Point group data processing device, point group data processing method and point group data processing program

Abstract

To suppress reduction in density of point group data while shortening a time required for laser scanning.(0006)MEANS FOR SOLVING THE PROBLEM: A point group data processing device 200 comprises: a data acquisition unit 201 that acquires point group data on a measurement object obtained by laser scanning, and image data on a photographing image photographing the measurement object; a superimposed data preparation unit 202 that prepares superimposed data superimposing the point group data onto the image data; a TIN mesh preparation unit 203 that prepares an aggregate of an area of a triangle surrounded by adjacent three points in the point group data; an image complexity calculation unit 204 that obtains complexity of an image within the area of the triangle on the basis of the superimposed data; and a point generation unit 205 that, when the complexity is smaller than a predetermined threshold, generates a point inside of the area of the triangle on the basis of a point defining the area of the triangle.SELECTED DRAWING: Figure 2

Description

[Technical field]

[0001] The present invention relates to a technique for handling point cloud data. [Background technology]

[0002] A laser scanning device with a camera is known (see, for example, Patent Document 1). Reducing the time required for surveying work is an important technical issue, and from that perspective, there is a demand for reducing the time required for laser scanning. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Patent Publication No. 2021-56069 Summary of the Invention [Problem to be solved by the invention]

[0004] For the same scan area, shortening the laser scan time will result in a sparser (coarser) scan.

[0005] In this context, an object of the present invention is to reduce the time required for laser scanning while suppressing a decrease in density of point cloud data. [Means for solving the problem]

[0006] The present invention is a point cloud data processing device that includes a data acquisition unit that acquires point cloud data of a measurement object obtained by laser scanning and image data of a captured image of the measurement object, a superposition data creation unit that creates superposition data by superimposing the point cloud data and the image data, a microregion aggregate creation unit that creates an aggregate of microregions between adjacent points in the point cloud data, an image complexity calculation unit that calculates the image complexity within the microregion based on the superposition data, and a point generation unit that generates points within the microregion based on points that define the microregion if the complexity is smaller than a predetermined threshold.

[0007] The present invention is a point cloud data processing method that acquires point cloud data of a measurement object obtained by laser scanning and image data of an image taken of the measurement object, creates superimposed data by superimposing the point cloud data and the image data, creates a collection of micro-areas between adjacent points in the point cloud data, calculates the image complexity within the micro-area based on the superimposed data, and if the complexity is smaller than a predetermined threshold, generates points within the micro-area based on the points that define the micro-area.

[0008] The present invention is a point cloud data processing program that is read and executed by a computer, which causes the computer to acquire point cloud data of a measurement object obtained by laser scanning and image data of an image taken of the measurement object, create superposition data by superimposing the point cloud data and the image data, create a collection of micro-areas between adjacent points in the point cloud data, calculate the image complexity within the micro-area based on the superposition data, and, if the complexity is smaller than a predetermined threshold, generate a point within the micro-area based on the points that define the micro-area. Effect of the Invention

[0009] According to the present invention, it is possible to reduce the time required for laser scanning while suppressing a decrease in density of point cloud data. [Brief description of the drawings]

[0010] [Figure 1] FIG. 1 is a perspective view of a laser scanner. [Diagram 2] FIG. 2 is a block diagram of a point cloud data processing device. [Diagram 3] FIG. 1 is a diagram illustrating an example of a TIN mesh. [Figure 4] FIG. 13 is an image diagram of generated points. [Diagram 5] 11 is a flowchart illustrating an example of a processing procedure. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0011] 1. First embodiment (Laser scanner with camera) 1 shows the appearance of a laser scanner (laser scanning device) 100 equipped with a camera 106. The laser scanner 100 includes a tripod 101 serving as a leg, a base unit 102 fixed to the upper part of the tripod 101, a horizontal rotation unit 103 serving as a rotating body capable of horizontal rotation (rotation about a vertical axis) on the base unit 102, and a vertical rotation unit 104 serving as a rotating body capable of vertical rotation (rotation about an axis extended in the horizontal direction) relative to the horizontal rotation unit 103. An operation panel (not shown) is provided on the back of the horizontal rotation unit 103.

[0012] 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. A smartphone or a tablet computer can also be used as the operation terminal.

[0013] The vertical rotation unit 104 includes an optical unit 105, which is an optical system that emits and receives laser scanning light. Inside the vertical rotation unit 104, a mirror that rotates vertically together with the vertical rotation unit 104 is disposed.

[0014] A light-emitting unit equipped with a light-emitting element that emits pulses of scanning light, and a light-receiving unit equipped with a light-receiving element are arranged inside the horizontal rotation unit 103. The scanning light emitted from the light-emitting unit is reflected by the 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 from the measurement object travels a reverse path to the emitted light and enters the optical unit 105, where it is separated by an optical system that separates the emitted light and the incident light (detection light) and received by the light-receiving unit.

[0015] As the vertical rotation unit 104 rotates vertically, the optical unit 105 rotates vertically. Laser scanning light is pulsed from the optical unit 105. This pulsed 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 rotates. In this case, the laser scanning light is pulsed from the optical unit 105 along the vertical angle direction (angle direction of the elevation angle and depression angle).

[0016] Vertical scanning is performed by emitting pulses of the above-mentioned 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.

[0017] By performing horizontal rotation at the same time as the vertical scan, the horizontal rotation unit 103 rotates horizontally, so that the vertical scan line moves so as to be shifted along the horizontal angle direction (horizontal direction). Note that if horizontal rotation is performed at the same time as vertical rotation, the vertical scan will not be completely along the vertical angle direction, but will be a slightly oblique line, even if it is only slightly.

[0018] Motors are used to rotate the horizontal rotation unit 103 and the vertical rotation unit 104. 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.

[0019] Each laser scanning light is a single pulse of distance measuring light, and a single 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 according to 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 distance measurement value and the irradiation direction of the laser scanning light.

[0020] Point cloud data is composed of position data of a large number of scan points. The form of point cloud data includes a form in which distance and direction data related to each point (each scan point) is output. It is also possible to calculate the position of each point in a specific coordinate system inside the laser scanner 100 and output the three-dimensional coordinate position of each point as point cloud data. In addition, the point cloud data obtained by laser scanning also includes information on the luminance (detected intensity of reflected light) of each scan point.

[0021] A camera 106 is disposed on the horizontal rotation unit 103. The camera 106 is a digital still camera that captures still images. The relationship between the position of the optical origin (center of the camera) of the camera 106 and the position of the optical origin of the laser scanner 100, and 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.

[0022] 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 use a configuration in which a moving image is taken with the camera 106 and frame images constituting the moving image are used. It is also possible to use a full-circle camera.

[0023] (Point cloud data processing device) 2 is a block diagram of a point cloud data processing device 200 that processes point cloud data acquired by the laser scanner 100. The point cloud data processing device 200 may be built into the laser scanner 100, or may be provided as a separate component.

[0024] The point cloud data processing device 200 is a computer, and is realized by executing an operating program for realizing the illustrated functional units by the CPU of the computer. A form in which some or all of the functional units in FIG. 2 are configured with dedicated hardware is possible. The point cloud data processing device 200 may be built into the laser scanner 100, or may be prepared as a separate housing. A form in which some or all of the functions of the point cloud data processing device 200 are realized by using a data processing server is also possible.

[0025] The point cloud data processing device 200 includes a data acquisition unit 201 , a superimposition data creation unit 202 , a TIN mesh creation unit 203 , an image complexity calculation unit 204 , and a point generation unit 205 .

[0026] The data acquisition unit 201 receives the point cloud data acquired by the laser scanner 100 and image data of the image captured by the camera 106. The superimposition data creation unit 202 superimposes the point cloud data acquired by the laser scanner 100 and the image data of the image captured by the camera 106, and generates data of a superimposed image (a composite image of the point cloud and the captured image).

[0027] The superimposition is performed with the corresponding positional relationship aligned. The relationship between the position and orientation of the optical system performing the laser scan of the laser scanner 100 and the optical system of the camera 106 is known, and it is possible to determine where in the shooting range of the camera 106 the points included in the point cloud data correspond. The above superimposition is performed with this corresponding positional relationship aligned. In other words, the above superimposition is performed by superimposing the point cloud and the shot image with the viewpoint position and viewing direction aligned.

[0028] The TIN mesh creation unit 203 creates a TIN mesh using three adjacent scan points in the superimposed image. TIN (triangulated irregular network) is data that represents the surface of an object as a set of triangles, and is also called an irregular triangulated network, an irregular triangular network, or an irregular triangular network. Figure 3 is an image diagram of an example of a TIN mesh in a superimposed image.

[0029] The TIN mesh creation unit 203 is an example of a micro-area aggregate creation unit that creates an aggregate of micro-areas between adjacent points in point cloud data. Also, a TIN is an example of an aggregate of micro-areas between adjacent points in point cloud data of the present invention.

[0030] The image complexity calculation unit 204 calculates the image complexity within a unit triangle in the TIN mesh (one of the many unit triangles that make up the TIN mesh: an example of a small area). A captured image of an object with a complex surface structure has a high image complexity. A smooth surface, etc. has a low image complexity.

[0031] Here, the feature points of the image within a unit triangle are extracted, and the variation in the distribution of these feature points is evaluated using the standard deviation to quantitatively determine the complexity (degree of complexity). Specifically, the standard deviation of the variation in the distance between adjacent feature points is calculated, and the complexity of the image is quantitatively evaluated based on this. The distribution and variation of brightness can also be used as parameters for evaluating the complexity. There is also a method of quantitatively evaluating the distribution of feature points and brightness using the fractal dimension to determine the complexity of the image.

[0032] In FIG. 3, the first unit triangular region is a region where the image complexity is relatively low, and the second unit triangular region is a region where the image complexity is relatively high.

[0033] The point generator 205 generates points based on the three scan points (three vertices) that define a unit triangle within the unit triangle whose complexity is equal to or less than a predetermined threshold value. In this example, a surface including the three vertices of the unit triangle is obtained, and points are generated so that the points are distributed at specific intervals on this surface.

[0034] An image of the generated points is shown in Figure 4. Figure 4 shows a case where points are generated in a first group of unit triangles with low complexity, and points are not generated in a second group of unit triangles with high complexity.

[0035] (An example of a processing procedure) 5 is a flowchart showing an example of a processing procedure. An operation program for executing the processing of FIG.

[0036] Prior to the processing, point cloud data is acquired using the laser scanner 100, and an image of the target of the laser scan is captured to obtain point cloud data and image data.

[0037] When the process starts, first, image data corresponding to the point cloud data obtained by the laser scanner 100 is obtained (step S101). Next, the point cloud obtained in step S101 is superimposed on an image to obtain a superimposed image (step S102). Next, a TIN mesh is created using the scan points in the superimposed image obtained in step S102 (step S103). An example of a TIN mesh is shown in FIG. 3.

[0038] After creating the TIN mesh, the image complexity in each unit triangle constituting the TIN mesh is calculated (step S104). Next, points are generated in the unit triangles whose complexity is equal to or less than a threshold (step S105).

[0039] (Advantages) In this embodiment, a point cloud data processing device 200 is used, which includes a data acquisition unit 201 that acquires point cloud data of a measurement object obtained by laser scanning and image data of an image captured of the measurement object, a superposition data creation unit 202 that creates superposition data by superimposing the point cloud data and the image data, a TIN mesh creation unit 203 that creates a collection of triangular areas surrounded by three adjacent points in the point cloud data, an image complexity calculation unit 204 that calculates the image complexity within the triangular area based on the superposition data, and a point generation unit 205 that generates points within the triangular area based on the points that define the triangular area if the complexity is smaller than a predetermined threshold.

[0040] According to the point cloud data processing device 200, it is possible to reduce the time required for laser scanning while suppressing the decrease in density of point cloud data. That is, even if the speed of laser scanning is increased and the scan density becomes sparse, if the image within a unit triangle constituting the TIN mesh is not complex (for example, a flat surface, etc.), the area is regarded as a flat surface, and new (additional) points are generated based on the scan points of the three points (three vertices) constituting the unit triangle, and the inside of the valley triangle is filled with the points. This makes it possible to suppress the decrease in density of point cloud data.

[0041] 2. Second embodiment In the first embodiment, points are not generated in areas where the image complexity is determined to be high from the captured image. Therefore, in this embodiment, re-laser scanning is performed on the areas determined to be highly complex. The re-laser scanning is performed under conditions that result in a higher point cloud density compared to the initial laser scanning. By performing re-scanning, the density of the laser scan points in areas determined to be highly complex from the captured image can be increased, and a point cloud that matches the complexity of the target can be obtained.

[0042] 3. Third embodiment It is also possible to focus on two adjacent scan points, evaluate the complexity of the image in the area between those two points, and if the complexity of that area is low, to interpolate a new point on the line connecting those two points in that area. It is also possible to perform the same processing on the image in the area surrounded by four adjacent scan points.

[0043] 4. Fourth embodiment A total station with a scanning function can also be used as the laser scanning device. In this case, the total station is equipped with a camera that captures an object to be measured. The relationship between the position and attitude of the optical system of this camera and the position and attitude of the optical system of the measurement laser light is known, and the position of the reflection point of the measurement laser light in the image captured by the camera can be determined. Therefore, the measurement point by the measurement laser light (the reflection point of the laser distance measuring light) can be superimposed on the captured image captured by the camera, and the superimposed image (superimposed data) described in the first embodiment can be obtained.

[0044] In this example, the non-prism laser positioning function of the total station is used to perform laser scanning by measuring the positions of multiple measurement points in a mesh (matrix) pattern, and point cloud data is obtained. In addition, the range of the laser scan is photographed by the camera, and photographed image data is obtained.

[0045] The obtained point cloud data and image data of the captured image are then subjected to the process in the flow shown in Figure 5. As a result, even if the density of the laser scan is made sparse in order to shorten the time required for measurement in laser scanning using a total station, points (scan points) are replenished in areas where the image complexity is determined to be low by image judgment, and a decrease in the density of the point cloud data can be suppressed. [Industrial Applicability]

[0046] The present invention can be used in technologies that handle point cloud data obtained by laser scanning. [Explanation of symbols]

[0047] 100...laser scanner, 101...tripod, 102...base section, 103...horizontal rotation section, 104...vertical rotation section, 105...optical section, 106...camera.

Claims

1. a data acquisition unit that acquires point cloud data of a measurement object obtained by laser scanning and image data of a captured image of the measurement object; a superimposition data creation unit that creates superimposition data by superimposing the point cloud data and the image data; a micro-area aggregate creation unit that creates an aggregate of micro-areas between a plurality of adjacent points in the point cloud data; an image complexity calculation unit that calculates an image complexity within the minute region based on the superimposed data; a point generating unit that generates a point within the micro-region based on a point that defines the micro-region when the complexity is smaller than a predetermined threshold value; A point cloud data processing device comprising:

2. Acquire point cloud data of the measurement object obtained by laser scanning and image data of a captured image of the measurement object, Creating superimposed data by superimposing the point cloud data and the image data; Creating a collection of minute regions between a plurality of adjacent points in the point cloud data; determining a degree of image complexity within the small region based on the superimposed data; A point cloud data processing method for generating points within a micro-region based on points defining the micro-region when the complexity is smaller than a predetermined threshold.

3. A program to be read and executed by a computer, To your computer Acquiring point cloud data of a measurement object obtained by laser scanning and image data of a photographed image of the measurement object; creating superimposed data by superimposing the point cloud data and the image data; creating a collection of minute regions between a plurality of adjacent points in the point cloud data; determining a degree of image complexity within the minute region based on the superimposed data; A point cloud data processing program that generates points within the micro-region based on points that define the micro-region when the complexity is smaller than a predetermined threshold.

Images