Laser scanning method, laser scanning device, and program
The two-stage laser scanning method with a widened light distribution gradient and increased scan density addresses the precision issues in laser scanning by accurately detecting reflective prisms, enhancing measurement accuracy and resolution.
Patent Information
- Application Number
- PCT/JP2024/017211
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2023-05-09
- Filing Date
- 2024-05-09
- Publication Date
- 2024-11-14
AI Technical Summary
Laser scanning techniques face challenges in achieving high precision due to saturation of light receiving elements when reflecting strong light from reflective prisms, leading to reduced distance measurement accuracy.
A two-stage laser scanning method is employed, where the first scan uses a reflective prism with known position, and a second scan is performed with a widened light quantity distribution gradient and higher scan density, using a neutral density filter to reduce light intensity and increase scan points, allowing for accurate detection and calculation of the reflective prism's optical center.
This approach enhances the accuracy of surveying by improving the measurement precision of reflective prisms and reducing errors, enabling reliable acquisition of point cloud data with increased resolution and effective range.
Smart Images

Figure JP2024017211_14112024_PF_FP_ABST
Abstract
Description
Laser scanning method, laser scanning device, and program
[0001] The present invention relates to laser scanning.
[0002] A surveying technique using a laser scanner is known (see, for example, Patent Document 1).
[0003] Japanese Patent Application Laid-Open No. 2022-152543
[0004] In surveying using a laser scanner, a laser scan is performed at the surveying site, and then, as post-processing, the point cloud data (laser scan data) obtained by the laser scan is used to create three-dimensional data or convert it into drawing data.
[0005] This technique requires obtaining laser scan data with high accuracy. In this context, the present invention aims to provide a technique for improving the accuracy of surveying work using laser scanning.
[0006] The present invention is a laser scanning method that performs a first laser scan using a laser scanner over an area that includes a reflecting prism whose position is known, detects the reflecting prism based on scan data obtained by the first laser scan, and performs a second laser scan over the reflecting prism with a wider gradient portion of the light intensity distribution of the scanning light compared to the first laser scan.
[0007] In one embodiment of the present invention, a second laser scan is performed on the reflecting prism at a higher scan density than in the first laser scan. In another embodiment of the present invention, the second laser scan is performed under conditions in which a plurality of scanning beams are reflected by the reflecting prism, and the direction of the optical center of the reflecting prism as seen from the laser scanner is calculated by calculating the center of gravity of the direction of the reflecting prism as seen from the laser scanner, weighted by the reflection intensity of the scanning beams.
[0008] In the present invention, the laser scan light in the second laser scan may be obtained by passing the first laser scan light through an optical aperture. In the present invention, the laser scan light in the second laser scan may have a beam shape in which the peak of the light amount distribution is lowered and the beam base is wider in the beam diameter direction than the laser scan light in the first laser scan.
[0009] In one embodiment of the present invention, the laser scanning light in the second laser scan is such that the reflected light from the reflecting prism can be detected when the optical axis of the laser scanning light is positioned off the reflecting surface of the reflecting prism. In another embodiment of the present invention, the laser scanning light is set to a condition where five or more reflection points of the laser scanning light on the reflecting prism are obtained in a direction in which the gradient portion of the light intensity distribution of the laser scanning light in the second laser scan is widened at the maximum expected distance from the laser scanner to the reflecting prism.
[0010] In one embodiment of the present invention, at the assumed minimum distance from the laser scanner to the reflecting prism, the beam diameter of the scanning light of the second laser scan is larger than the effective prism diameter of the reflecting prism, the effective prism diameter being half the objective diameter of the laser scanner, and the beam diameter of the scanning light is defined as a range of light amounts in a direction perpendicular to the optical axis within which effective reflected light from a reflector having the same reflectivity as the reflecting prism at the minimum distance can be obtained. In another embodiment of the present invention, the beam cross section of the laser scanning light in the second laser scan as viewed from the direction of the optical axis has a longitudinal shape, and the spacing of the laser scanning light in a direction perpendicular to the longitudinal direction of the longitudinal shape of the laser scanning light at the reflecting prism is narrower than the spacing of the laser scanning light in the longitudinal direction.
[0011] In the present invention, an example of an embodiment is to obtain a captured image of the target of the first laser scan, create a depth map that predicts distance information from the capture viewpoint of the target in the captured image, create a point cloud image that is an image drawn with points when the point cloud obtained by the first laser scan is viewed from the viewpoint of the laser scan, and compare the distribution of the predicted distances in the depth map with the distribution of distance measurements of each point in the point cloud image, thereby detecting points with abnormalities in distance information.
[0012] The present invention is a program that is read and executed by a computer, which causes the computer to perform a first laser scan using a laser scanner over an area that includes a reflecting prism whose position is known, detect the reflecting prism based on scan data obtained by the first laser scan, and perform a second laser scan over the reflecting prism in a state in which the gradient portion of the light intensity distribution of the scanning light is wider than in the case of the first laser scan.
[0013] The present invention is a laser scanning device that includes a target detection unit that detects a reflecting prism based on scan data obtained by a first laser scan using a laser scanner over an area that includes a reflecting prism whose position is known, and a scan condition setting unit that sets conditions for a second laser scan of the reflecting prism in a state in which the gradient portion of the light intensity distribution of the scanning light is wider than in the case of the first laser scan.
[0014] The present invention is a laser scanning method that performs a first laser scan using a laser scanner, captures an image of the target of the first laser scan, creates a depth map that predicts distance information from the shooting viewpoint of the target in the captured image, creates a point cloud image that is an image drawn with points when the point cloud obtained by the first laser scan is viewed from the viewpoint of the laser scan, and detects points with abnormalities in distance information by comparing the distribution of the predicted distances in the depth map with the distribution of distance measurements of each point in the point cloud image.
[0015] The present invention is a program that is read and executed by a computer, and causes the computer to acquire data of a first laser scan using a laser scanner, acquire data of a captured image of the target of the first laser scan, create a depth map that predicts distance information from the shooting viewpoint of the target in the captured image, create a point cloud image that is an image drawn with points when the point cloud obtained by the first laser scan is viewed from the viewpoint of the laser scan, and detect points with abnormalities in distance information by comparing the distribution of the predicted distances in the depth map with the distribution of distance measurements of each point in the point cloud image.
[0016] According to the present invention, the accuracy of surveying work using laser scanning can be improved.
[0017] It is a conceptual diagram of an embodiment. It is an external view of a laser scanner. It is a block diagram of a laser scanner. It is a block diagram of a calculation device. It is a flowchart showing an example of a processing procedure. It is a diagram showing the light amount distribution of a scanning light. It is a principle diagram explaining the prism effective diameter of a reflecting prism.
[0018] 1. First Embodiment (Overview) Figure 1 shows an overview of a surveying site. Here, a case where laser scanning is performed to obtain three-dimensional data of the terrain will be described. The object of surveying is not particularly limited, and may be the target of civil engineering work or its results, or the exterior or interior of a building or structure. Figure 1 shows a laser scanner 100, which is a device for performing laser scanning, reflecting prisms 200 and 300, which serve as targets for providing reference points, a worker (user) 150 operating the laser scanner 100, and a tablet computer 400 carried by the worker 150.
[0019] The reflecting prisms 200 and 300 are optical reflective targets for surveying that reflect incident light by changing its direction by 180°. FIG. 1 shows an example of a total reflection prism as the reflecting prisms 200 and 300. The reflecting prism may be configured to accommodate incident light from a specific direction. In this example, two or more reflecting prisms are used. Here, the case where two reflecting prisms are used is shown. As the target, in addition to a reflecting prism, an optical reflective target with high reflectivity and retroreflective properties can be used.
[0020] The tablet computer 400 is used as a computing device that performs various processes based on the point cloud data acquired by the laser scanner 100. The functions of the tablet computer 400 can also be performed by a PC (personal computer), a data processing server, or dedicated hardware. It is also possible for the laser scanner 100 itself to have the functions of the tablet computer 400.
[0021] In this example, the positions of the reflecting prisms 200 and 300 in an absolute coordinate system or a specific local coordinate system are known, and a first laser scan (full-circumference scan) is performed on the entire circumference in a state where the position and orientation of the laser scanner 100 are unknown. The full-circumference scan is performed under conditions for obtaining point cloud data of the measurement target, but under these conditions, the reflected light from the reflecting prisms 200 and 300 is too strong, causing saturation of the light receiving elements of the laser scanner 100 and resulting in a decrease in the distance measurement accuracy of the reflecting prisms 200 and 300.
[0022] Therefore, a second laser scan (target scan) is performed with a neutral density filter inserted in the optical path to obtain point cloud data with guaranteed distance measurement accuracy for the reflecting prisms 200 and 300. The reflecting prisms 200 and 300 are identified based on the detected intensity of the reflected light and / or the captured image.
[0023] Then, based on the point cloud data of the reflecting prisms 200, 300 obtained in the second laser scan, the directions of the optical centers of the reflecting prisms 200, 300 relative to the laser scanner 100 are calculated. Here, for multiple reflection points on the reflecting prisms 200, 300, the direction of each reflection point as seen from the laser scanner 100 is weighted by the reflection intensity, and the center of gravity is calculated. In this way, the directions of the optical centers of the reflecting prisms 200, 300 as seen from the optical origin of the laser scanner 100 are statistically determined. This processing is performed inside the laser scanner 100.
[0024] Data on the direction and distance from the laser scanner 100 to each point obtained by the full-circle scan, data on the direction and distance from the laser scanner 100 to the optical center positions of the reflecting prisms 200 and 300 obtained by the target scan, and identification information of the reflecting prisms 200 and 300 are sent from the laser scanner 100 to the tablet computer 400.
[0025] The following processing is performed in the tablet computer 400. Since the positions of the reflecting prisms 200 and 300 in the coordinate system used here are known, the position and orientation of the laser scanner 100 in that coordinate system are determined by resection based on data on the direction and distance from the laser scanner 100 to the optical center positions of the reflecting prisms 200 and 300. Once the position and orientation of the laser scanner 100 in that coordinate system are determined, the coordinates in that coordinate system of each point obtained by the full-circle scan can be determined. This processing is performed in the tablet computer 400.
[0026] The above is the basic operation. In addition to this, in this embodiment, the tablet computer 400 performs the following processes: determining whether the point cloud data sent from the laser scanner 100 to the tablet computer 400 includes an area that requires another laser scan; notifying the operator 150 of the result of this determination; and determining the conditions for a rescan to resolve the abnormality.
[0027] 2 shows the appearance of a laser scanner (laser scanning device) 100. The laser scanner 100 includes a tripod 101 serving as a leg, a base 102 fixed to the top of the tripod 101, a horizontal rotation unit 103 which is a rotating body capable of horizontal rotation (rotation about a vertical axis) on the base 102, and a vertical rotation unit 104 which is a rotating body capable of vertical rotation (rotation about an axis extending horizontally) relative to the horizontal rotation unit 103. An operation panel (not shown) is also provided on the back of the horizontal rotation unit 103.
[0028] The laser scanner 100 can also be operated using a terminal (remote control). In this case, wireless communication is performed between the terminal and the laser scanner 100. It is also possible to use the tablet computer 400 shown in FIG. 1 as the operation terminal.
[0029] The vertical rotation unit 104 includes an optical unit 105, which is an optical system that emits and receives laser scanning light. Inside the horizontal rotation unit 103, there are arranged a light-emitting unit (reference numeral 112 in FIG. 3 , which will be described later) that includes a light-emitting element that emits pulses of scanning light, a light-receiving unit (reference numeral 113 in FIG. 3 , which will be described later) that includes a light-receiving element, and an optical system that separates and combines the optical axes of the emitted light and incident light (detection light). The scanning light emitted from the light-emitting unit passes through the optical system, is reflected by a mirror (not shown) that rotates vertically together with the vertical rotation unit 104, and is emitted to the outside from the optical unit 105. The scanning light reflected by the measurement object enters the optical unit 105, travels the opposite path to the emitted light, is separated by the optical system, and is received by the light-receiving unit.
[0030] As the vertical rotation unit 104 rotates vertically, the optical unit 105 also rotates vertically. Laser scanning light is emitted in pulses from the optical unit 105. This pulse emission occurs along a direction (vertical plane) perpendicular to the rotation axis (axis extending horizontally) of the vertical rotation unit 104 while the vertical rotation unit 104 is rotating. In this case, the laser scanning light is emitted in pulses from the optical unit 105 along the vertical angle direction (angle direction of the elevation angle and depression angle).
[0031] Vertical scanning is performed by emitting pulses of the laser scanning light while vertically rotating the vertical rotation unit 104. While horizontally rotating the horizontal rotation unit 103 and vertically rotating the vertical rotation unit 104, pulses of laser scanning light are emitted from the optical unit 105, and the reflected light from the object is received by the optical unit 105, thereby performing laser scanning on the measurement object.
[0032] Simultaneously with the vertical scan, the horizontal rotation unit 103 rotates horizontally, causing the vertical scan line to move so as to be offset along the horizontal angle direction (horizontal direction). If horizontal rotation is also performed during vertical rotation, the vertical scan will not be perfectly aligned with the vertical angle direction, but will instead be a slightly oblique line. The horizontal rotation unit 103 and the vertical rotation unit 104 are rotated by motors. The horizontal rotation angle of the horizontal rotation unit 103 and the vertical rotation angle of the vertical rotation unit 104 are precisely measured by encoders. By measuring these horizontal rotation angles and vertical rotation angles, the direction of the optical axis of the scan light can be measured.
[0033] Each laser scanning light is a single beam of pulsed distance measurement light, and each laser scanning light measures the distance to the scanning point, which is the reflection point where the laser scanning light hits. This distance measurement is performed based on the well-known principle of optical distance measurement. The position of the scanning point (the reflection point of the laser scanning light) relative to the laser scanner 100 is calculated from this measured distance value and the irradiation direction of the laser scanning light.
[0034] Point cloud data is composed of position data of multiple scan points. One form of point cloud data is to output distance and direction data for each point (each scan point). It is also possible to calculate the position of each point in a specific coordinate system within the laser scanner 100 and output the three-dimensional coordinate position of each point as point cloud data. Furthermore, the point cloud data obtained by laser scanning also includes information on the brightness (detected intensity of reflected light) of each scan point.
[0035] A camera 106 is disposed on the horizontal rotation unit 103. The camera 106 is a digital still camera that captures still color images (RGB images). The relationship between the optical origin (camera center) of the camera 106 and the optical origin of the laser scanner 100, as well as the relationship between the direction of the optical axis of the camera 106 and the direction of the optical axis of the laser scanner 100, are known.
[0036] By repeatedly and continuously taking still images with the camera 106 while rotating the horizontal rotation unit 103, a full-circle image (a 360° panoramic image) can be obtained. It is also possible to obtain a panoramic image of a specific range. It is also possible to take a moving image with the camera 106 and use the frame images that make up the moving image.
[0037] (Block Diagram of Laser Scanner) Figure 3 is a block diagram of the laser scanner 100. The laser scanner 100 includes an operation content accepting unit 111, a light emitting unit 112, a light receiving unit 113, a distance measuring unit 114, a direction acquiring unit 115, a point cloud data generating unit 116, a horizontal rotation drive control unit 117, a vertical rotation drive control unit 118, a light emission timing control unit 119, a target detecting unit 120, a scan condition setting unit 121, a target direction calculating unit 123, a communication device 124, a memory unit 125, and a camera 106. The functional units denoted by reference numerals 111 and 114 to 125 are realized by a computer included in the laser scanner 100. It is also possible to configure some or all of these functional units using dedicated hardware.
[0038] The operation content receiving unit 111 is configured with an operation panel (for example, a touch panel or a touch panel display). A user operates the operation content receiving unit 111 to operate the laser scanner 100 and perform various settings. It is also possible for the operation content to be received using a terminal or a smartphone.
[0039] The light emitting unit 112 has a light emitting element for emitting laser scanning light, an optical system, and a drive system for the light emitting element. The light receiving unit 113 has a light receiving element for receiving the laser scanning light reflected from the target, an optical system, and a circuit for handling the detection signal output from the light receiving element.
[0040] The distance measurement unit 114 calculates the distance from the optical origin of the laser scanner 100 to the reflection point of the laser scanning light based on the principle of optical distance measurement. This is the same principle as that of a normal laser rangefinder. The direction acquisition unit 115 acquires the direction of the optical axis of each laser scanning light from the horizontal rotation angle of the horizontal rotation unit 103 and the vertical rotation angle of the vertical rotation unit 104.
[0041] The point cloud data generator 116 generates point cloud data, which is data describing the position of each scan point in a coordinate system with the position of the laser scanner 100 as the origin. For example, the point cloud data is composed of a group of data on the direction and distance from the laser scanner 100 for each scan point. The point cloud data also associates the detected intensity of reflected light from each scan point.
[0042] The horizontal rotation drive control unit 117 controls the horizontal rotation drive of the horizontal rotation unit 103. The vertical rotation drive control unit 118 controls the vertical rotation drive of the vertical rotation unit 104. The light emission timing control unit 119 controls the light emission timing of the light emitting element of the light emitting unit 112.
[0043] The target detection unit 120 detects the target reflecting prisms 200 and 300 from the laser scan point cloud obtained by the first laser scan, which is the full-circle scan (step 103 in FIG. 5). This detection is performed as follows.
[0044] First, the full-circle scan in step S103 in FIG. 5 is performed to obtain point cloud data of the measurement object. The measurement object may be a terrain, building, road, bridge, etc., and the reflectivity of the scanning light is lower than that of the reflecting prisms 200 and 300. Therefore, the full-circle scan is performed under conditions that allow detection of the reflection of the scanning light from these measurement objects. On the other hand, under these scan conditions, the reflected light from the reflecting prisms 200 and 300 is too strong, saturating the light-receiving elements of the light-receiving unit 113 and reducing distance measurement accuracy.
[0045] Therefore, points (reflection points) where the output of the light receiving element exceeds a preset threshold are detected as points of reflected light from the reflecting prisms 200 and 300. In addition, since the output waveform of the light receiving element takes on a shape due to saturation (the upper end becomes flat and a shape that leaves a tail on the time axis), there is also a method of detecting reflected light from the reflecting prisms 200 and 300 by evaluating this waveform. There is also a method of combining these methods. Reflected light from the reflecting prisms 200 and 300 is detected using these principles.
[0046] Furthermore, when detecting only based on the detection intensity of reflected light, there is a possibility that reflected light from a mirror or shiny metal that is not a reflecting prism 200, 300 may be mistakenly detected as a reflecting prism 200, 300. Therefore, in this example, an image captured by the camera 106 in the direction determined to be reflected light from the reflecting prism 200, 300 is compared with a previously prepared image of the reflecting prism 200, 300 to determine the authenticity of the reflected light determined to be from the reflecting prism 200, 300. This makes it possible to reliably obtain the direction of the reflecting prism 200, 300 as seen from the laser scanner 100.
[0047] There is also a method for identifying the reflected light from the reflecting prisms 200, 300 as follows. Here, the positions of the two reflecting prisms 200, 300 in the absolute coordinate system are known, and the distance L between them is also known. Therefore, two bright points whose detection intensity exceeds a threshold and whose distance between them is L are detected as reflected light from the reflecting prisms 200, 300. The above processing is performed in the target detection unit 120.
[0048] The scan condition setting unit 121 sets the scan conditions for the full-circle scan in step S103 of Fig. 5 and the scan conditions for the target scan in step S105. Note that the target scan in step S105 is performed under conditions that reduce the intensity of the scan light (scan light reflected from the reflection points) received by the light-receiving elements compared to the first full-circle scan (step S103), by using methods such as placing a neutral density filter in front of the light-emitting element, placing a neutral density filter in front of the light-receiving element, lowering the voltage applied to the light-emitting element, or a combination of these methods. This ensures the measurement accuracy of the reflecting prisms 200 and 300.
[0049] The target direction calculation unit 123 calculates the direction of the reflecting prisms 200 and 300 as seen from the laser scanner 100 based on the point cloud data obtained by scanning the target.
[0050] The processing performed by the target direction calculation unit 123 will be described below. First, the direction of the reflecting prisms 200 and 300 as seen from the laser scanner 100 is known through processing in the target detection unit 120. However, the accuracy of this determination is not high. Here, the reflected light reflected from the direction of the reflecting prisms 200 and 300 is extracted from the point cloud data obtained in the target scan. Typically, the scan density is set so that multiple reflection points are generated from one reflecting prism, and multiple beams of light are reflected from each reflecting prism. Therefore, the center of gravity of the angle measurements measured by the laser scanner at multiple points corresponding to these multiple reflected beams is calculated through statistical processing. The angle measurements at each point are weighted by the intensity of the reflected light to determine the center of gravity of the angle measurements. The center of gravity of these angle measurements is obtained as the direction of the optical center of the reflecting prisms 200 and 300 as seen from the laser scanner 100.
[0051] In this way, the directions of the reflecting prisms 200 and 300 relative to the laser scanner 100 are obtained. The directions of the reflecting prisms 200 and 300 obtained here are obtained in a coordinate system (machine coordinate system) with the laser scanner 100 as the origin. The positions of the reflecting prisms 200 and 300 in the absolute coordinate system (global coordinate system) are measured in advance as reference points and are known. Each reference point is assigned a reference coordinate name (e.g., a specific code). This is for identifying the reference points on a drawing (map) of the survey site. The reflecting prisms 200 and 300 are marked with this identification code, and the identification code is detected from the image captured by the camera 106. This identification information, the position information of the reflecting prisms 200 and 300 obtained by the target scan, and the known position information of the reflecting prisms 200 and 300 are associated with each other.
[0052] The above processing is performed in the target direction calculation unit 123.
[0053] The communication device 124 communicates with other devices. Communication is performed using a wireless line such as a wireless LAN. Wired communication is also possible. The storage unit 125 stores various data handled in the operation of the laser scanner 100, operation programs, and obtained point cloud data.
[0054] (Data Processing Device) Figure 4 is a block diagram of a laser scan data processing device (arithmetic device) configured using the tablet computer 400 of Figure 1. The tablet computer 400 implements the following functional units: a point cloud data receiving unit 401, an image data receiving unit 402, a point cloud data adjusting unit 403, a unit for determining whether or not rescanning is necessary 404, a notification unit 405, and a unit for setting rescanning conditions 406. These functional units are implemented by the CPU of the tablet computer 400 executing dedicated application programs. The tablet computer 400 also includes a communication unit 407, a touch panel display 408, and a storage device 409.
[0055] The point cloud data receiving unit 401 receives point cloud data (laser scan point cloud) acquired by the laser scanner 100. This point cloud data is data that associates the direction and distance, measurement time, and received light intensity of reflected light of each point obtained by the first laser scan, the full-circle scan (step S103 in FIG. 5 ), as seen from the laser scanner 100. The point cloud data receiving unit 401 also receives information on the direction of the optical centers of the reflecting prisms 200 and 300 as seen from the laser scanner 100, and identification information for the reflecting prisms 200 and 300, obtained by the second laser scan, the target scan (step S105), and calculation of the target position based on the target scan (step S106).
[0056] The image data receiving unit 402 receives image data of images captured by the camera 106. The point cloud data adjusting unit 403 performs processing on the point cloud data received by the point cloud data receiving unit 401, such as removing noise and replacing point data determined to be abnormal with point data obtained by rescanning.
[0057] The rescanning necessity determination unit 404 determines whether the point cloud data that has undergone the point cloud data adjustment process includes a range that requires rescanning. Examples of ranges that require rescanning include a point cloud range containing points with abnormal light reception intensity (too strong or too weak), a range where there are no (or few) reflection points in an area where reflection points should be found when compared with image data, a point cloud range containing points with abnormal distance measurements (suspected multiple reflections), a range with a low density of points (reflection points), a range where a target has not been detected, a range where the intensity of reflected light from a target is lower than a specified value, a range where a notable event has been found in the point cloud, and a point cloud that satisfies predetermined conditions. These determinations are made using predetermined thresholds.
[0058] The notification unit 405 performs processing to notify whether or not a rescan is required and the details of the cause of the rescan. Specifically, it performs processing to display the details of the notification on the touch panel display 408. The details of the notification are specific details of the above-mentioned determination to the operator 150. For example, the touch panel display 408 of the tablet computer 400 displays messages such as "There is an area where the density of points is insufficient," "There is an area that cannot be measured," or "The target cannot be detected."
[0059] The rescan condition setting unit 406 performs the process of step S112. The rescan conditions in step S112 are set using the position and orientation of the laser scanner 100 in the laser scans of steps S103 and S105. For example, if the received light intensity is too strong, a dimming scan condition is set in which the intensity of light incident on the light receiving element is reduced. Alternatively, if the received light intensity is too weak, a scan condition is set in which the intensity of light incident on the light receiving element is increased (for example, the light emission intensity of the light emitting element is increased). Alternatively, if there are no (or few) reflection points in an area where there should be reflection points, a scan condition is set in which the intensity of light incident on the light receiving element is increased (for example, the light emission intensity of the light emitting element is increased).
[0060] For example, if the distance measurement value is abnormal (multiple reflections are suspected), low-light scanning conditions are set, which reduce the intensity of light incident on the light-receiving element. By reducing the intensity of the received light, the effects of weak multiple reflections can be suppressed. For example, if the density of points (reflection points) is low, scanning conditions are set that increase the density of points on the reflective surface, such as slowing the rotation speed of the horizontal and / or vertical rotation parts of the laser scanner and increasing the emission frequency of the scanning light.
[0061] The communication device 407 communicates with the laser scanner 100 and other devices. Communication is performed using a wireless line. The touch panel display 408 accepts various operations by the worker 150, displays various data, and displays notification information. The storage device 409 stores various data handled by the tablet computer 100, various functions of the tablet computer 100, and programs and data for executing various processes executed by the tablet computer 100.
[0062] (Example of Processing) An example of processing will be described below. First, the premise of the processing will be described. At the surveying site (site where laser scanning is performed) in Figure 1, reflecting prisms 200 and 300 that serve as targets for providing reference points are installed, and their positions (coordinate values) are known in the absolute coordinate system.
[0063] The position and orientation of the laser scanner 100 in the absolute coordinate system are unknown. The laser scanner 100 has a leveling function that is electrically driven. In addition, as an initial setting, either a manual mode in which a process related to rescanning is executed at the instruction of the operator 150 when a data abnormality is detected, or an automatic mode in which a process related to rescanning is executed automatically without the instruction of the operator 150 can be set.
[0064] When the laser scanner 100 is installed at the surveying site and started up, the processing of the flowchart in Fig. 5 is executed. Here, steps S101 to S107 are executed by the laser scanner 100, and steps S108 to S113 are executed by the tablet computer 400. A program for executing the flowchart in Fig. 5 is stored in an appropriate storage medium and executed by the CPU of the computer that controls the operation of the laser scanner 100 and the CPU of the tablet computer 300.
[0065] First, a leveling process is performed (step S101). In the leveling process, an operation is performed to ensure the horizontality of the electrically driven laser scanner 100. Next, an image of the entire circumference of the laser scanner 100 is taken (step S102).
[0066] In this 360° photographing, a panoramic image is obtained by intermittently photographing with the camera 106 while rotating the horizontal rotation unit 103 360°. This 360° photographing may be performed simultaneously with the next 360° scan (step S103). Note that if the range of the measurement target is not 360° but is limited (for example, a range of 120°), photographing is performed within that range (photographing of the measurement target range).
[0067] Next, a full-circle scan is performed (step S103). The full-circle scan is performed under conditions for acquiring point cloud data of the measurement target. When a specific range is the measurement target, point cloud data for that range is adopted from the point cloud data acquired by the full-circle scan. Laser scanning limited to a specific range is also possible.
[0068] Next, the scanning reflected light from the targets (reflecting prisms 200, 300) is detected from the point cloud data obtained by the full-circle scan (step S104). This process is performed by the target detection unit 120. In this process, the reflected light from the reflecting prisms 200, 300 is identified by utilizing the fact that the intensity of the reflected light from the reflecting prisms 200, 300 is particularly high.
[0069] Once the reflecting prisms 200 and 300 are detected, a laser scan (target scan) (step S105) is performed in which the detection intensity of the scanning light (the intensity of the reflected light incident on the light receiving element) is reduced. In the target scan (step S105), a neutral density filter is placed in front of the light receiving element or the light emitting element, thereby reducing the intensity of the light received by the light receiving element compared to the full-circle scan (step S103).
[0070] This is because, under the conditions of the full-circle scan in step S103, the intensity of the reflected light from the reflecting prisms 200 and 300 is too strong, causing saturation in the light-receiving elements and reducing the accuracy of the distance measurement data from the reflecting prisms 200 and 300. By reducing the light-receiving intensity in the light-receiving elements compared to the full-circle scan (step S103), saturation in the light-receiving elements is suppressed, and the accuracy of the distance measurement data from the reflecting prisms 200 and 300 can be ensured.
[0071] There are several methods for determining the degree to which the light incident on the light receiving element is reduced. The first method is to determine the degree of light reduction (attenuation amount) in advance. The second method is to determine the degree of light reduction according to the intensity of the light reflected from the reflecting prism detected in step S104. In this case, a neutral density filter capable of varying the amount of attenuation is used.
[0072] The target scan (step S105) allows for highly accurate measurement of the positions of the target reflecting prisms 200 and 300. The results of the target scan are used to calculate the position and orientation of the laser scanner 100 by resection.
[0073] In the target scan (step S105), a laser scan is performed on the entire circumference to extract only the point cloud data of the target (reflecting prisms 200, 300) obtained in step S104, or a laser scan is performed limited to the direction of the reflecting prisms 200, 300 obtained in step S104.
[0074] Furthermore, in the target scan (step S105), scanning conditions are adopted in which the scan density is increased and the gradient portion of the light intensity distribution of the scanning light is made larger compared to the full-circle scan in step S103. The scan density is defined as the number of scan points per unit area.
[0075] First, the increase in scan density will be described. Here, the emission frequency of the scan light is not changed, and the horizontal rotation speed of the horizontal rotation unit 103 and the vertical rotation speed of the vertical rotation unit 104 are made slower than in step S103, thereby increasing the scan density in step S105 compared to the scan density in step S103. Note that the scan density can also be changed by adjusting the emission frequency of the scan light.
[0076] Here, the above conditions are set so that the number of scan points (reflection points of the scan light) in the reflecting prisms (targets) 200, 300 detected in step S104 is 5 to 7 or more in the direction of the spread of the Gaussian distribution of the scan light (the direction in which the tail of the beam widens).
[0077] A specific example of setting the scan density will be described below. Here, it is assumed that the maximum distance from the laser scanner to the reflecting prism is 100 m. Here, the horizontal rotation speed of the horizontal rotation unit 103 and the vertical rotation speed of the vertical rotation unit 104 are set so that the reflected light from the reflecting prism of the prism scanning light at a distance of 100 m will have 5 to 7 or more points in the direction of spread of the Gaussian distribution.
[0078] For example, if the cross section of the beam is circular, the horizontal rotation speed of the horizontal rotation unit 103 and the vertical rotation speed of the vertical rotation unit 104 are set so that 5 x 5 points of reflection are obtained from the reflecting prism at a distance of 100 m.
[0079] The number 5 has the following meaning. To obtain accuracy in the calculation of the center of gravity, which will be described later, the distribution of the scanning reflected light from the reflecting prism (distribution of points) needs to have enough points to draw the outline of a Gaussian distribution. To meet this condition, three points in the direction of the spread of the Gaussian distribution of the beam (the direction in which the tail widens) is the limit, and five or more points are required. For this reason, the condition of five to seven points mentioned above is adopted.
[0080] In general, the scan density in the target scan (step S105) set by the above method is greater than the scan density in the full-circle scan in step S103. Therefore, the horizontal rotation speed of the horizontal rotation unit 103 and the vertical rotation speed of the vertical rotation unit 104 in the laser scan in step S105 are slower than those in the laser scan in step S103. Note that the maximum distance from the laser scanner to the reflecting prism is set to 100 m, but this maximum value can also be set under other conditions, such as 50 m or 70 m.
[0081] The optical center position (hereinafter referred to as the optical center) of a reflecting prism is set as a design factor. Although a reflecting prism has a specific size, the optical axis of the scanning light does not necessarily coincide with this optical center. Therefore, the scanning conditions are adjusted so that multiple scanning light beams are incident on the reflecting prism, resulting in a large number of reflection points. The reflection points are weighted by their reflection intensities to calculate the center of gravity of the multiple reflection points as seen from the laser scanner. This allows the direction of the optical center of the reflecting prism as seen from the laser scanner to be determined with the highest possible accuracy. To improve this measurement accuracy, the scan density is set in the target scan (step S105). In other words, by ensuring a large number of scanning points on the reflecting prism, the accuracy of the calculation of the center of gravity of the reflecting prism as seen from the laser scanner is improved.
[0082] Next, we will explain how to widen the gradient portion of the light intensity distribution. In this embodiment, in addition to improving the angle measurement accuracy of the reflecting prism by increasing the number of scan points as described above, the gradient portion of the light intensity distribution in the direction perpendicular to the optical axis of the scanning light is widened (expanded) compared to the full-circle scan in step S103.
[0083] An example of the light intensity distribution of the scanning light is shown in Figure 6. The vertical axis in Figure 6 represents the light intensity value normalized by the peak value, and the horizontal axis represents the relative value of the beam diameter captured at the corresponding angle direction (for example, vertical angle or horizontal angle). Note that the line passing through the light emission center (or light reception center) and the point of maximum value of the light intensity distribution is the optical axis of the scanning light.
[0084] (A) is the light intensity distribution before passing through the optical aperture. (B) is the light intensity distribution after passing through the optical aperture. When the optical aperture is a linear slit, the light intensity distribution shown in Figure 6 changes in the direction perpendicular to the extension direction of the slit. When the optical aperture is a pinhole (circular hole), the light intensity distribution shown in Figure 6 changes in all directions.
[0085] The change in the light intensity distribution in Figure 6 is caused by the diffraction phenomenon when the light passes through the optical aperture. As shown in Figure 6(B), the gradient of the light intensity distribution becomes wider after passing through the optical aperture compared to before passing through. In other words, the base of the mountain-shaped beam shape becomes gentler, and the base becomes wider in the radial direction.
[0086] 6A and 6B are shown at the same peak position because they are normalized by the peak value of the light amount, but the actual peak value of the light amount of the beam waveform (B) is lower than that of the beam waveform (A) because the base is wider. Therefore, the laser scanning light shown in (B) has a lower peak of the light amount distribution than the laser scanning light shown in (A), and the base of the beam is wider in the beam diameter direction.
[0087] The gradient portion of the light quantity distribution is a portion where the light quantity in the beam diameter direction monotonically decreases or increases. The gradient portion can be defined as a region where the change in light quantity monotonically decreases or increases in the range of 90% to 5% of the peak value.
[0088] The degree of beam spread in (B) is set so that (effective prism diameter) < (beam diameter) at the minimum distance assumed during prism scanning. The minimum distance is the minimum value of the measurement range determined by the specifications of the laser scanner as a product. For laser scanners used for surveying, a minimum distance of several meters is generally adopted.
[0089] Here, the prism effective diameter is determined by the principle shown in FIG. 7 . Note that the objective diameter in FIG. 7 is the effective diameter of the objective lens of the light receiving system of the laser scanner 100. The beam diameter is the width of the light intensity distribution that can be effectively detected as reflected light when a reflecting prism is the target. In the case of FIG. 6 , for the beam (B), the above beam diameter is the range of the light intensity distribution in the beam diameter direction (direction perpendicular to the optical axis) where effective reflected light (detectable reflected light) is obtained from a reflector having the same reflectivity as the reflecting prism at the assumed minimum distance from the reflecting prism. In other words, the above beam diameter is defined as the range of light intensity in the direction perpendicular to the optical axis where effective reflected light is obtained from a reflector having the same reflectivity as the reflecting prism at the assumed minimum distance.
[0090] The optical diaphragm is arranged inside the horizontal rotation unit 103 in a state in which insertion or non-insertion can be selected at a position on the optical axis in front of the light-emitting unit 112. The optical diaphragm is not used in step S103, and is inserted on the optical axis of the scanning light in front of the light-emitting unit in step S105.
[0091] Based on the target scan in step S105, the direction from the laser scanner 100 to the optical center of the reflecting prism is calculated based on the multiple reflected lights from the reflecting prism. In this case, instead of treating the multiple reflected lights from the reflecting prism equally, the angle measurement values of the reflection points on the reflecting prism are weighted by the intensity of the reflected lights to determine the center of gravity. This improves the measurement position of the direction of the optical center of the reflecting prism as seen from the laser scanner 100.
[0092] Specifically, the following statistical processing calculations are performed. Here, the direction of the optical center of the reflecting prism as viewed from the laser scanner, as determined by the statistical calculations, is defined as (H0, V0). H is the horizontal angle of the horizontal rotation unit 103, and V is the vertical angle of the vertical rotation unit 104 (the angle in the direction of elevation or depression). Furthermore, the measurement direction (angle measurement value) of the scan point as viewed from the laser scanner based on the scan light reflected by the reflecting prism is defined as (Hi, Vi) (i = 1, 2, 3, 4, ...), and the received light intensity of each scan light is defined as Ii. In this case, (H0, V0) is calculated using H0 = Σ(Ii × Hi) / ΣIi and V0 = Σ(Ii × Vi) / ΣIi. Here, Ii is the weighting coefficient.
[0093] The scanning light beam with a broadened beam tail shown in Fig. 6B has scanning light components that are reflected from the reflecting prism at the tail, even at the same scanning speed, so the number of effective reflection points from the reflecting prism increases compared to when the scanning light beam shown in Fig. 6A is used. Note that, because the reflecting prism has high reflectivity, it is possible to obtain effective reflected light that can be detected as a point cloud even at the tail of the beam.
[0094] When beam (B) is used, the detected intensity of the reflected light tends to be stronger (larger) as the optical axis of the scanning light is closer to the optical center of the reflecting prism, and weaker (smaller) as the optical axis of the scanning light is farther from the optical center of the reflecting prism. This is because the farther the optical axis of the scanning light is from the optical center of the reflecting prism, the larger the tail of the beam that does not enter the reflecting prism.
[0095] Therefore, coupled with the effect of increasing the number of scan points described above, the weighting described above functions effectively, improving the calculation resolution and calculation accuracy of the optical center of the reflecting prism.
[0096] 6A, the scanning light having the light intensity distribution shown in FIG. 6A is used to perform the laser scan (target scan) on the reflecting prism in step S105. In this case, the width of the gradient portion of the light intensity distribution is narrow, so the number of scan points is reduced, and the contribution of the reflected light from the tail of the beam to the calculation of the center of gravity described above is small, resulting in a decrease in statistical accuracy compared to when the beam shown in FIG. 6B is used.
[0097] Note that the scanning light with a wider gradient in the light intensity distribution (beam in FIG. 6B) has a wider base due to the use of the diffraction phenomenon in the optical aperture, resulting in a larger diameter for the scan points (reflection points). Therefore, when used for the full-circle scan in step S103, the resolution decreases. Furthermore, since the density of the light intensity decreases, effective reflected light cannot be obtained for targets that are not highly efficient reflectors such as reflecting prisms. This results in disadvantages such as missing points and a reduced effective range. In this sense, the scanning light with a wider gradient in the light intensity distribution shown in (B) is not suitable for the full-circle scan in step S103.
[0098] Therefore, in this embodiment, in the laser scanning of step S103 to obtain point cloud data of the scan target, priority is given to range, reliable acquisition of the target point cloud, and resolution, and scanning light having a light intensity distribution with a relatively narrow gradient portion of the light intensity distribution shown in Figure 6(A) is used. On the other hand, in the target scanning of step S105 to determine the direction of the optical center of the reflecting prism as seen from the laser scanner by calculating the center of gravity, scanning light having a relatively wide gradient portion of the light intensity distribution shown in Figure 6(B) is used to pursue accuracy in calculating the center of gravity related to the direction of the target (reflecting prism).
[0099] In the above example, one type of optical diaphragm is used to expand the gradient portion of the light intensity distribution of the scanning light. However, multiple optical diaphragms with different diaphragm diameters or an optical diaphragm with a variable diaphragm diameter may be prepared. In this case, the diaphragm diameter of the optical diaphragm is selected according to the distance to the reflecting prism obtained in step S103, and the size (width) of the gradient portion of the light intensity distribution according to that distance is selected.
[0100] Specifically, the process is as follows: First, an appropriate relationship between the distance L from the laser scanner to the reflecting prism and the magnitude of the gradient portion of the light intensity distribution of the scanning light is obtained in advance. Then, in step S103, the distance L to the reflecting prism is obtained. Then, in step S105, the above-mentioned "appropriate relationship between the distance L and the magnitude of the gradient portion of the light intensity distribution of the scanning light" is referenced, and an aperture diameter of the optical aperture that provides the optimal "magnitude of the gradient portion of the light intensity distribution" at the distance L is selected.
[0101] After step S105, the process proceeds to step S106. In step S106, the accurate direction (direction of the optical center) of the reflecting prisms 200 and 300 as seen from the laser scanner 100 is calculated based on the point cloud data of the reflecting prisms 200 and 300 obtained in step S105 (step S106). In this process, the direction of the optical center of the reflecting prism from the laser scanner 100 is calculated by calculating the center of gravity related to the measurement angles of the multiple reflected lights described above. This process is performed by the target direction calculation unit 123. At this time, identification information of the reflecting prisms 200 and 300 is also obtained based on the image captured by the camera 106.
[0102] Next, the image data obtained in step S102 and the point cloud data obtained in step S103 at this point are transmitted to the tablet computer 400 (step S107). If a rescan is performed, the point cloud data obtained by the rescan is transmitted to the tablet computer 400. In addition to the image data and point cloud data, the position information and identification information of the reflecting prisms 200 and 300, which are targets for providing reference points, are also transmitted to the tablet computer 400.
[0103] The tablet computer 400 that has received the transmitted data performs the following processes. First, it performs a point cloud data adjustment process (step S108). In the point cloud data adjustment process in step S108, noise is removed and point data determined to be abnormal is replaced with point data obtained by rescanning. If rescanning has not been performed, only noise removal is performed in step S108.
[0104] Next, it is determined whether or not rescanning is required for the point cloud data that has been subjected to the point cloud data adjustment processing (step S109). This processing is performed by the rescanning requirement determination unit 404 in FIG.
[0105] If rescanning is necessary, the process proceeds from step S109 to step S110. If rescanning is not necessary, a notification process (step S114) is performed to notify the operator 150 that "rescanning is not necessary." In this case, the content of the notification is displayed on the touch panel display 408 of the tablet computer 400. This process is performed by the notification unit 405.
[0106] In step S110, a notification regarding the need for rescanning is made. Here, the notification regarding the need for rescanning is displayed on the touch panel display 408 of the tablet computer 400. This process is performed by the notification unit 405.
[0107] If a rescan command is received from the operator 150 after the notification in step S110, the process proceeds from step S111 to step S112; otherwise, the process ends. Note that if an automatic mode is selected in which a rescan is automatically performed when a data abnormality is detected, the process proceeds from step S111 to step S112.
[0108] In step S112, rescanning conditions corresponding to the content determined in step S109 are set. This process is performed by rescanning condition setting unit 406 in FIG. 4. The conditions for the rescanning are set using the position and orientation of laser scanner 100 in the full-circle scanning in step S103. In other words, the position and orientation of laser scanner 100 used in setting the rescanning conditions are the same as the position and orientation of laser scanner 100 used in the full-circle scanning in step S103.
[0109] After the scan conditions for the rescan (rescan conditions) have been set, the rescan conditions are sent to the laser scanner 100 (step S113). The laser scanner 100, which has received the rescan conditions, performs another laser scan (rescan) using the rescan conditions. The rescan may not be a full-circle scan, but may be a scan of a specific limited range. If the target has already been detected, steps S104 to S106 during the rescan may be omitted. The rescan may be set to be performed multiple times, not just once.
[0110] When a rescan is performed, in step S108, which is performed after the rescan, the range of points that need to be rescanned and are included in the point cloud obtained by the circumferential scan (step S103) is replaced with points newly obtained by the rescan. The position and orientation information of the laser scanner 100 is the same during the circumferential scan and the rescan. Therefore, the replacement of the points is easy.
[0111] 5 includes point cloud data, data on the detection intensity of each point, data on the direction of the reflecting prisms 200 and 300 from the laser scanner 100 obtained by target scanning, identification data, and measurement time. At this stage, the position of the laser scanner 100 is unknown, and the position of the point cloud data obtained by circumferential scanning in the absolute coordinate system is unknown (the coordinate system of the point cloud data obtained by circumferential scanning is a machine point coordinate system with the laser scanner 100 as the origin).
[0112] Therefore, the position and orientation of the laser scanner 100 in the absolute coordinate system are determined by resection based on the known positions of the reflecting prisms 200 and 300 in the absolute coordinate system and the directions of the reflecting prisms 200 and 300 from the laser scanner 100 obtained by target scanning. As a result, the coordinate values in the absolute coordinate system of each point of the point cloud data obtained by the laser scanner 100 are determined.
[0113] (Advantages) By adopting scanning conditions suitable for target scanning during target scanning in step S105, the measurement accuracy of the direction of the target (reflecting prism) as seen from the laser scanner obtained in step S106 is improved, which in turn increases the positional accuracy in the absolute coordinate system of the point cloud data obtained in step S103.
[0114] That is, by improving the accuracy of the directions of the reflecting prisms 200 and 300 that provide the reference position from the laser scanner 100, the accuracy of calculating the position and orientation of the laser scanner 100 by the resection using the reference position improves. This improves the accuracy of the coordinate conversion of the point cloud data obtained in step S103 into an absolute coordinate system.
[0115] Furthermore, after the laser scan is completed, the scan results are immediately checked and the check results are fed back to the worker 150. This reduces the time and effort required for surveying using the laser scanner 100. For example, if there is a problem with the obtained point cloud data, the rescanning conditions to resolve the problem are automatically set, making the rescanning process more efficient. Furthermore, this rescanning makes it possible to eliminate point cloud data with low accuracy, thereby increasing the accuracy of the point cloud data ultimately obtained.
[0116] The position and orientation of the laser scanner 100 used in setting the rescanning conditions are the same as those used in the full-circle scanning in step S103. This eliminates the need for a registration process (matching process) when identifying the correspondence between the point cloud data obtained by the full-circle scanning in step S103 and the point cloud data obtained by the rescanning. This makes it easy to replace abnormal points and reduces the occurrence of errors during this process.
[0117] 2. Second Embodiment Before step S105, the image data obtained in step S102 and the point cloud data obtained in step S103 may be transmitted to the tablet computer 400, and the processing from step S108 onward may be performed before step S105. In this case, the processing can be made more efficient. Furthermore, because a determination as to whether rescanning is required is reliably made before the laser scanner 100 is moved, rescanning and data processing can be performed efficiently.
[0118] 5, assume that the work is completed after step S107, the laser scanner 100 is powered off, and the laser scanner 100 is moved. In this case, if a rescan is required, the position and orientation data of the laser scanner 100 must be acquired again during the rescan. Furthermore, the point cloud data acquired by the rescan must be aligned with the point cloud data acquired initially (a process of aligning the coordinate systems). This increases the workload and can cause problems such as errors during the alignment process, making it inefficient. According to this embodiment, step S109 is executed before step S105, thereby preventing the above-mentioned problems from occurring.
[0119] The transmission of the results of the circumferential scan in step S103 may be performed when step S103 is completed, or may be performed simultaneously with the scan while step S103 is being executed. In the latter case, the point cloud data obtained by the circumferential scan can be sent to the tablet computer 400 at an earlier timing.
[0120] In this embodiment, it is also possible to configure the laser scanner 100 so that target scanning cannot be performed until it has been determined whether or not rescanning is necessary for the point cloud data obtained by the full-circle scanning in step S103. In this case, the laser scanner 100 may be moved because it is mistakenly assumed that processing has ended even though there is a data abnormality or the like.
[0121] 5 are the same (or can be considered the same) as the scan conditions for the target scan in step S105, the point cloud data for the target scan in step S105 is requested from the laser scanner 100. In this case, the point cloud data for the target scan in step S105 is used as the data for the rescan. In this case, the effort of rescanning using the laser scanner 100 can be eliminated.
[0122] 4. Fourth Embodiment GNSS is used to measure the approximate position in the absolute coordinate system of the laser scanner 100. This measurement may contain errors, so a simple measurement, for example, using the GNSS function of a smartphone, may be used.
[0123] This measurement determines the approximate positional relationship between the laser scanner 100 and the reflecting prisms 200 and 300, making it possible to identify the reflecting prisms 200 and 300 as viewed from the laser scanner 100.
[0124] 5. Fifth Embodiment An example of a case where a noteworthy phenomenon is found in the point cloud when determining whether or not rescanning is necessary will be described. For example, in the circumferential scan of step S103, there is a region where the positions of many points distributed on a plane are dispersed in a direction perpendicular to the surface. In this case, it is expected that the surface to be scanned is uneven. In such a case, in order to obtain a more detailed view of the unevenness, rescanning is performed under conditions where the scan density is higher than that of the circumferential scan.
[0125] In the above case, the discrete state of the distribution positions of the points is judged using a threshold value, and based on the result of this judgment, a group of points in which a notable event has been detected is extracted.
[0126] 6. Sixth Embodiment There is a technology for detecting defects (cracks, corrosion, peeling, etc.) on concrete surfaces by laser scanning. An example of applying the present invention to this technology will be described. In this case, a determination condition is set in advance so that if a point in a locally depressed portion is detected in the point cloud constituting the smooth surface obtained in the circumferential scan of step S103, that area (range) is determined to be an area requiring rescanning. In addition, the rescan is set to be performed under conditions with a higher scan density than the first circumferential scan (step S103).
[0127] 7. Seventh Embodiment An example will be described in which a point cloud that satisfies a predetermined condition for determining whether or not rescanning is required is determined to be a point cloud to be rescanned. For example, it is possible to set the condition such that only vertical wall surfaces are determined to require rescanning, or only planes facing vertically downward are determined to require rescanning. For example, in laser scanning of a bridge, it may be desired to preferentially scan the vertical surfaces of the bridge piers in detail and with high density. In this case, the determination condition is set so that only vertical surfaces are selected as points to be rescanned in the determination of step S109. In other words, the condition is set in advance so that point clouds facing a specific direction are determined to be points to be rescanned.
[0128] 8. Eighth Embodiment Rather than replacing point cloud data that requires rescanning with point cloud data obtained by rescanning, a process of superimposing the former point cloud data and the latter point cloud data may be performed. The positions of each point in the former point cloud data do not necessarily coincide with the positions of each point in the latter point cloud data. Therefore, by superimposing the two point cloud data, the density of the point clouds can be increased.
[0129] 9. Ninth Embodiment An example of the determination of "whether rescanning is necessary" in step S109 will be described below. Here, a depth camera is used as the camera 106. A depth camera is an image capturing device (or technology thereof) that can acquire depth information of a captured object (distance information from the capturing viewpoint). There are several techniques for acquiring depth information. Here, a depth camera is used that uses a neural network (depth estimation AI model) that utilizes deep learning to predict distance information of an object captured in a captured image. In this method, distance information of the captured object is acquired (predicted) by software processing the captured image.
[0130] A method for obtaining distance information from a captured image will be described below. First, various objects captured in an image captured by the camera 106 are extracted from the captured image. For example, individual objects such as the road surface, walls, roofs, handrails, pipes, pillars, overhead wires, road surface, and vehicles are identified and extracted. This process is performed using image processing software. Furthermore, the images of the extracted objects are classified into types such as the road surface, walls, roofs, handrails, pipes, pillars, overhead wires, road surface, and vehicles.
[0131] Then, depth information (distance information) of each extracted object is acquired. Here, the depth information is information about the distance from the optical origin (projection center) of the camera 106 to the object.
[0132] The images captured by the camera 106 are monocular images. Here, a neural network (depth estimation AI model) using deep learning is used to predict distance information of objects captured in the captured monocular images, and a depth map (distance map) of the captured images is created. In the depth map, distance information (distance from the optical origin of the camera 106) of the objects captured in the captured images is expressed (digitized) using shading and color tones.
[0133] In the depth map, image information relating to each extracted object and the depth information (distance information) described above are integrated to provide depth information (distance information) of each object in the captured image.
[0134] In this embodiment, the depth map is compared with a point cloud image of point cloud data obtained by the laser scanner 100. A point cloud image is an image drawn using points obtained by laser scanning as viewed from the viewpoint of the laser scanner. In addition, in this point cloud image, an image of the points is created in which changes in shading and color tone are added to the image according to distance information for each point. Since the relationship between the position and orientation of the optical system of the laser scanner 100 and the optical system of the camera 106 is known, the correspondence between the depth map and the point cloud image can be determined.
[0135] Here, distance data estimated from an image is compared with distance data obtained by laser scanning. This comparison compares the distribution of distances. Specifically, for a specific area of the surveying target, the distribution of distance information shown in the depth map is compared with the distribution of distance information shown in the point cloud image. That is, for corresponding parts of the depth map and point cloud image, the state of change in perspective shown in the depth map is compared with the state of change in perspective shown in the point cloud image.
[0136] The measured distance values of laser scan points with abnormal distance information due to multiple reflections, etc., are longer than normal distances. Therefore, when comparing the distribution of distances of points in a point cloud image with the distribution of distances of the photographed object in a depth map, differences arise at the above-mentioned abnormal points. By evaluating this difference, the presence or absence of abnormal points in the laser scan point cloud can be determined.
[0137] For example, suppose there are two object parts (object part 1 and object part 2) that are two-dimensionally adjacent when viewed from a viewpoint. These two object parts have a perspective relationship, with object part 1 in the foreground and object part 2 in the background. Here, suppose that the laser scan data for object part 1 contains erroneous distance information due to multiple reflections or the like. In this case, a difference occurs in the perspective relationship between object part 1 and object part 2 in the depth map and the point cloud image. In some cases, the perspective relationship may be reversed.
[0138] By evaluating the difference in perspective, it is possible to determine whether or not there are any points in the point cloud image that have abnormal distance information, i.e., whether or not there are any abnormal points in the laser scan point cloud.
[0139] The result of the above determination is displayed on the display of the tablet computer 400 and notified to the operator 150. The rescanning may be performed automatically, or the operator 150 may perform an operation based on the above notification.
[0140] The position information of the scan points related to the multiple reflections of the scanning light is an erroneous measurement value, and it is desirable to eliminate it by some method. According to this embodiment, by comparing the distance information of the object shown in the depth map created based on the captured image with the distance information of the points actually measured by laser scanning, it is possible to determine whether or not there are any points with abnormal distance information included in the point cloud obtained by laser scanning.
[0141] 10. Tenth Embodiment It is also possible to replace the processing of step S109 with "defective point cloud presence / absence determination." In this case, the part indicated by reference numeral 404 in FIG. 4 becomes a "defective point cloud presence / absence determination unit." For example, even without rescanning, the accuracy of the final point cloud can be ensured by replacing an inappropriate point cloud with a point cloud obtained from a laser scanner installed at another machine point. In this case, rescanning is unnecessary, and the inappropriate point cloud that needs to be replaced with another appropriate point cloud is determined to be a "defective point cloud."
[0142] 11. Eleventh Embodiment In step S105, it is also possible to adopt a mode in which the angular interval of the scanning light in step S103 is not changed, that is, the horizontal rotation speed of the horizontal rotation unit 103 and the vertical rotation speed of the vertical rotation unit 104 are not changed, and only the optical aperture is inserted onto the optical axis.
[0143] In this case, the density of the optical axes of the scanning light on the surface to be scanned at a specific distance remains the same, but the number of reflected light beams from the reflecting prism (the number of reflection points) increases compared to when the conditions of step S103 are adopted. This is because the light reflected by the reflecting prism at the base of the expanded beam is measured as a reflection point. In other words, when a reflecting prism is the target, using the beam in Figure 6 (B) can increase the number of scan points compared to when the beam in (A) is used. In other words, the same effect as when the scan density is essentially increased can be obtained.
[0144] Therefore, if the accuracy of calculating the direction from the laser scanner to the optical center of the reflecting prism by calculating the center of gravity in step S106 can be ensured, it is also possible to adopt in step S105 scanning conditions that increase the gradient portion of the light intensity distribution of the scanning light by using an optical aperture, without changing the angular interval of the scanning light in step S103.
[0145] 12. Twelfth Embodiment An example will be described below in which the beam cross section (cross section viewed from the optical axis direction) of the scanning light used in step S105 has a longitudinal shape such as an ellipse. The shape of the beam cross section of the scanning light may be a longitudinal shape such as an ellipse due to the light emitting element or the optical system of the light emitting unit, or when a slit is used as an optical aperture.
[0146] For example, suppose the reflecting prism is positioned in a roughly horizontal direction as viewed from the laser scanner. The beam cross section of the scanning light in the horizontal direction is elongated (the long side is the V angle (vertical angle) direction), and the gradient portion of the light intensity distribution spreads in this elongated direction. In this case, since the spread of the beam in the horizontal direction (H angle direction) is relatively small, the horizontal rotation speed of the horizontal rotation unit 103 is adjusted (slowed down) to intentionally narrow the spacing of the scanning light in the horizontal direction and increase the density of optical axes per unit horizontal angle.
[0147] This prevents the number of reflection points obtained from the reflecting prism from being uneven in the horizontal and vertical directions. In this mode, the number of reflection points on the reflecting prism in the horizontal direction is ensured by adjusting the spacing between the optical axes of the scanning light, and the number of reflection points on the reflecting prism in the vertical direction is ensured by expanding the gradient portion of the light intensity distribution of the scanning light.
[0148] The opposite case to the above example will now be described. It is assumed that the reflecting prism is arranged in a roughly horizontal direction as viewed from the laser scanner. In this case, the beam cross section of the scanning light in the horizontal direction is horizontally elongated (longitudinal in the horizontal direction), and the gradient portion of the light intensity distribution spreads in this horizontal direction. In this case, since the beam spread in the vertical direction is relatively small, the vertical rotation speed of the vertical rotation unit 104 is adjusted (slowed down) to intentionally narrow the spacing of the scanning light in the vertical angle direction and increase the density of optical axes per unit vertical angle. It is also possible to narrow the spacing of the scanning light in the vertical angle direction by increasing the emission frequency of the scanning light.
[0149] An example of specific scanning conditions will be described below. As described in the first embodiment, the laser scanning in step S105 is preferably performed under conditions in which there are five or more reflection points in the direction in which the gradient portion of the light intensity distribution is widened. However, this is only when the effect of improving the accuracy of statistical processing using the gradient portion of the light intensity distribution can be obtained. In the direction perpendicular to the longitudinal direction of the beam cross section having the above-mentioned longitudinal shape, the gradient portion of the light intensity distribution cannot be expected to be used to much, so it is necessary to narrow the intervals between the scanning beams and further increase the number of reflection points.
[0150] Although it depends on the degree of longitudinal shape of the beam cross section (for example, the ratio of the major axis to the minor axis in the case of an ellipse), as a guideline, it is preferable that the spacing between the scanning beams in the direction perpendicular to the longitudinal direction of the longitudinal beam cross section is at least twice the spacing between the scanning beams in the longitudinal direction. The laser scanning conditions in step S105 are adjusted so as to satisfy this condition.
[0151] Specifically, the rotation speed of the horizontal rotation unit 103, the rotation speed of the vertical rotation unit 104, and the position or positions of the emission frequency of the scanning light are adjusted so as to satisfy the above conditions at the position of the reflecting prism or at the maximum expected distance.
[0152] 2, scanning light incident from the horizontal direction is reflected by a rotating mirror in the vertical rotation unit 104, and laser scanning is performed in the vertical angle direction. In this format, if the beam cross section of the scanning light (cross section viewed from the optical axis direction) is longitudinal (e.g., elliptical), the longitudinal direction of the beam cross section rotates in the vertical angle direction.
[0153] This rotation may affect the calculation of the direction of the reflecting prism in statistical processing. In other words, for a longitudinal beam cross section, the longitudinal direction in which the beam diameter expands is the direction in which the gradient portion of the light intensity distribution of the scanning light is widened, and this direction becomes an effective beam component during statistical calculation. On the other hand, the direction perpendicular to the longitudinal direction has a narrow gradient portion of the light intensity distribution and contributes little to statistical calculation. Therefore, the spacing of the scanning light beams is narrowed in the direction perpendicular to the longitudinal direction of the beam cross section of the scanning light striking the reflecting prism to prevent differences in the effectiveness of statistical calculation depending on the direction. Specifically, if the beam expands in the V direction at the reflecting prism, the spacing of the scanning light beams in the H direction is narrowed. Conversely, if the beam expands in the H direction at the reflecting prism, the spacing of the scanning light beams in the V direction is narrowed.
[0154] An example will be described below. Here, the scanning conditions in step S105 are adjusted in response to the rotation of the beam cross section of the scanning light. It is assumed that the beam cross section of the scanning light is longitudinal (e.g., elliptical). It is also assumed that the relationship between the vertical angle of the optical axis of the scanning light and the longitudinal direction of the beam cross section is known in advance. It is also assumed that the gradient portion of the light intensity distribution is wide in the longitudinal direction.
[0155] First, the approximate direction of the target reflecting prism is obtained from the result of step S103. Next, the longitudinal direction of the cross-sectional shape of the scanning light beam in that direction is obtained. Next, the H angle component (horizontal angle component) and V angle component (vertical angle component) of the longitudinal direction are calculated. Next, the interval between the scanning lights in the H angle direction and the V angle direction are set based on the calculated H angle component and V angle component.
[0156] For example, if the longitudinal contribution of the beam cross section in the H angle direction is large (i.e., the longitudinal beam tends to tilt sideways), the beam component in the V angle direction will be relatively small, so the condition is to narrow the spacing of the scanning light in the V angle direction. Specifically, the rotation speed of the vertical rotation unit 104 is slowed down and / or the emission frequency of the scanning light is increased.
[0157] Conversely, if the longitudinal contribution of the beam cross section in the V-angle direction is large (i.e., the longitudinal beam tends to be vertical), the beam component in the H-angle direction will be relatively small, so the condition is to narrow the spacing of the scanning light in the H-angle direction. Specifically, the rotation speed of the horizontal rotation unit 103 is slowed down and / or the emission frequency of the scanning light is increased.
[0158] This embodiment can be understood as an invention in which a laser scanning method includes a first laser scan using a laser scanner over an area including a plurality of reflective prisms whose positions are known, detection of the plurality of reflective prisms based on scan data obtained by the first laser scan, and a second laser scan of the plurality of reflective prisms in a state in which the gradient portion of the light intensity distribution of the scan light is wider than in the case of the first laser scan, in which the shape of the beam cross section of the scan light in the second laser scan is longitudinal, and the gradient portion of the light intensity distribution is wider in the direction of the longitudinal shape, the direction of the reflective prism as viewed from the laser scanner is obtained based on the first laser scan, the longitudinal direction of the longitudinal shape is obtained based on the direction, and the scanning conditions of the second laser scanner are set based on the longitudinal direction.
[0159] As described above, when the beam cross section of the scanning light has a longitudinal shape, the rotation of the beam cross section of the scanning light during the laser scanning in step S105 may result in a decrease in the accuracy of the calculation of the direction of the reflecting prism as seen by the laser scanner through statistical processing. In this embodiment, this problem is addressed by narrowing the spacing of the scanning light in a direction perpendicular to the longitudinal direction of the beam cross section. That is, when the beam cross section of the scanning light has a longitudinal shape, adjusting the spacing of the scanning light in a direction perpendicular to the longitudinal direction of the longitudinal shape prevents differences in accuracy in the H-angle and V-angle directions when calculating the direction of the reflecting prism through statistical processing. Specific adjustments are made as described at the end of the thirteenth embodiment.
[0160] 14. Fourteenth Embodiment The present invention can also be applied to the mechanical point backsight method, which uses one reflecting prism to provide a reference point. In this case, a laser scanner is installed at the position of known point A as the mechanical point, and a full-circle scan and target scan (prism scan) are performed with the reflecting prism installed at known point B as the scanning target.
[0161] In target scanning, a laser scan is performed on a reflecting prism placed at known point B as the backsight point. The scanning conditions for target scanning are the same as those in the other embodiments. Here, the direction of the reflecting prism as seen from the laser scanner is obtained by the statistical calculation described in the first embodiment using the results of the target scan. Since the position of the laser scanner is known as known point A and the position of the reflecting prism is also known as known point B, the attitude of the laser scanner in the coordinate system used here is calculated.
[0162] By knowing the position and orientation of the laser scanner, coordinates can be assigned to the point cloud data obtained in the circumferential scan. For example, assume that the positions of known points A and B are given in an absolute coordinate system. In this case, the position (known point A) and orientation of the laser scanner in the absolute coordinate system are given, and therefore coordinates in the absolute coordinate system are assigned to the point cloud data obtained in the circumferential scan.
[0163] 15. Fifteenth Embodiment A configuration in which a plurality of the above-described embodiments are combined is also possible.
[0164] 100...laser scanner, 101...tripod, 102...base unit, 103...horizontal rotation unit, 104...vertical rotation unit, 105...optical unit, 106...camera, 150...worker, 200...reflecting prism (target that provides a reference point), 300...reflecting prism (target that provides a reference point), 400...tablet computer.
Claims
1. A laser scanning method comprising: performing a first laser scan by a laser scanner over an area including a reflecting prism whose position is known; detecting the reflecting prism based on scan data obtained by the first laser scan; and performing a second laser scan over the reflecting prism in a state in which the gradient portion of the light quantity distribution of the scanning light is wider than in the first laser scan.
2. A laser scanning method according to claim 1, wherein a second laser scan is performed on the reflecting prism at a higher scan density than in the first laser scan.
3. A laser scanning method as described in claim 1 or 2, wherein the second laser scan is performed under conditions in which multiple scanning beams are reflected by the reflecting prism, and the direction of the optical center of the reflecting prism as seen by the laser scanner is calculated by calculating the center of gravity of the direction of the reflecting prism as seen by the laser scanner weighted by the reflection intensities of the multiple scanning beams.
4. A laser scanning method according to any one of claims 1 to 3, wherein the laser scanning light in the second laser scan is obtained by passing the first laser scanning light through an optical aperture.
5. A laser scanning method according to any one of claims 1 to 4, wherein the laser scanning light in the second laser scan has a beam shape in which the peak of the light quantity distribution is lowered and the beam tail is wider in the beam diameter direction compared to the laser scanning light in the first laser scan.
6. A laser scanning method according to any one of claims 1 to 5, wherein the laser scanning light in the second laser scan is capable of being detected as reflected light from the reflecting prism when the optical axis of the laser scanning light is positioned away from the reflecting surface of the reflecting prism.
7. A laser scanning method as claimed in any one of claims 1 to 6, wherein the condition is set such that, at the maximum expected distance from the laser scanner to the reflecting prism, five or more reflection points of the laser scanning light in the reflecting prism in a direction in which the gradient portion of the light quantity distribution of the laser scanning light in the second laser scan is widened are obtained.
8. A laser scanning method according to any one of claims 1 to 7, wherein, at an assumed minimum distance from the laser scanner to the reflecting prism, the beam diameter of the scanning light of the second laser scan is larger than the effective prism diameter of the reflecting prism, the effective prism diameter being half the objective diameter of the laser scanner, and the beam diameter of the scanning light is defined as a range of light amounts in a direction perpendicular to the optical axis within which effective reflected light is obtained from a reflector having the same reflectance as the reflecting prism at the minimum distance.
9. A laser scanning method according to any one of claims 1 to 8, wherein a beam cross section of the laser scanning light in the second laser scan viewed from the direction of the optical axis has a longitudinal shape, and the spacing of the laser scanning light in a direction perpendicular to the longitudinal direction of the longitudinal shape of the laser scanning light at the reflecting prism is narrower than the spacing of the laser scanning light in the longitudinal direction.
10. A laser scanning method according to any one of claims 1 to 9, comprising: obtaining a captured image of the target of the first laser scan; creating a depth map that predicts distance information from a capturing viewpoint of the target in the captured image; creating a point cloud image that is an image rendered with points of the point cloud obtained by the first laser scan when viewed from the viewpoint of the laser scan; and detecting points with abnormalities in distance information by comparing the distribution of the predicted distances in the depth map with the distribution of distance measurements of each point in the point cloud image.
11. A program to be read and executed by a computer, causing the computer to execute the following: a first laser scan with a laser scanner over an area including a reflecting prism whose position is known; detection of the reflecting prism based on scan data obtained by the first laser scan; and a second laser scan of the reflecting prism in a state in which the gradient portion of the light quantity distribution of the scanning light is wider than in the case of the first laser scan.
12. A laser scanning device comprising: a target detection unit that detects a reflecting prism whose position is known based on scan data obtained by a first laser scan by a laser scanner in an area including the reflecting prism; and a scan condition setting unit that sets conditions for a second laser scan of the reflecting prism in a state in which the gradient portion of the light quantity distribution of the scanning light is wider than in the case of the first laser scan.
13. A laser scanning method comprising: performing a first laser scan using a laser scanner; capturing an image of a target of the first laser scan; creating a depth map that predicts distance information from a viewpoint of capturing the target in the captured image; creating a point cloud image that is an image rendered by points when the point cloud obtained by the first laser scan is viewed from the viewpoint of the laser scan; and detecting points with abnormalities in distance information by comparing the distribution of the predicted distances in the depth map with the distribution of distance measurements for each point in the point cloud image.
14. A program to be read and executed by a computer, causing the computer to perform the following operations: acquiring data of a first laser scan using a laser scanner; acquiring data of images of the subject of the first laser scan; creating a depth map that predicts distance information from the shooting viewpoint of the subject in the captured image; creating a point cloud image that is an image drawn with points when the point cloud obtained by the first laser scan is viewed from the viewpoint of the laser scan; and detecting points with abnormalities in distance information by comparing the distribution of the predicted distances in the depth map with the distribution of distance measurements of each point in the point cloud image.
Citation Information
Patent Citations
Vehicle detection device
JP2013134667A
Survey data processing device, survey data processing method and survey data processing program
JP2022152543A
Laser scanning device, laser scanning method, and program for laser scanning
JP2023041321A
