Method for measuring position of alternative target, point cloud measurement apparatus, point cloud information processing apparatus, and medium

Abstract

To identify an instrument point of a point cloud measurement apparatus by using a substitute target instead of a dedicated target.SOLUTION: A point cloud measurement unit measures a measurement point cloud including a substitute target. A point cloud matching unit uses a reference model including a reference model point cloud that has been previously generated and assigned coordinates representing relative positional information corresponding to shape features of the substitute target. The point cloud matching unit compares the reference model point cloud with the measurement point cloud, extracts at least one measurement model point clouds corresponding to the reference model point cloud from the measurement point cloud, and generates one or more measurement models. A coordinate transformation parameter generation unit generates coordinate transformation parameters representing movement amounts required to move either the reference model point cloud or the measurement model point cloud so that one coincides with the other, on the basis of the coordinate differences between the reference model point cloud and the measurement model point cloud. A measurement model position estimation unit estimates at least one position of the measurement model using the coordinate transformation parameters.SELECTED DRAWING: Figure 4

Description

[Technical Field]

[0001] The present disclosure relates to a method for measuring the position of a substitute target that substitutes for a dedicated target for combining multiple measurement point clouds measured from different positions using a point cloud measurement device. [Background technology]

[0002] Conventionally, laser scanners and the like have been known as point cloud measurement devices that acquire three-dimensional data of a survey target. A process known as registration, in which multiple measurement point clouds acquired from different viewpoints by such point cloud measurement devices are matched and combined into a single point cloud, is known (Patent Document 1). [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Publication No. 2022-157660 Summary of the Invention [Problem to be solved by the invention]

[0004] In order to perform point cloud measurements from multiple different viewpoints and combine the point clouds through registration, dedicated targets are required to identify the instrument position of the point cloud measurement device for each viewpoint. Dedicated targets are installed as needed as reference points for identifying the instrument position of the point cloud measurement device. Examples of dedicated targets include black-and-white checkerboards, spheres, and reflecting prisms. These are specialized parts incorporating various optical technologies, and are expensive to obtain, which limits the number available. As a result, there may be a shortage of dedicated targets at some work sites, and tasks such as repositioning may be required, necessitating careful planning of target installation in advance, resulting in various burdens.

[0005] Therefore, one aspect of the present disclosure aims to grasp the instrument points of a point cloud measurement device using an alternative target that replaces a dedicated target in order to grasp the instrument points of the group measurement device. [Means for solving the problem]

[0006] In order to achieve the above object, a position measurement method for a substitute target that substitutes a dedicated target for combining a plurality of measurement point clouds measured from different positions using a point cloud measurement device according to the present disclosure includes: a measurement point cloud measurement step in which a point cloud measurement unit capable of measuring a point cloud to which coordinates that serve as relative position information with respect to an instrument point where measurement is performed measures a measurement point cloud including a substitute target; and a point cloud matching step in which a reference model including a reference model point cloud that has been generated in advance and to which coordinates that serve as relative position information that represent shape features of the substitute target are assigned, and a point cloud matching step in which a reference model point cloud is compared with the measurement point cloud to match the measurement point cloud. a measurement model generation step of extracting one or more measurement model point clouds corresponding to the reference model point cloud from the group and generating one or more measurement models; a coordinate transformation parameter generation step of a coordinate transformation parameter generation unit generating a coordinate transformation parameter representing the amount of movement required to move one of the reference model point cloud or the measurement model point cloud to the position of the other based on the difference in coordinates between the reference model point cloud and the measurement model point cloud; and a measurement model position estimation step of a measurement model position estimation unit estimating at least one position of the measurement model using the coordinate transformation parameter. [Effects of the Invention]

[0007] By using the alternative target position measurement method of the present disclosure, which uses the above-mentioned means, it is possible to determine the instrument points of the point cloud measurement device using an alternative target that replaces the dedicated target, rather than a dedicated target. [Brief explanation of the drawings]

[0008] [Figure 1] FIG. 10 is a diagram illustrating an example of a dedicated target. [Figure 2] 1 is an overall configuration diagram of a point cloud measurement system according to an embodiment of the present disclosure. [Figure 3] FIG. 10 is a sequence chart illustrating a method for measuring the position of a substitute target according to an embodiment of the present disclosure. [Figure 4] FIG. 10 is a schematic diagram illustrating the generation of a reference model. [Figure 5] FIG. 1 is a schematic diagram illustrating generation of a measurement model and position measurement. [Figure 6] FIG. 10 is a diagram illustrating an example of an alternative target. [Figure 7] FIG. 1 is a diagram showing the layout in an evaluation test of an example. [Figure 8] FIG. 1 is a diagram showing the layout in an evaluation test of an example. [Figure 9] 10A and 10B are diagrams illustrating evaluation results of the coordinate accuracy of representative position points of alternative targets according to an embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0009] <Introduction> When using a point cloud measurement device, such as a Terrestrial Laser Scanner (TLS), it is necessary to scan the object or location to ensure there are no blind spots, so the TLS must be repositioned and scanned repeatedly.When doing so, there are points that can be considered identical, corresponding points, or overlapping points between the different installation positions of the TLS, i.e., between the 3D point cloud data acquired for each instrument point, so a process called "registration" is performed to combine these points into a single point.

[0010] Typical registration techniques use black and white checkered targets TG1 and sphere targets TG4, as shown in Figure 1. These are dedicated targets that serve as useful reference points and corresponding points when there are no distinctive landmarks at the site where point cloud acquisition is being performed. By using dedicated targets, the center coordinates of the dedicated targets can be determined, and the relative positions of instrument points can be determined using these center coordinates, allowing point clouds to be registered. There are also targetless products on the market that can automatically track instrument points using GNSS or IMU, but using targets provides higher coordinate accuracy for point clouds.

[0011] One example of a dedicated target is a black-and-white checkered target, which is attached to the top of a pole or other location at the work site. However, it must be aligned with the instrument for each scan, requiring manual rotation of the target. Sphere targets, on the other hand, do not require rotation for alignment, but are more expensive due to their precise geometric shape. These dedicated targets are surprisingly expensive because they are optically processed and machined components, and preparing a sufficient number of them is costly. Furthermore, the limited number of targets requires the time and effort of juggling and repositioning a small number of targets.

[0012] Therefore, in this disclosure, a method for reducing registration costs has been investigated by using common components or equipment that are relatively easy to obtain at the work site as registration targets instead of dedicated targets.

[0013] In the registration disclosed herein, common components and equipment at work sites, more specifically, work-related tools that are normally present at work sites but are not normally used as survey targets, are used as substitute targets for dedicated targets. The work-related tools referred to here include cones such as color cones (registered trademark), safety cones, triangular cones, and pylons, as well as construction machinery such as hydraulic excavators, backhoes, and heavy machinery. These work-related tools are preferable because they are normally present at work sites and do not need to be specially procured.

[0014] These substitute targets are scanned to obtain point clouds, from which a "reference model" of the substitute target is generated. Using this, it is possible to extract another substitute target, a "measurement model," with a shape that can be regarded as the same as the reference model from unknown point clouds measured from different positions. Then, by using the reference model or measurement model as a virtual target, registration between different measurement point clouds becomes possible without using dedicated targets. The configuration and processing flow for this purpose are explained below.

[0015] <Configuration> 2 shows an example of the configuration of a point cloud measurement device or a point cloud measurement system for using the alternative target position measurement method of the present disclosure. The configuration of a point cloud measurement system 1 including at least a point cloud measurement device 20 will be described below with reference to these figures.

[0016] The point cloud measurement system 1 has at least two functions. One is a point cloud measurement function, and the other is a point cloud information processing function. Therefore, it is provided with a point cloud measurement device 20 that has at least the point cloud measurement function. The point cloud information processing function may be performed by a point cloud information processing device 10 described below, or the functions of the point cloud information processing device 10 may be performed integrally by the point cloud measurement device 20, or it may be ensured by connecting to another information processing device such as a server device installed in a remote location with which wired or wireless communication is possible, and communicating information with that information processing device.

[0017] An example of the point cloud measurement device 20 is a 3D laser scanner such as a TLS installed on the ground, but the implementation is not limited to this and the device may be mounted on an unmanned aerial vehicle (UAV) or other moving body. An example of the configuration of the point cloud measurement device 20 may include, for example, an imaging unit 21, a point cloud measurement unit 22, and a GNSS (Global Navigation Satellite System) 23. Furthermore, the point cloud measurement device 20 is connected to other devices via a communication unit 24 so as to be able to electrically communicate with them via wire or wirelessly.

[0018] The point cloud measurement unit 22 may be a so-called 3D laser scanner that scans by scanning a laser beam (distance measurement beam) back and forth in the vertical direction within a predetermined range and rotating it horizontally to acquire a point cloud and generate point cloud information. This allows the relative distance from the point cloud measurement unit 22 to the target object to be measured based on the time it takes for the laser beam to hit the object, reflect, and return. Furthermore, the irradiation direction of the laser beam (horizontal angle and vertical angle) is also detected, and the relative angle of the measurement point is measured based on this. The point cloud measurement unit 22 calculates the three-dimensional coordinates of each point based on this relative distance and relative angle data. Therefore, the acquired point cloud is assigned three-dimensional coordinates in a predetermined coordinate system (e.g., the TLS coordinate system).

[0019] The imaging unit 21 may be a camera that captures a two-dimensional image of light in a predetermined wavelength range, such as visible light or infrared light, and may have the function of capturing an image of a surveying object and generating information such as RGB intensity at each pixel. This image information may include horizontal angle information relative to the point cloud measurement device at the time of capturing the image, and may also be capable of calculating angle information relative to the direction of laser light irradiation in the point cloud measurement unit 22. This may enable the position of each point acquired by the point cloud measurement unit 22 to be associated with a position in the image captured by the imaging unit 21. This may allow each point cloud data to be assigned a scalar value such as RGB intensity.

[0020] The GNSS 23 functions as, for example, a GPS and can acquire global position coordinate information. This makes it possible to acquire position information of the point cloud measurement device 20, i.e., viewpoint position information of the point cloud measurement device 20. This viewpoint position information functions as a laser light irradiation base point in the point cloud measurement unit 22 and as an imaging base point in the imaging unit 21. If the point cloud measurement device 20 is installed at a known point, position information can be acquired through input operations by the operator without using the GNSS.

[0021] In addition, the point cloud measurement unit 22 can include viewpoint position information (laser light irradiation base point) in the point cloud information based on absolute position information of the point cloud measurement device 20. Furthermore, the point cloud measurement unit 22 may have a function to convert relative position information (a predetermined coordinate system, for example, the TLS coordinate system) of the point cloud acquired by laser scanning into another coordinate system. Similarly, the imaging unit 21 may have a function to include viewpoint position information (imaging base point) in the image information based on position information of the point cloud measurement device 20.

[0022] The point cloud information processing device 10 may have the function of calculating the positional relationship between multiple point cloud information of an object obtained from different positions and aligning (registering) that information. It may also have the function of creating a three-dimensional model representing a predetermined shape using the measured point cloud information. The point cloud information processing device 10 may be configured as a general-purpose computer with installed software, such as a handheld device such as a smartphone or tablet, or a laptop PC. These devices may have a data input unit 11 such as a mouse or touch panel, an output unit 12 for displaying images, a memory unit 13 for saving data, a communication unit 14, and a processing unit 30. These components may communicate with devices other than the point cloud information processing device via wired or wireless connections, allowing for external information input and display and storage of processing results. Furthermore, the point cloud information processing device 10 does not need to have its own hardware computing resources. It may be a portable information terminal that simply inputs and outputs information by communicating with a processing unit 30 implemented on a remotely installed server or other so-called cloud computer.

[0023] The point cloud information processing device 10 may include a processing unit 30, which is realized by executing software, programs, applications, etc. using hardware computational resources such as a central processing unit of a general-purpose computer, and which performs each process, and which includes a reference model generation unit 31, a representative position point determination unit 32, a point cloud matching unit 33, a coordinate transformation parameter generation unit 34, a measurement model position estimation unit 35, and a point cloud combination unit 36. Data input to the processing unit 30 via an input unit 11 and output as a result of processing, such as intermediate or final output, is output to an output unit 12, stored in a memory unit 13, and can be communicated via a communication unit 14. These computers may include a medium such as a memory unit 13 that stores a program for executing the functions of the processing unit 30.

[0024] The reference model generation unit 31 has the function of generating a reference model including a reference point cloud model consisting of multiple point clouds representing alternative targets, using a point cloud for generating a reference model including an alternative target measured by the point cloud measurement device 20.

[0025] The reference model is a virtual three-dimensional model representing the geometric characteristics of the alternative target, and may be composed of a reference model point cloud extracted from the measured point cloud. The reference model generation unit 31 may automatically generate the reference model from the measured point cloud based on, for example, predetermined geometric characteristics, point cloud density, color (RGB intensity), etc. Alternatively, the user may select a predetermined range from the point cloud displayed on the output unit 12 using the input unit 11, and the reference model may be generated based on the selected range.

[0026] Measurements do not need to be performed multiple times, and a reference model may be generated based on a point cloud acquired by measurement from only one viewpoint. The generated reference model does not need to perfectly reproduce the shape features of the substitute target on which it was based, and some blind spots and occlusions may be allowed. Of course, a reference model may also be generated using point clouds acquired by multiple point cloud measurements or by measuring the point cloud while moving around the surrounding area.

[0027] The reference model generation unit 31 may have a trimming function for removing unnecessary points from the reference model generation point cloud, for example, a function for automatically removing noise points based on a predetermined point cloud density distribution or the like.

[0028] Furthermore, the reference model may be model data other than a point cloud, such as mesh data composed of points and surfaces, as long as a means or program for converting it into a format that can be compared with the point cloud is available.

[0029] Creating a reference model of such a substitute target is different from conventional point cloud registration using a dedicated target, at least in that, as mentioned above, the dedicated target uses its center position coordinates, but the dedicated target itself is not modeled as a point cloud and used as a reference point.

[0030] The representative position point determination unit 32 has a function of determining a reference model representative position point whose relative position with respect to the reference model satisfies a predetermined condition.

[0031] A representative position point is a point that represents a representative position in a reference model or a measurement model (described later), and does not necessarily have to be on the model. It may be any point whose relative position with respect to the reference model is determined. A representative position point in a reference model corresponds to the center position coordinates of a conventional dedicated target, and is useful for determining the position of a substitute target. Therefore, multiple representative position points may be determined for one reference model.

[0032] Therefore, determining a representative position point is not necessarily required, but is merely for the convenience of point cloud registration. If a representative position point is not determined, the position, etc. can be estimated using any point or point cloud included in the reference model point cloud that constitutes the reference model.

[0033] The point cloud matching unit 33 has the function of comparing the reference model point cloud with the measurement point cloud measured by the point cloud measurement device 20 using a reference model including a reference model point cloud that has been generated in advance before matching, extracting one or more measurement model point clouds that correspond to the reference model point cloud from the measurement point cloud, and generating one or more measurement models.

[0034] A measurement model is a virtual model extracted from a measurement point cloud by matching a point cloud that corresponds to a reference model and can be considered the same as the reference model. If a measurement model can be extracted from a measurement point cloud, it means that a substitute target has been installed there, and if the position of the measurement model can be estimated, the position of the substitute target can be estimated. This will be discussed later.

[0035] Point cloud matching can be achieved by employing known methods and algorithms. The most well-known point cloud matching method is ICP (Iterative Closest Point), which determines how to change the position and orientation of one point cloud (moving point cloud) relative to another point cloud (fixed point cloud), iteratively finds neighboring points of each point, and aligns the point clouds by minimizing the distance between those neighboring points. Other known point cloud matching algorithms include NDT (Normal Distributions Transform), GMMReg, SVR, and derivatives of these. Note that these point cloud matching algorithms do not need to perfectly match two measured point clouds; they can align them so that errors and distances converge and a certain degree of agreement is achieved.

[0036] The coordinate transformation parameter generation unit 34 has the function of generating coordinate transformation parameters, which are transformation matrices such as rotation amounts, translation amounts, etc., required to move the position of one point cloud (moving point cloud, source point cloud) to the position of another point cloud (fixed point cloud, target point cloud) using a point cloud matching algorithm.

[0037] A transformation matrix is, for example, R, which represents a rotation matrix.ij and the translation amount T x , T y , T z Using this transformation matrix, you can rotate and translate each point in a point cloud, and move the entire point cloud to the position of another point cloud.

[0038] The measurement model position estimation unit 35 has a function of estimating at least one position of the measurement model using the coordinate transformation parameters.

[0039] If a representative position point has been determined for the reference model, the position of the measurement model can be estimated simply by moving the representative position point using coordinate transformation parameters such as the transformation matrix described above. Even if a representative position point has not been determined, the position of the measurement model can be estimated by moving any point or point group in the reference model point group using coordinate transformation parameters.

[0040] The point cloud combining unit 36 ​​has a function of combining a plurality of measurement point clouds measured from different positions using the position information of one or more measurement models estimated above.

[0041] In other words, the point cloud combining unit 36 ​​of the present disclosure performs point cloud registration by using the representative position points of the reference model or the representative position points of the measurement model as the position coordinates of the substitute target instead of the central position coordinates of the dedicated target. The point cloud registration method is the same as when using a dedicated target, so a description thereof will be omitted.

[0042] <Processing flow> Next, a method for measuring the position of an alternative target that replaces a dedicated target for combining multiple measurement point clouds measured from different positions using the point cloud measurement device 20 according to this embodiment will be described with reference to the sequence chart shown in Figure 3.

[0043] In step S101, the point cloud measuring device 20 measures a point cloud for generating a reference model including a substitute target (reference model measuring step).

[0044] 4 is a schematic diagram for explaining the generation of a reference model. In this figure, a point cloud scan, which is a reference model point cloud measurement RPM, is performed using an alternative target ATG1 in an appropriate positional relationship and a point cloud measurement device 20. The alternative target ATG1 is, for example, a cone.

[0045] In step S102, the reference model generation unit 31 uses the point cloud for generating the reference model included in the point cloud measured in step S101 to generate a reference model RM1 including a reference model point cloud RMP1 consisting of multiple point clouds representing the shape characteristics of the alternative target ATG1 (reference model generation step).

[0046] In step S103, the representative position point determination unit 32 determines a reference model representative position point RPP1 whose relative position with respect to the reference model RM1 satisfies a predetermined condition (reference model representative position point determination step). For example, in Fig. 4, the top of the cone is determined as the representative position point.

[0047] Another example of the reference model will now be described. FIG. 6 is a schematic diagram illustrating another example of the reference model. In this diagram, a shovel is used as the surrogate target ATG3. When using such a surrogate target, it is important to note that the geometric features to be extracted as the reference model are immovable parts. There is no problem in using an immovable object such as a cone as the reference model as is, but for a work tool that can move, such as a shovel, it is necessary to create a reference model using immovable parts, such as the back of the vehicle where the counterweight is located, as geometric features. Therefore, in this example, the rear panel of the vehicle is used as the reference model RMP3.

[0048] Furthermore, when the coordinate accuracy of the reference model is highly reliable, such as when generating a reference model over a relatively large area like a shovel, multiple representative position points may be determined, such as RPP4 to 6. These representative position points can be treated as the central position coordinates of the dedicated target.

[0049] Next, the generation of a measurement model and position measurement will be explained. Fig. 5 is a schematic diagram showing the generation of a measurement model and position measurement. In this diagram, for the sake of convenience of explanation, it is assumed that the viewpoint position (irradiation base point of laser light) of the point cloud measurement device 20 coincides with the representative position point of the reference model RM1. Even if this is not the case, if the relative position between the irradiation bright point of the laser light and the reference model RM1 is known, it is possible to perform coordinate calculations taking this into consideration.

[0050] It is assumed that one or more alternative targets are installed within a range where the point cloud can be acquired by the point cloud measurement device 20. For ease of explanation, it is assumed here that one alternative target ATG2, which is a cone, is installed. This cone has a shape that can be regarded as the same as the alternative target ATG1 described above.

[0051] In step S104, the point cloud measurement unit 22 measures a measurement point cloud using the measurement model point cloud measurement MPM so that the alternative target ATG2 is included (measurement point cloud measurement step). Here, the measured point cloud includes both a point cloud that matches the reference model and a point cloud that does not match.

[0052] In step S105, the point cloud matching unit 33 compares the reference model point cloud RMP with the measurement point cloud using a reference model RM1 including the reference model point cloud RMP1, extracts one or more measurement model point clouds MMP corresponding to the reference model point cloud from the measurement point cloud, and generates one or more measurement models (measurement model generation step). For ease of understanding, this diagram shows an example in which only one measurement model is extracted, but multiple measurement models may be extracted from one measurement point cloud. Also, in this diagram, point clouds that match with the reference model are indicated by solid white circles, and non-matching point clouds are indicated by dotted colored circles.

[0053] In step S106, the coordinate transformation parameter generation unit 34 generates a coordinate transformation parameter CTP that represents the amount of movement required to move either the reference model point group or the measurement model point group to the position of the other, based on the difference in coordinates between the reference model point group and the measurement model point group (coordinate transformation parameter generation step).

[0054] In step S107, the measurement model position estimation unit 35 estimates at least one position of the measurement model using the coordinate transformation parameters (measurement model position estimation step). Once the position of the representative position point RPP1 of the reference model has been determined, it can be said that the destination of the representative position point RPP when moved based on the transformation matrix, which is the coordinate transformation parameter CTP, is the representative position point MPP1 of the measurement model MM1. This is because the alternative targets ATG1 and ATG2, which are the basis of the reference model RM1, have shape features that can be considered equivalent, and it can be said that the measurement model point group MMP1, which is a point group corresponding to the reference model point group RMP1 of the reference model RM1, was detected from the point group obtained by scanning the shape features of the alternative target ATG2.

[0055] Information on the representative position points of the measurement model estimated by the measurement model position estimation unit 35 can be stored in association with the measurement point cloud data including the representative position points.

[0056] The accuracy of the positions of the measurement model point group MMP1 estimated in this way will be explained in the examples described later.

[0057] Once the required number of representative position points MPP of the measurement model have been obtained in this manner, in the step following step S107, the point cloud combination unit 36 ​​combines multiple point clouds measured from different positions using the position information of the acquired measurement model (point cloud combination step).

[0058] When there are multiple measurement point cloud data measured from different instrument points, if there is a common measurement model representative position point, the point cloud combination unit 36 ​​performs point cloud registration of the multiple measurement point cloud data measured from different instrument points by using this as a corresponding point or reference point for calculating the relative positions between the instrument points, as equivalent to the central position coordinate of the dedicated target.

[0059] This allows the instrument points of the group measurement device to be identified using a substitute target instead of a dedicated target, and the point clouds can be combined through registration, thereby reducing the cost of registration.

[0060] This concludes the description of the embodiments of the present disclosure.

[0061] <Example> Hereinafter, examples of the present disclosure will be described. The position measurement method according to the embodiment of the present disclosure was evaluated by the following method. (Evaluation conditions) The coordinate accuracy was verified by calculating the error between the coordinates of the representative position points calculated using the coordinate change parameters obtained by ICP matching the reference model and the measurement model and the true coordinates of the locations corresponding to the representative position points on the measurement model. A 700mm triangular cone and a 20t class hydraulic excavator were used as substitute targets. A TLS was used as the point cloud measurement device. A TS was used as the surveying device for verification. The true coordinates of the reference points were measured using a TS and a dedicated black and white checker target. The substitute targets and the equipment used for measurement were initially positioned as shown in Figures 6 and 7. The distance from the point cloud measurement device to the substitute target ATG1, which is a cone, was 2.5 m in the X direction (distance direction), and the distance to the substitute target ATG, which is a shovel, was 10 m in the X direction (distance direction).

[0062] (Evaluation Procedure) (1) As shown in Figures 7 and 8, we measured point clouds of various alternative targets installed in positions that met theoretically ideal conditions for obtaining reference model point clouds suitable for ICP matching. (2) The alternative targets included in the measurement point cloud were extracted and used as the “reference model.” (3) The "representative position point" on the reference model was determined. (4) A point cloud corresponding to a substitute target included in the measurement point cloud installed at an arbitrary position was extracted and used as a "measurement model." In this case, cones were placed at 5 m intervals, from 5 to 50 m in the distance direction (X direction) from the instrument point (10-time measurement model extraction). Shovels were placed at 10 m intervals, from 10 to 60 m in the distance direction (X direction) from the instrument point (6-time placement, 6-time measurement model extraction). (5) The reference model and the measurement model were aligned using ICP matching, and the coordinate transformation parameters were calculated. (6) The coordinate transformation parameters obtained by ICP matching were used to transform the coordinates of the representative position points of the reference model. (7) ICP matching was performed on the representative position points that were transformed into corresponding coordinates for each measurement point group. (8) The coordinate transformation parameters obtained by ICP matching were used to register the measurement point clouds. (9) The center position coordinates of the dedicated target were measured using TS's non-prism measurement. (10) Using the dedicated target, TS, arranged as shown in Figures 7 and 8, the coordinates of the representative position points of the alternative targets ATG1 and ATG2, whose separation distance was changed in the distance direction (X direction), were measured and used as the true coordinates. (11) In addition, point cloud measurements were performed using TLS on alternative targets ATG1 and ATG2, with the separation distance changed in the distance direction (X direction). (12) The TS coordinate system was transformed into the TLS coordinate system using the central position coordinates of the dedicated target measured in (10). (13) We generated a reference model and a measurement model for ICP matching. (14) ICP matching was performed for the reference model and the measurement model. (15) The error between the coordinates of the reference model representative position points transformed using the coordinate transformation parameters obtained by ICP matching and the true coordinates measured in (10) was calculated.

[0063] (Evaluation results) Figure 9 shows the results of the evaluation of the coordinate accuracy of the representative position points of the surrogate targets obtained using the above evaluation method. This table shows the error from the true value in each direction for each measurement. Generally, for coordinate measurements within 50 m, the measurement error must be within 50 mm. In measurements using a cone as the surrogate target ATG1, the errors in the distance direction (X direction) and the orthogonal direction (Y direction) were approximately 10 mm or less, but the error in the height direction (Z direction) was up to 30 mm compared to the true value. Furthermore, in measurements using a shovel as the surrogate target ATG2, the errors in the distance direction, orthogonal direction, and height direction were all within 10 mm, with particularly high accuracy in the height direction. Regardless of the surrogate target used, the measurement error was within 50 mm in coordinate measurements within 50 m, and the surrogate targets can be used as targets for point cloud merging.

[0064] Regarding the above evaluation results, it is believed that the horizontal accuracy was high because the triangular cone has an axisymmetric shape in the height direction. Also, it is believed that the hydraulic excavator had stable coordinate accuracy because it was high above the ground and had a large surface area.

[0065] An example of the configuration of this embodiment is as follows. [1] A method for measuring the position of a substitute target that substitutes for a dedicated target for combining a plurality of measurement point clouds measured from different positions using a point cloud measurement device, comprising: a measurement point cloud measurement step in which a point cloud measurement unit capable of measuring a point cloud to which coordinates are assigned as relative position information with respect to the instrument point where the measurement was performed measures a measurement point cloud including the alternative target; a measurement model generation step in which a point cloud matching unit compares the reference model point cloud with the measurement point cloud using a reference model including a reference model point cloud that has been generated in advance and has been assigned coordinates that serve as relative position information representing shape features of the alternative target, and extracts one or more measurement model point clouds that correspond to the reference model point cloud from the measurement point cloud, thereby generating one or more measurement models; a coordinate transformation parameter generation step in which a coordinate transformation parameter generation unit generates, based on a difference in coordinates between the reference model point cloud and the measurement model point cloud, a coordinate transformation parameter representing an amount of movement required to move one of the reference model point cloud or the measurement model point cloud to the position of the other; a measurement model position estimating step in which a measurement model position estimating unit estimates at least one position of the measurement model using the coordinate transformation parameters. [2] [1] The method for measuring a target position according to the present invention, a reference model measurement step in which, at least before the measurement model generation step, the point cloud measurement device measures a point cloud for generating a reference model including the substitute target; a reference model generation step in which a reference model generation unit uses the reference model generation point cloud to generate a reference model including a reference model point cloud consisting of a plurality of point clouds representing the alternative target. [3] [2] The method for determining the position of a substitute target according to [2], a reference model representative position point determination step in which, after at least the reference model generation step, a representative position point determiner determines a reference model representative position point whose relative position with respect to the reference model satisfies a predetermined condition; a measurement model position estimation unit that estimates at least one measurement model representative position point whose relative position with respect to the measurement model satisfies a predetermined condition using the reference model representative position point and the coordinate transformation parameters in the measurement model position estimation step; [4] A method for measuring the position of a substitute target described in any one of [1] to [3], wherein the substitute target is any work-related tool that is normally present at a work site but is not normally used as a survey target. [5] A method for measuring the position of an alternative target, comprising a point cloud combining step in which a point cloud combining unit combines multiple measurement point clouds measured from different positions using position information of one or more measurement models measured using the method for measuring the position of an alternative target described in any one of [1] to [3]. [6] A point cloud measurement device that combines multiple measurement point clouds using an alternative target that replaces a dedicated target, a point cloud measurement unit capable of measuring a point cloud to which coordinates are assigned as relative position information with respect to the instrument point where the measurement was performed; a point cloud matching unit that uses a reference model including a reference model point cloud that has been generated in advance and has been assigned coordinates that serve as relative position information representing the shape features of the substitute target, compares the reference model point cloud with a measurement point cloud measured by the point cloud measurement unit, extracts one or more measurement model point clouds that correspond to the reference model point cloud from the measurement point cloud, and generates one or more measurement models; a coordinate transformation parameter generation unit that generates, based on the coordinates of the reference model point cloud and the measurement model point cloud, coordinate transformation parameters that represent the amount of movement required to move one of the reference model point cloud or the measurement model point cloud to the position of the other; a measurement model position estimation unit that estimates at least one position of the measurement model using the coordinate transformation parameters; a point cloud combining unit that combines a plurality of measurement point clouds measured from different positions using position information of the one or more measurement models. [7] The point cloud measurement device described in [6] further includes a reference model generation unit that generates a reference model including a reference model point cloud consisting of a plurality of point clouds representing the alternative target using a point cloud for generating a reference model measured by the point cloud measurement unit. [8] A point cloud information processing device that combines a plurality of measurement point clouds using an alternative target that substitutes for a dedicated target, a point cloud matching unit that uses a reference model including a reference model point cloud that has been generated in advance and has been assigned coordinates that serve as relative position information representing the shape characteristics of the alternative target, compares the reference model point cloud with a measurement point cloud measured by a point cloud measurement device that is capable of measuring a point cloud that has been assigned coordinates that serve as relative position information with respect to the instrument point that performed the measurement, extracts one or more measurement model point clouds that correspond to the reference model point cloud from the measurement point cloud, and generates one or more measurement models; a coordinate transformation parameter generation unit that generates, based on the coordinates of the reference model point cloud and the measurement model point cloud, coordinate transformation parameters that represent the amount of movement required to move one of the reference model point cloud or the measurement model point cloud to the position of the other; a measurement model position estimation unit that estimates at least one position of the measurement model using the coordinate transformation parameters; a point cloud combining unit that combines a plurality of measurement point clouds measured from different positions using position information of the one or more measurement models. [9] The point cloud information processing device described in [8] further includes a reference model generation unit that generates a reference model including a reference model point cloud consisting of a plurality of point clouds representing the alternative target using a point cloud for generating a reference model measured by the point cloud measurement device.

[10] A medium storing a position measurement program for a substitute target that substitutes a dedicated target for combining a plurality of measurement point clouds measured from different positions using a point cloud measurement device using a computer, a measurement model generation step in which a point cloud matching unit uses a reference model including a reference model point cloud that has been generated in advance and to which coordinates serving as relative position information representing shape features of the substitute target are assigned, compares the reference model point cloud with a measurement point cloud including the substitute target measured by a point cloud measurement device, and extracts one or more measurement model point clouds that correspond to the reference model point cloud from the measurement point cloud, thereby generating one or more measurement models; a coordinate transformation parameter generation step in which a coordinate transformation parameter generation unit generates, based on a difference in coordinates between the reference model point cloud and the measurement model point cloud, a coordinate transformation parameter representing an amount of movement required to move one of the reference model point cloud or the measurement model point cloud to the position of the other; A medium storing a substitute target position measurement program for causing a computer to execute a measurement model position estimation step in which a measurement model position estimation unit estimates at least one position of the measurement model using the coordinate transformation parameters. [Explanation of symbols]

[0066] 1. Point cloud measurement system 10 Point cloud information processing device 11 Input section 12 Output section 13 Storage section 14 Communications Department 20 Point cloud measurement device 21 Imaging unit 22 Point cloud measurement unit 23 GNSS 24 Communications Department 30 Processing section 31 Reference model generation unit 32 Representative position point determination section 33 Point Cloud Matching Section 34 Coordinate transformation parameter generation unit 35 Measurement model position estimation unit 36 Point group joint TG dedicated target ATG Alternate Target RM Reference Model RPM reference model point cloud measurement RMP Reference Model Point Cloud RPP Reference Model Representative Position Point MM Measurement Model MPM measurement model point cloud measurement MMP Measurement Model Point Cloud MPP Measurement model representative position point CTP coordinate transformation parameters

Claims

1. A method for measuring the position of a substitute target that substitutes for a dedicated target for combining a plurality of measurement point clouds measured from different positions using a point cloud measurement device, comprising: a measurement point cloud measurement step in which a point cloud measurement unit capable of measuring a point cloud to which coordinates are assigned as relative position information with respect to the instrument point where the measurement was performed measures a measurement point cloud including an alternative target; a measurement model generation step in which a point cloud matching unit compares the reference model point cloud with the measurement point cloud using a reference model including a reference model point cloud that has been generated in advance and has been assigned coordinates that serve as relative position information representing shape features of the alternative target, and extracts one or more measurement model point clouds that correspond to the reference model point cloud from the measurement point cloud, thereby generating one or more measurement models; a coordinate transformation parameter generation step in which a coordinate transformation parameter generation unit generates, based on a difference in coordinates between the reference model point cloud and the measurement model point cloud, a coordinate transformation parameter representing an amount of movement required to move one of the reference model point cloud or the measurement model point cloud to the position of the other; a measurement model position estimating step in which a measurement model position estimating unit estimates at least one position of the measurement model using the coordinate transformation parameters.

2. 2. The method for measuring a target position according to claim 1, a reference model measurement step in which, at least before the measurement model generation step, the point cloud measurement device measures a point cloud for generating a reference model including the substitute target; a reference model generation step in which a reference model generation unit uses the reference model generation point cloud to generate a reference model including a reference model point cloud consisting of a plurality of point clouds representing the alternative target.

3. 3. The method for determining the position of a substitute target according to claim 2, a reference model representative position point determination step in which, after at least the reference model generation step, a representative position point determiner determines a reference model representative position point whose relative position with respect to the reference model satisfies a predetermined condition; a measurement model position estimation unit that estimates at least one measurement model representative position point whose relative position with respect to the measurement model satisfies a predetermined condition using the reference model representative position point and the coordinate transformation parameters in the measurement model position estimation step;

4. 4. The method for measuring the position of a surrogate target according to claim 1, wherein the surrogate target is any work-related tool that is normally present at a work site and is not normally used as a survey target.

5. A method for measuring the position of an alternative target, comprising a point cloud combining step in which a point cloud combining unit combines multiple measurement point clouds measured from different positions using position information of one or more measurement models measured using the method for measuring the position of an alternative target described in any one of claims 1 to 3.

6. A point cloud measurement device that combines multiple measurement point clouds using an alternative target that replaces a dedicated target, a point cloud measurement unit capable of measuring a point cloud to which coordinates are assigned as relative position information with respect to the instrument point where the measurement was performed; a point cloud matching unit that uses a reference model including a reference model point cloud that has been generated in advance and has been assigned coordinates that serve as relative position information representing the shape features of the substitute target, compares the reference model point cloud with a measurement point cloud measured by the point cloud measurement unit, extracts one or more measurement model point clouds that correspond to the reference model point cloud from the measurement point cloud, and generates one or more measurement models; a coordinate transformation parameter generation unit that generates, based on the coordinates of the reference model point cloud and the measurement model point cloud, coordinate transformation parameters that represent the amount of movement required to move one of the reference model point cloud or the measurement model point cloud to the position of the other; a measurement model position estimation unit that estimates at least one position of the measurement model using the coordinate transformation parameters; a point cloud combining unit that combines a plurality of measurement point clouds measured from different positions using position information of the one or more measurement models.

7. The point cloud measurement device according to claim 6, further comprising a reference model generation unit that generates a reference model including a reference model point cloud consisting of a plurality of point clouds representing the alternative target using the point cloud for generating the reference model measured by the point cloud measurement unit.

8. A point cloud information processing device that combines a plurality of measurement point clouds using an alternative target that substitutes for a dedicated target, a point cloud matching unit that uses a reference model including a reference model point cloud that has been generated in advance and has been assigned coordinates that serve as relative position information representing the shape characteristics of the alternative target, compares the reference model point cloud with a measurement point cloud measured by a point cloud measurement device that is capable of measuring a point cloud that has been assigned coordinates that serve as relative position information with respect to the instrument point that performed the measurement, extracts one or more measurement model point clouds that correspond to the reference model point cloud from the measurement point cloud, and generates one or more measurement models; a coordinate transformation parameter generation unit that generates, based on the coordinates of the reference model point cloud and the measurement model point cloud, coordinate transformation parameters that represent the amount of movement required to move one of the reference model point cloud or the measurement model point cloud to the position of the other; a measurement model position estimation unit that estimates at least one position of the measurement model using the coordinate transformation parameters; a point cloud combining unit that combines a plurality of measurement point clouds measured from different positions using position information of the one or more measurement models.

9. The point cloud information processing device according to claim 8, further comprising a reference model generation unit that generates a reference model including a reference model point cloud consisting of a plurality of point clouds representing the alternative target, using the point cloud for generating the reference model measured by the point cloud measurement device.

10. A medium storing a position measurement program for a substitute target that substitutes a dedicated target for combining a plurality of measurement point clouds measured from different positions using a point cloud measurement device using a computer, a measurement model generation step in which a point cloud matching unit uses a reference model including a reference model point cloud that has been generated in advance and to which coordinates serving as relative position information representing shape features of the substitute target are assigned, compares the reference model point cloud with a measurement point cloud including the substitute target measured by a point cloud measurement device, and extracts one or more measurement model point clouds that correspond to the reference model point cloud from the measurement point cloud, thereby generating one or more measurement models; a coordinate transformation parameter generation step in which a coordinate transformation parameter generation unit generates, based on a difference in coordinates between the reference model point cloud and the measurement model point cloud, a coordinate transformation parameter representing an amount of movement required to move one of the reference model point cloud or the measurement model point cloud to the position of the other; A medium storing a substitute target position measurement program for causing a computer to execute a measurement model position estimation step in which a measurement model position estimation unit estimates at least one position of the measurement model using the coordinate transformation parameters.

Images