JP2014163707A - Road deformation detection device, road deformation detection method and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To realize automated work of detecting unevenness on the basis of three-dimensional coordinate data for a point group extracted from a structure surface along a road.SOLUTION: Point group extraction means 20 sets a reference line passing through an unevenness inspection position on an even reference surface assuming a structure surface along a road, and extracts first and second attention point groups in a point group that are located along the reference line within a preset nearby range, from each of one side and the other side of the reference line divided at the unevenness inspection position. Unevenness determination means 22 uses a prescribed significant level to apply an F-test to a hypothesis that regression coefficients of both attention point groups coincide with each other, with respect to a regression analysis with setting a coordinate along the reference line at each point constituting the attention point groups and a coordinate along a normal direction to the reference surface as an explanatory variable and a dependent variable respectively, and when the hypothesis is rejected, determines the presence of unevenness corresponding to the unevenness inspection position.

Description

  The present invention relates to a technique for detecting an uneven state (non-landscape) as a deformation of a road surface and a surface of a structure provided along a road.

  Maintenance of roads is important for ensuring the safety of road users, and road surface property surveys that detect road surface irregularities are being conducted as part of the management. In the conventional road surface property survey, road surface cracks, rutting, and longitudinal unevenness data are measured, and maintenance indexes are calculated based on these data to evaluate road serviceability.

  Conventionally, as a measuring device for detecting the unevenness of the road surface, a laser scanner is used to irradiate the road surface with light, and the reflected light from the road surface is measured to determine the distance between the device and the road surface, thereby measuring the unevenness of the road surface. There was a thing (patent documents 1, 2). The measuring device is mounted on a vehicle, and a road surface profile can be obtained by measuring the distance between the device and the road surface while the vehicle is running.

  On the other hand, a mobile mapping system has recently been developed as a technique for acquiring three-dimensional point cloud data representing the shape of a feature along a road using a laser scanner mounted on a vehicle (Patent Document 3). In this system, a laser scanner mounted on an automobile irradiates a laser obliquely downward or obliquely upward from the top of the vehicle body. The optical axis of the laser is scanned in the horizontal direction, and laser pulses are emitted at every minute angle within the scanning angle range. The distance is measured based on the time from the laser emission to the reception of the reflected light, and at that time, the laser emission direction, time, and the position / posture of the vehicle body are measured. From these measurement data, point group data representing the three-dimensional coordinates of the point reflecting the laser pulse is obtained. In addition, the system captures video using a digital camera simultaneously with the acquisition of point cloud data. The image can be used when the user designates a measurement target part in data analysis.

JP-A-8-86645 JP-A-9-287933 JP 2009-204615 A

  The laser scanner of the conventional measuring apparatus has a structure in which scanning with a laser beam is performed along the width direction of the automatic measuring vehicle mounted, and the scanning width is about the vehicle width. Therefore, the road surface to be measured is limited to the vicinity of the center of the lane, the end portion of the lane is difficult to measure, and there is a problem that the measuring vehicle must be run for each lane on a road with multiple lanes.

  In this regard, the mobile mapping system can acquire a high-density point group in a wide range including not only the road but also the roadside area at a time. On the other hand, analysis such as interpretation of features based on point cloud data obtained by the system is performed manually using an editing tool or the like in 3D CAD, and automation is under development. Similarly, if the road unevenness is detected manually, there is a problem that the burden on the operator is heavy. In particular, the unevenness of the road surface is not necessarily as clear as the features installed on the road or on the side of the road, and people can visually interpret the unevenness with high accuracy from the surface shape formed by the point cloud. Is not easy.

  The present invention relates to a road deformation detection device, a road deformation detection method, and a program capable of automating work for detecting unevenness based on three-dimensional coordinate data of a point cloud extracted from a structure surface along a road. The purpose is to provide.

  A road deformation detection device according to the present invention is a device for detecting unevenness on the surface of a target structure based on three-dimensional coordinate data of a point cloud extracted from the surface of the structure along the road, A reference line passing through the unevenness inspection position on an assumed flat reference surface is set, and one of the reference lines divided at the unevenness inspection position and the other side of each of the points set in advance Point group extraction means for extracting first and second attention point groups located along the reference line within a range, coordinates of the points constituting the attention point group, and the reference plane along the reference line With respect to regression analysis with coordinates along the normal direction to the explanatory variable and dependent variable, respectively, the hypothesis that the regression coefficients of the two points of interest match is F-tested at a predetermined significance level, and the hypothesis is rejected If the But having a uneven surface determining means for determining a present.

  In another road deformation detection device according to the present invention, the unevenness determination means tests the hypothesis by a Chow test.

  In still another road deformation detection device according to the present invention, the point cloud extraction means sets the non-land inspection positions at a plurality of locations on the reference line, and the same predetermined number at each non-land inspection position. The first attention point group and the second attention point group consisting of the points are set, and the non-land determination means follows an F value common to each other at the non-land inspection position. Is calculated, and information representing the variation of the F value along the reference line is generated.

  In the road deformation detection device, the point cloud extraction unit may set the reference plane corresponding to the surface of the road and set the reference line along the crossing direction of the road.

  In the road deformation detection device, the point cloud extraction unit sets a band-like region having a predetermined width along the reference line on the reference plane as the vicinity range, and sets the reference plane on the reference plane. The point-of-interest group can be constituted by points where the projected positions are within the band-like region.

  A road deformation detection method according to the present invention is a method for detecting unevenness on the surface of a target structure based on three-dimensional coordinate data of a point cloud extracted from the surface of the structure along the road, A reference line passing through the unevenness inspection position on an assumed flat reference surface is set, and one of the reference lines divided at the unevenness inspection position and the other side of each of the points set in advance A point group extracting step for extracting first and second attention point groups positioned along the reference line within a range; coordinates of the points constituting the attention point group along the reference line; and the reference plane With respect to regression analysis with coordinates along the normal direction to the explanatory variable and dependent variable, respectively, the hypothesis that the regression coefficients of the two points of interest match is F-tested at a predetermined significance level, and the hypothesis is rejected Corresponding to the uneven inspection position. Having a uneven surface determining step determines that the uneven surface is present.

  The program according to the present invention is a program for causing a computer to perform data analysis for detecting unevenness on the surface of the target structure based on the three-dimensional coordinate data of the point cloud extracted from the surface of the structure along the road. The computer sets a reference line passing through the non-land inspection position on a flat reference surface assuming the surface, and from each of the one side and the other side of the reference line divided at the non-land inspection position, respectively. , Point group extraction means for extracting first and second attention point groups located along the reference line within a preset vicinity range, and each point constituting the attention point group. With respect to regression analysis using coordinates along the reference line and coordinates along the normal direction to the reference plane as explanatory variables and dependent variables, respectively, a hypothesis that the regression coefficients of the two points of interest match is a predetermined significance level. so And test uneven surface determining means for determining the uneven surface is present that corresponds to the uneven surface inspection position when the hypothesis is rejected, to function as a.

  ADVANTAGE OF THE INVENTION According to this invention, automation of the operation | work which detects unevenness based on the three-dimensional coordinate data of the point cloud extracted from the structure surface along a road is achieved.

1 is a block diagram showing a schematic configuration of a road deformation detection system according to an embodiment of the present invention. It is a schematic flowchart of the unevenness detection process by the road deformation detection system which concerns on embodiment of this invention. It is a typical perspective view which shows the reference plane, reference line, and test | inspection cross section which are set in a non-land | grand | ground detection process. It is a typical top view explaining the partial space set by a non-land | grand | ground detection process. It is a schematic diagram of the inspection cross section explaining the setting of the non-land inspection position and the extraction of the attention point group in the non-land detection processing. It is a schematic diagram of the test | inspection cross section explaining the said regression analysis about the attention point group of the both sides of a non-land | gear inspection position. It is a schematic diagram which shows the example of the road surface profile and F value along a reference line. It is a schematic diagram which shows the other example of the road surface profile along a reference line, and F value. It is a schematic diagram which shows the further another example of the road surface profile and F value along a reference line.

  Hereinafter, a road deformation detection system 2 according to an embodiment of the present invention (hereinafter referred to as an embodiment) will be described with reference to the drawings. This system is a data analysis device that detects unevenness on the surface of the target structure based on the three-dimensional coordinate data of the point cloud extracted from the surface of the structure along the road. The structures along the road include various things such as side walls and tunnels in addition to the road surface. On the other hand, the target structure of this system basically has a surface that is expected to be smooth, and this system detects unevenness on the surface, and there is unevenness. Based on the above, the deformation of the structure is detected.

  The point cloud data is acquired by, for example, a laser scanner mounted on a vehicle traveling on the ground like the above-described mobile mapping system. In addition, measurement may be performed by installing a laser scanner on the ground. Here, a plurality of measurement points constituting the point group are discretely located in the space. In order for this point group to capture uneven shapes such as the road surface, it is necessary to perform laser scanning at a density corresponding to the uneven scale. In this respect, a laser scan performed from the height of a vehicle, a tripod or the like can obtain a sufficient scanning density, unlike a laser scan performed from a height such as an aircraft.

  FIG. 1 is a block diagram showing a schematic configuration of a road deformation detection system 2. The system includes an arithmetic processing device 4, a storage device 6, an input device 8, and an output device 10. As the arithmetic processing device 4, it is possible to make dedicated hardware for performing various arithmetic processing of this system, but in this embodiment, the arithmetic processing device 4 uses a computer and a program executed on the computer. Built.

  A CPU (Central Processing Unit) of the computer constitutes the arithmetic processing unit 4 and functions as a point group extraction unit 20 and a non-land determination unit 22 described later.

  The storage device 6 is composed of a hard disk or the like built in the computer. The storage device 6 stores a program for causing the arithmetic processing unit 4 to function as the point group extraction unit 20 and the unevenness determination unit 22 and other programs, and various data necessary for the processing of this system. For example, the storage device 6 stores point cloud data of the analysis target space as the processing target data.

  The input device 8 is a keyboard, a mouse, or the like, and is used for a user to operate the system.

  The output device 10 is a display, a printer, or the like, and is used for displaying the result of analysis of unevenness by this system to the user by screen display, printing, or the like.

  Hereinafter, a case will be described as an example where the target structure for non-land detection is a paved road surface and the road deformation detection system 2 detects non-land using a road surface profile along the crossing direction of the road. FIG. 2 is a schematic flowchart of road surface unevenness detection processing by the road deformation detection system 2.

  Here, a virtual road surface having no unevenness is set as a reference plane, and a reference line extending in the road crossing direction is set on the flat reference plane. The reference line is the x-axis of the xyz orthogonal coordinate system, and the longitudinal direction of the road (or the traveling direction of the vehicle) and the normal direction of the reference plane are the y-axis and z-axis, respectively.

  As will be described later, the inspection of unevenness is performed on the reference line, and the unevenness in the xz section of the road is evaluated. That is, an xz cross section that is a non-land inspection cross section corresponding to the reference line is set (S10). FIG. 3 is a schematic perspective view showing a reference plane, a reference line, and an inspection section. The reference plane 30 is set substantially parallel to the road surface, and the reference plane 30 defines an xy plane. A reference line 34 is set on the reference plane 30 along the crossing direction (x-axis direction) of the road 32. Then, an inspection cross section 36 orthogonal to the reference plane 30 is set along the reference line 34.

  The point cloud extraction means 20 extracts from the point cloud data stored in the storage device 6 a point cloud located along a reference line within a preset neighborhood range as a partial point cloud corresponding to the examination section (S20). . For example, a band-like area having a predetermined width along the reference line is set on the reference plane as the vicinity range, and a measurement point where the position projected on the reference plane falls within the band-like area is extracted.

Specifically, the point group extraction means 20 sets a partial space that includes a road surface of a rectangular parallelepiped shape having sides along the x, y, and z axes and that performs unevenness detection. FIG. 4 is a schematic plan view for explaining the partial space. For example, the size η _{x in} the x direction of the partial space is set to about 5 m corresponding to the road width, and the size η _{y in the} y direction is set to a size including a plurality of measurement points in the direction. Here, the scan position of the laser in the y direction moves as the vehicle travels. The size η _{y in} the y direction of the partial space can be set so as to include, for example, scan positions of about 3 to 5 cycles. In the present embodiment, the size η _y is 50 cm. Regarding the z direction, if objects such as tunnel ceilings and road signs exist above the road surface, they become noise for the analysis of the road surface shape. Therefore, the point cloud extraction range (position, size) in the z direction is set in advance so as to exclude the point cloud caused by them. Alternatively, only the points acquired by the obliquely downward laser scanner among the points obtained by the obliquely downward laser scanner and the obliquely upward laser scanner may be used for the road surface shape analysis.

  In addition, from the processing contents of the unevenness determination means 22 described later, even if the reference surface and the road surface are not parallel, the unevenness detection is not so much affected, and the reference surface is set with a slight angle with respect to the road surface. It is permissible to do.

The point cloud extraction means 20 sets the unevenness inspection position on the reference line (S30). Figure 5 is a schematic view of an inspection section for explaining the extraction of the attention point group in the configuration and the next step S40 the uneven surface inspection position P _C in step S30. FIG. 5 shows an example of the reference line 34 and an image 40 projected onto the inspection cross section of the partial point group existing in the partial space. When the x-coordinate of the uneven surface inspection position P _C and x _C, point group extraction unit 20, the first point of interest comprising a measurement point having x <x _C becomes x-coordinate of the extracted portion point group in step S20 a group _{G a,} and a second target point group _{G B} formed of the measurement points having x ≧ _{x C} becomes x-coordinate is extracted (S40). In the present embodiment, attention point group G _A, a predetermined number measurement points constituting the G _B from close to uneven surface inspection position P _C, respectively in the order n _A, is selected by n _B. In the present embodiment, n _A and n _B are set to the same number, for example, n _A = n _B = 12.

Note that n _A and n _B are preferably the same number, but may be different. Further, for example, reference lines at uneven surface inspection position _{P C} from the same width w on opposite sides interval _{[x C -w, x C)} , [x C, attention point group measurement points existing _x C + w] in the G _a, may be a _{G B.}

Uneven surface determining means 22 target point group extracted by the point cloud extracting section 20 G _A, whether discontinuity in the road surface using a G _B at uneven surface inspection position P _C is generated, using a regression analysis An assay is performed (S50). The unevenness determination means 22 uses coordinates along the reference line of each measurement point constituting the target point group, that is, the x coordinate as explanatory variables, and coordinates along the normal direction with respect to the reference plane, that is, the z coordinate as a dependent variable. Perform regression analysis. Figure 6 is a schematic view of an inspection section illustrating the regression analysis, projected onto the test section on either side of the target point groups uneven surface inspection position P _C is shown. 6, a straight line 50 the coordinates of each measurement point for the target point group G _A (x, z) is a regression line obtained by regression analysis using, similarly, straight lines 52 for group target point G _B It is the regression line obtained by regression analysis. Uneven surface determining means 22 the regression coefficient representing the regression line 50, 52 of the target point group G _A, G _B match, i.e. a road surface point of interest group G _A located on both sides of the uneven surface inspection position P _C was obtained road surface condition in the road surface obtained attention point group G _B is F-test hypotheses (null hypothesis) at a predetermined significance level to be equivalent to. Then, the alternative hypothesis of the retrace has occurred a change in the road surface condition on both sides of Furiku inspection position P _C if hypothesis is rejected, is Furiku the testing position P _C or vicinity thereof there is correct judge.

  In this embodiment, a Chow test is used as the F test model. The Chow test calculates the F value according to the following equation.

_Here, RSS _A, RSS _B is the residual sum of squares regression analysis for each interest point group _G A, _{G B.} Further, in the Chow test target point group _G A, perform regression analysis also coordinates of the point group combined both _{G B (x, z),} RSS A, B is the residual sum of squares of the regression analysis . As described above, n _A and n _B are the number of measurement points constituting the attention point groups G _A and G _B , respectively. k is the number of explanatory variables, and is 1 here.

For example, uneven surface determining means 22 is a 95% probability, if the null hypothesis is rejected That at significance level alpha = 0.05, the uneven surface of the road surface at a location corresponding to the inspection position P _C is generated determination (S60). Incidentally, the F value in the equation (1) follows an F distribution with (k + 1, n _A + n _B −2k−2) degrees of freedom. The rejection range of the F value is determined according to the degree of freedom for each significance level α, and the null hypothesis is rejected if the F _value is larger than the boundary value F ₀ determined according to α and the degree of freedom.

The point group extraction means 20 and the unevenness determination means 22 determine whether or not the unevenness inspection has been completed for all the positions to be inspected that are set in the crossing direction of the road for the partial point group (S70). Is present (in the case of “Yes” in S70), the processing of steps S30 to S60 is repeated by changing the inspection position. For example, the point group extraction unit 20 Furiku inspection position P _C of the x target point groups by sequentially moving by a predetermined distance along the axis G _A, or to extract G _B, interest point groups G _A at the measurement point based, G _B is moved to search for non-land along the reference line.

On the other hand, when no uninspected position remains (in the case of “No” in S70), the unevenness determination means 22 evaluates the road surface property based on the F value obtained at each point of the reference line ( S80). Specifically, the degree of road surface unevenness is evaluated by the magnitude of the F value, or the degree of damage to the road surface property is determined by the occurrence ratio of the inspection position where the F value becomes a rejection area in the evaluation section set along the reference line. Can be evaluated. Here, by making the number of measurement points constituting the attention point groups G _A and G _{B the} same at each inspection position, the degree of freedom becomes the same, and the boundary value of the F value at each inspection position is common. That is, since the respective F values can be simply compared between the inspection positions on the reference line without conversion, the above-described road surface property can be easily evaluated. For example, the unevenness determination means 22 determines that the degree of unevenness is large at a location where the F value exceeds a certain threshold, or the occurrence rate of the inspection position where the F value becomes a rejection area exceeds a certain threshold. It is determined that the degree of damage is large at the position of the reference line.

Steps S <b> 10 to S <b> 80 described above are evaluation of unevenness with a one-dimensional road surface profile in the x-axis direction. FIGS. 7 to 9 are schematic diagrams showing examples of road surface profiles and F values along the reference line, with measurement points indicated by ◆ and F values indicated by ◯. The horizontal axis is the x coordinate (unit is meter), the left vertical axis is the z coordinate (unit is meter) of the measurement point with the height of the laser scanner as the origin, and the right vertical axis is Represents the F value. A thick horizontal line indicates a value 3.49 of the boundary value F ₀ when α is 0.05 and the degree of freedom is (2, 20). In the example of FIG. 7, there is one depression on the road surface with both slopes, and the F value at that location is higher than the other portions. In the example of FIG. 8, the unevenness of the road surface spreads as a whole, and places with high F values appear dispersed in the width. In the example of FIG. 9, one place where the F value is extremely high appears on the left side of the road surface with a single slope. For example, the place is extracted as a locally damaged place or a road shoulder or a side groove.

  The point cloud extraction means 20 newly sets the inspection cross section in a position shifted in the y direction every time the unevenness evaluation (S20 to S80) on the inspection cross section set in step S10 is completed. For example, the point cloud extraction means 20 sets the inspection cross section at equal intervals λ along the longitudinal direction of the road as shown in FIG. For example, λ can be about 5 m. In this way, by setting inspection sections at a plurality of locations in the longitudinal direction of the road, the evaluation of unevenness with a two-dimensional road surface profile in the xy plane is performed.

  The above-mentioned detection of unevenness by the road deformation detection system 2 is performed at each position along the reference line, and the road surface property can be locally evaluated. On the other hand, unevenness can be detected over the entire width, and unevenness can be detected by setting inspection sections at arbitrary intervals in the longitudinal direction of the road, so that the road surface can be evaluated as a whole. That is, according to the road deformation detection system 2 according to the present invention, both micro evaluation and macro evaluation are possible.

In the above-described embodiment, a point group in a partial space in which the planar shape is a band shape having a width η _{y in the} y direction is used. By using measurement points at a plurality of positions in the y direction in this way, it is possible to reduce the influence of variations caused by various factors on the coordinate values of the point cloud data obtained by the mobile mapping system.

In the set partial space, changing the number of measurement points n _A and n _B to be extracted as the point of interest group changes the range in the x direction in which the point of interest group is distributed. In the unevenness detection method of the above-described embodiment, for example, if n _A and n _B are increased or the width w in the x direction for extracting the point of interest is increased, it is difficult to detect unevenness with a small size in the x direction. . Therefore, by setting n _A and n _B , or w according to the scale of the unevenness that is the detection target, the scale that is not the detection target is less affected by the small unevenness, and the target scale is The land can be detected suitably. For example, the width in the x direction of rutting of a road with a large amount of traffic of a large vehicle can be larger than a road with a small amount of traffic of a large vehicle. Therefore, in the road deformation detection system 2, n _A and n _B , or w can be configured to be changed so that rutting with different scales can be discriminated and detected. For example, in the process described with reference to FIG. 2, even if the processes for steps S40 to S70 and the process of S80 are performed with a plurality of types of n _A and n _B , or w in the analysis for each set inspection section. Good.

  In the above-described embodiment, the reference line is set in the crossing direction of the road. However, the same unevenness inspection can be performed by setting the reference line in an arbitrary direction such as a longitudinal direction.

_A model other than the Chow test can be used as the F test model for testing the hypothesis that the regression coefficients of the attention point groups G _A and G _B match.

  Moreover, the surface of the structure for detecting unevenness is not limited to the road surface, and the present invention can be similarly applied to, for example, side walls, slopes, and flat ceiling panels of tunnels. Furthermore, for the inner surface having the curvature of a tunnel whose cross-section in the road crossing direction is horseshoe-shaped, egg-shaped, or circular, a reference surface that has a curvature but is smooth is assumed, assuming the inner surface. It is possible to perform unevenness inspection along a reference line set on the surface. Specifically, since a straight reference line can first be set in the longitudinal direction of the road, the unevenness inspection can be basically performed as in the above-described embodiment. Further, even when the reference line is a curve, an equation representing the reference line is obtained. Therefore, in the above-described embodiment, it is possible to perform the unevenness inspection using the linear regression analysis as a regression analysis of the curve. .

  2 Road deformation detection system, 4 arithmetic processing device, 6 storage device, 8 input device, 10 output device, 20 point group extraction means, 22 unevenness determination means, 30 reference plane, 34 reference line, 36 inspection section.

Claims (7)

A road deformation detection device that detects unevenness on the surface of a target structure based on three-dimensional coordinate data of a point cloud extracted from the surface of the structure along the road,
A reference line passing through the non-land inspection position is set on a flat reference surface assuming the surface, and each of the point groups in advance from one side and the other side of the reference line divided at the non-land inspection position. Point cloud extraction means for extracting first and second attention point clouds located along the reference line within a set neighborhood range;
With respect to regression analysis in which coordinates along the reference line of each point constituting the point of interest group and coordinates along the normal direction with respect to the reference plane are respectively an explanatory variable and a dependent variable, the regression coefficients of the two point of interest groups F-test the hypothesis that the two coincide with each other at a predetermined significance level, and when the hypothesis is rejected, a non-land determination means that determines that there is a non-land corresponding to the non-land inspection position;
A road deformation detection device comprising: In the road deformation detection device according to claim 1,
The road irregularity detecting means tests the hypothesis by a Chow test. In the road deformation detection device according to claim 1 or 2,
The point group extraction means sets the non-land inspection positions at a plurality of locations on the reference line, and the first attention point group and the second group of points having the same predetermined number of points at each non-land inspection position. Set the point of interest for
The unevenness determination means calculates an F value that follows a common F distribution at each of the unevenness inspection positions, and generates information representing the variation of the F value along the reference line,
A road deformation detection device characterized by the above. In the road deformation detection device according to claim 3,
The road group detection unit, wherein the point group extraction unit sets the reference plane corresponding to a road surface, and sets the reference line along a crossing direction of the road. In the road deformation detection device according to claim 3,
The point group extraction unit sets a band-like area having a predetermined width along the reference line on the reference plane as the neighborhood range, and a position projected on the reference plane is within the band-like area. A road deformation detection device characterized in that the point-of-interest group is constituted by points. A road deformation detection method for detecting unevenness on the surface of a target structure based on three-dimensional coordinate data of a point cloud extracted from the surface of the structure along the road,
A reference line passing through the non-land inspection position is set on a flat reference surface assuming the surface, and each of the point groups in advance from one side and the other side of the reference line divided at the non-land inspection position. A point group extraction step of extracting first and second attention point groups located along the reference line within a set neighborhood range;
With respect to regression analysis in which coordinates along the reference line of each point constituting the point of interest group and coordinates along the normal direction with respect to the reference plane are respectively an explanatory variable and a dependent variable, the regression coefficients of the two point of interest groups A non-land determination step that F-tests the hypothesis that the two are coincident with each other at a predetermined significance level, and determines that there is a non-land corresponding to the non-land inspection position when the hypothesis is rejected;
A road deformation detection method characterized by comprising: A program for causing a computer to perform data analysis for detecting unevenness on the surface of a target structure based on three-dimensional coordinate data of a point cloud extracted from the surface of the structure along a road, the computer comprising:
A reference line passing through the non-land inspection position is set on a flat reference surface assuming the surface, and each of the point groups in advance from one side and the other side of the reference line divided at the non-land inspection position. Point cloud extracting means for extracting first and second focused point clouds located along the reference line within a set neighborhood range; and
With respect to regression analysis in which coordinates along the reference line of each point constituting the point of interest group and coordinates along the normal direction with respect to the reference plane are respectively an explanatory variable and a dependent variable, the regression coefficients of the two point of interest groups F-test the hypothesis that the two coincide with each other at a predetermined significance level, and when the hypothesis is rejected, function as non-land judgment means for judging that the land corresponding to the non-land inspection position exists. Program.

Images