 1 is calculated in step S1. Since the laser data is obtained by sweeping laser light in a direction orthogonal to the flight direction indicated by the arrow in FIG. 3A, on the ground position plane, as shown in FIG. It is arranged in a zigzag pattern with respect to the direction, and each point is acquired as a series of data having height information (elevation information). The number given to each measurement point in FIG. 3A indicates the data acquisition order.
[0022]
Further, as shown in FIG. 3B, flag data indicating the laser scanning direction, more precisely, the direction when the reflected light is acquired, is added to each acquired laser data. In the example shown in FIG. 3B, for the laser data obtained by laser scanning to the right with respect to the flight direction, the flag “0” is set, and the laser data obtained by laser scanning to the left is obtained. Is assigned a flag “1”.
[0023]
In this embodiment, the calculation of the center point sequence 2 is performed using the flag data. First, the laser data is tracked in the order of data acquisition, and the flag changes on each tracking direction line (laser scanning line 4). Extract points. FIG. 3B shows the change point of the flag with a large black circle.
[0024]
Next, the center points on the ground position plane of the two flag change points extracted on each tracking direction line 4 are calculated, and these are arranged on the flight direction line to obtain the center point sequence 2. The center point obtained in FIG. 3C is indicated by a white circle. Further, FIG. 4A shows a state in which the center point sequences 2 for all the courses obtained as described above are connected by a straight line, and FIG. 4B shows the second flight course in which the center point sequence 2 is obtained. Data 1-2 is shown. For easy understanding of the arrangement of the center points 2, the center point row 2 is shown as being connected by a straight line in FIG.
[0025]
After the course center point sequence 2 is extracted as described above, the scan start line and the end line are extracted as the course start / end line 6 (S2). FIG. 4C shows the course start / end line 6 of all courses, and FIG. 4D shows the course start / end line 6 of the second course. In FIG. 1, the course start / end line 6 is extracted after the course center point sequence 2 is extracted. However, this is reversed, and the course start / end line 6 is extracted prior to the course center point sequence 2 extraction. You can also.
[0026]
Next, area division is performed in step S3. Prior to area division, a mesh is first set in the edge match processing area. FIG. 5A shows an enlarged view near the lower side of FIG. 2, and FIG. 5B shows a state where a mesh is set.
[0027]
After setting the mesh as described above, the attribution of each cell in the mesh to the course is determined by the following method. The cell attribution determination is performed by a procedure of concentrically searching the course center point sequence 2 from the cell of interest, and determining the course to which the course center point sequence 2 searched first is the belonging course of the cell of interest. .
[0028]
For example, assuming that the cell A is the target cell in FIG. 5C, when a search circle that gradually increases the size of the diameter from the center of the cell A is drawn, the center point row 2-2 of the second course is first on the search circle 7A. Therefore, the cell A is determined to belong to the second course. As shown in the cell B, when the course start / end line 6-2 is detected on the search circle 7B-0 and the course start / end line 6-2 is the scan end position, the center of the cell B is determined. The course attribute “2” is not given below the line segment 8 that connects the end of the course start / end line and the course start / end line 6-2, and in the example shown in FIG. The center point sequence 2-2 of the second course is detected first, but this is ignored, and then the cell is detected due to the presence of the center point sequence 2-3 of the third course detected on the search circle 7B-3. A course attribute of the third course is assigned to B. FIG. 6A shows the result of the area processing, and FIG. 6B shows a state in which the flight course data 1-2 of the second course is superimposed on the result. In FIG. 6A, the assigned course attributes are indicated by Roman numerals. For example, an area where “I” is entered belongs to the first course.
Next, in step S4, the flight course data 1 is selected based on the result of area division obtained as described above. The selection of the flight course data 1 is performed by clipping the flight course data 1 with the divided area 3 to which the course attribute of the same course is assigned. FIGS. 6C to 6E show the processing results, that is, the adoption ranges of the first to third flight course data 1 with hatching.
[0029]
In the above, the region division is performed by searching the belonging course for all the cells in the overlapping range. However, as shown in FIG. 7, the region division may be performed by widening the center point row 2 at a predetermined ratio. it can. FIG. 7A shows the middle of the widening operation, and shows a state in which the widening boundary line 9 is widened in the width direction at an equal ratio. In this state, the course attribute “2” is given to the region sandwiched by the central widening boundary line 9-2, and the region division in the overlapping region is completed by repeating this widening operation.
[0030]
In this method, it is also possible to select the flight course data 1 by weighting between flight courses. FIG. 7B shows a process in which the second course is weighted. First, the widening process is preferentially performed on the second course, and the attribute “2” is given to the area. Thereafter, when the widening boundary line 9 of another course, for example, the widening boundary line 9-1 of the first course in FIG. It is not adopted as the boundary line of the region 3.
[0031]
In this way, by performing the weighting of the region division by changing the number of widening operations or the widening ratio at the time of the widening operation, the flight course data 1 close to the center point sequence 2 can be preferentially adopted.
[0032]
FIG. 8 shows another application example. In this example, the flight course data 1 is acquired in three courses that are substantially parallel and two courses that intersect these courses. In this application example, for convenience of explanation, the three flight course data 1 running in the vertical direction in FIG. 8A are referred to as the first, second, and third courses from the left, and the two courses intersecting these from the bottom Called the 4th and 5th course. In order to facilitate understanding of the shape of each course, FIG. 8B to FIG. 8F show flight course data 1 in which each course is hatched. FIG. 8 shows only the vicinity of one side of the course start / end line 6 in an enlarged manner.
[0033]
When performing edge matching of these five courses, the center point sequence 2 and the course start / end line 6 of each course are extracted as described above. FIG. 9A shows the center point sequence 2 extracted, and FIG. 9B shows the center point sequence 2-2 in the second course. FIG. 9C shows the course start / end line 6, and FIG. 9D shows the start / end line 6-2 in the second course.
[0034]
Thereafter, region division is performed. FIG. 10A shows a state after the area division in this example, and Roman numerals in the figure indicate course attributes given to the area.
[0035]
As shown in FIG. 10A, the laser data plane is divided into a plurality of areas by a plurality of area dividing lines as a result of the area division. Then, as shown in FIG. Each flight course data 1 is overlaid on top of each other, and only the data in the area where the course attributes match is selected as the adopted data. FIG. 10B shows a state where the second flight course data 1-2 is superimposed on the area division data, and FIG. 10C shows the area attribute “II” from the flight course data 1-2 of the second course. Only the divided area 3-3 that is "" is selected as adopted data.
[0036]
FIG. 11 and below show an embodiment in which weighting is given to each flight course data 1 for the processing object shown in FIG. In this embodiment, weighting is given to the first to third flight course data 1-1 to 1-3, and the weighted course data is used when the center point sequence 2 and the course start / end line 6 are extracted. Is preferentially adopted. FIG. 11A shows unweighted flight course data 1, that is, the center point sequences 2-4 and 2-5 of the fourth and fifth flight course data 1-4 and 1-5 are weighted flight course data. 1, that is, it is not defined after crossing the central point sequence 2-1 to 2-3 of the first to third flight course data 1, and FIG. 11 (b) is accompanied by the fourth and fifth flights. The course start lines 6-5 and 6-6 of the course data 1 are not defined.
[0037]
Thus, after the weighted flight course data 1 is prioritized and the center point sequence 2 and the course start / end line 6 are obtained, as shown in FIG. 10, the region division is performed as shown in FIG. As is clear from comparison with FIG. 10 (a), the course attributes are given centering on the weighted course, and thereafter, each flight course data 1 is represented by area division data. Clip and finish the process.
[0038]
【The invention's effect】
As is apparent from the above description, according to the present invention, data uniformity can be ensured, and versatility can be enhanced by eliminating the restriction due to the flight course at the time of laser data acquisition.
[Brief description of the drawings]
FIG. 1 is a flowchart illustrating the present invention.
FIG. 2 is a diagram showing flight course data to be processed.
FIG. 3 is a diagram illustrating an extraction step of a center point sequence.
FIG. 4 is an explanatory diagram showing a state where a center point sequence extraction and course start / end line extraction step is completed;
FIG. 5 is an explanatory diagram showing area division.
FIG. 6 is an explanatory diagram from the end of region division to the end of edge match processing.
7 is a diagram showing a modification of FIG.
FIG. 8 is a diagram illustrating another application example of the present invention.
FIG. 9 is an explanatory diagram showing a center point sequence extraction and course start / end line extraction step completion state for the processing target of FIG. 8;
FIG. 10 is an explanatory diagram from the end of region division to the end of edge match processing;
FIG. 11 is a diagram showing another embodiment of the present invention.
FIG. 12 is a diagram illustrating an output state by edge match processing.
FIG. 13 is a diagram showing a conventional example.
[Explanation of symbols]
1 Flight course data 2 Center in width direction 3 Divided area 4 Laser scanning line

Claims (7)

Overlap area data acquired from the same area by multiple flight course data acquired by an aircraft laser scanner that swings the laser in a direction orthogonal to the flight direction from the flying object on the predetermined flight course toward the ground Including an edge matching method of laser data to be used as laser data in an overlapping region by selecting any of the plurality of overlapping region data,
The programmed computer system uses the distance from the target point to the center in the width direction defined as a continuous curve or straight line of each flight course as the laser data of an arbitrary target point in the overlap region. An edge matching method for laser data, in which an overlap region data of a flight course to which a minimum center in the width direction belongs is selected. Overlap area data acquired from the same area by multiple flight course data acquired by an aircraft laser scanner that swings the laser in a direction orthogonal to the flight direction from the flying object on the predetermined flight course toward the ground Including an edge matching method of laser data to be used as laser data in an overlapping region by selecting any of the plurality of overlapping region data,
A programmed computer system
After obtaining the width direction center defined as a continuous curve or a straight line for each flight course,
The overlapping region is divided into regions while giving course attributes of flight course data to which the center in the width direction to which an arbitrary target point is closest belongs for all points in the overlapping region,
An edge matching method for laser data, wherein processing is performed for selecting overlapping area data of flight course data matching a course attribute assigned to the divided area as laser data of the divided area. The laser data edge matching method according to claim 2, wherein the plurality of flight courses including the overlapping area are given weights for determining priority in the area dividing operation. Overlap area data acquired from the same area by multiple flight course data acquired by an aircraft laser scanner that swings the laser in a direction orthogonal to the flight direction from the flying object on the predetermined flight course toward the ground Including an edge matching method of laser data to be used as laser data in an overlapping region by selecting any of the plurality of overlapping region data,
A programmed computer system
After obtaining the width direction center defined as a continuous curve or a straight line for each flight course,
By repeating to interfere with widening area in the other flight course adjacent the widening operation for widening the width direction on both sides by a predetermined ratio a region from the width direction center around the widthwise center for each flight course data, Giving the course attribute of the flight course data to which each center in the width direction belongs to each widening operation target region, and dividing the overlapping region into regions,
An edge matching method for laser data, wherein processing is performed for selecting overlapping area data of flight course data matching a course attribute assigned to the divided area as laser data of the divided area. 5. The laser data edge matching method according to claim 4, wherein weighting of the flight course data is performed by varying the widening ratio in the area division. The center of the flight course in the width direction is the tracking of the flight course laser data in the order of data acquisition, and two flag change points on each tracking direction line extracted from the flag data indicating the laser scan direction added to each laser data. 6. An edge matching method for laser data according to claim 1 , wherein a center point on the ground position plane is calculated and obtained as a center point sequence obtained by arranging the center points in a flight direction line. . 7. The laser data edge matching method according to claim 1, wherein a mesh is set in the overlap region, and the overlap region data is selected in units of cells in the mesh.

Images