Point cloud image manipulation system

Abstract

To solve a conventional problem, that is, to provide a point group image operation system that can handle measurement points using three-dimensional coordinate information while visually observing an image representing a three-dimensional point group in two dimensions.SOLUTION: A point group image operation system of the present invention includes measurement point storage means, mapping means, display means, point designation means, and attribute acquisition means. Among these, the attribute acquisition means is means for reading, from the measurement point storage means, attributes of original measurement points corresponding to designated image measurement points. The mapping means arranges original measurement points acquired during one rotation of a laser scanner in a first axis direction so as to be arranged in order of measurement time, and arranges image measurement points arranged in this single axis in a second axis direction perpendicular to the first axis direction so as to be arranged in order of measurement time.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a technology for processing measurement points acquired by a laser scanner, and more specifically to a point cloud image manipulation system that can process measurement points with three-dimensional coordinates (hereinafter referred to as "three-dimensional measurement points") after representing them as two-dimensional images. [Background technology]

[0002] Recently, there has been an increasing demand for topographical information (spatial information), and an increasing number of people are requesting spatial information on facilities, such as their shape and location, in order to more effectively manage and utilize facilities installed on the ground. At the same time, advanced maintenance and management of social infrastructure is positioned as an important issue in realizing Society 5.0, which is currently being promoted by the public and private sectors. Furthermore, as autonomous driving technology becomes more practical, there is a strong demand from many quarters for various spatial information on roads, including road edges (road boundaries).

[0003] Traditionally, two-dimensional (2D) planar drawings (floor plans), such as topographical maps, have been the mainstream for showing spatial information. Although plan views sometimes show "height information" such as contour lines and endpoint elevations, they are primarily focused on showing planar positions, making it difficult to grasp the target area as a three-dimensional (3D) space. However, recent advances in measurement technology have made it easy to obtain a large number of sets of three-dimensional measurement points (hereafter referred to as "3D point clouds"), and advances in information technology have also made it easier to handle these three-dimensional point clouds.

[0004] For example, to obtain a 3D point cloud of terrain, preferred measurement methods include aerial photogrammetry, airborne laser measurement, terrestrial laser measurement, and MMS (Mobile Mapping System). Among these, MMS is a measurement method in which a vehicle equipped with a laser scanner, camera, a satellite positioning system (GNSS: Global Navigation Satellite System) for acquiring its own position, an IMU (Inertial Measurement Unit), an odometry, and other sensors moves around the ground, and while moving, the laser scanner rotates and emits a laser, thereby obtaining measurement points at all points (i.e., a 3D point cloud).

[0005] The 3D point cloud obtained by measuring a measurement target (such as terrain) using an MMS or similar device is generally used as a 3D model (hereinafter referred to as a "3D model"). This 3D model represents the measurement target using 3D coordinates, and is a terrain model represented by a DSM (Digital Surface Model), DEM (Digital Elevation Model), or DTM (Digital Terrain Model).

[0006] Typically, a 3D model is composed of multiple small regions obtained by dividing the target planar area. These small regions, also known as meshes, are formed, for example, by dividing the area into orthogonal grids, and each small region has a representative point. Because the 3D point clouds obtained by measurement are often random data (data with irregular arrangements on a plane), geometric calculations are often used to assign height information to the representative points of the small regions. Examples of geometric calculation methods include the triangulated irregular network (TIN) method, which calculates height using an irregular triangulation network formed from random data; the nearest neighbor method, which uses the nearest laser measurement point; inverse distance weighting (IDW), the Kriging algorithm, and the averaging method. Because 3D models allow for a two-dimensional and three-dimensional understanding of the measurement target, they can be used for a wider variety of purposes than traditional plan views.

[0007] Because 3D models merely present spatial information based on a three-dimensional coordinate system, in most cases, their actual use involves assigning meaning to each feature, such as a facility or building. That is, measurement points are grouped by feature, and attributes such as type (road edge, office building, sign, etc.) and installation date are assigned to each group (feature) before the 3D model is used. Efforts have been made to automatically process these features (hereinafter referred to as "feature data processing"), such as overlaying two-dimensional information like images or topographical maps and automatically extracting them, or, in recent years, processing using artificial intelligence (AI). However, complete processing without misrecognition or missing data is difficult, and the reality is that operators still perform the process manually, visually checking from the beginning (or as a supplement to automatic processing).

[0008] Although 3D models are arranged in three-dimensional space, they are conventionally displayed on a flat (i.e., two-dimensional) surface such as a display screen to be viewed. This means that operators must visually view the 3D model displayed on a flat surface while converting it into feature data. This has led to inconveniences such as difficulty in grasping the depth direction of the display, or difficulty in specifying the desired measurement point due to the large number of measurement points displayed. To address this issue, technologies have been proposed to make it easier for operators to recognize 3D models. For example, Patent Document 1 proposes a technology for converting a 3D model into a two-dimensional image and then handling the 3D point cloud. [Prior art documents] [Patent documents]

[0009] [Patent Document 1] Patent Publication No. 2015-141147 Summary of the Invention [Problem to be solved by the invention]

[0010] Patent Document 1 describes a method for imaging actual 3D measurement points by projecting them onto a predetermined plane, a technology for creating images by reversing the usual photometric procedure. This allows for intuitive understanding of features without determining interior and exterior orientation parameters, i.e., reducing the burden associated with spatial calculation processing. Converting a 3D point cloud into a 2D image and using it in this way is advantageous because it makes it easier for operators to recognize features. However, because the imaged 3D point cloud only contains 2D coordinate information, there is insufficient information for 3D feature data processing.

[0011] The object of the present invention is to solve the conventional problems, that is, to provide a point cloud image manipulation system that allows you to handle measurement points using three-dimensional coordinate information while visually viewing an image that represents a three-dimensional point cloud on a two-dimensional surface. [Means for solving the problem]

[0012] The present invention focuses on the fact that each measurement point acquired during one rotation of the laser scanner is placed on a plane and imaged, and the points in that image are linked to the original three-dimensional measurement points, and is an invention based on an idea that has not been seen before.

[0013] The point cloud image manipulation system of the present invention comprises a measurement point storage means, a mapping means, a display means, a point designation means, and an attribute acquisition means. Among these, the measurement point storage means is a means for storing three-dimensional measurement points (hereinafter referred to as "original measurement points") acquired by a moving laser scanner. The mapping means is a means for arranging the original measurement points read from the measurement point storage means on a plane and generating a two-dimensional image (hereinafter referred to as a "point cloud image") based on the attributes of the original measurement points. For convenience, the points representing the original measurement points in the point cloud image will be referred to as "image measurement points." The point designation means is a means for allowing an operator to designate a desired position in the point cloud image displayed on the display means, and the attribute acquisition means is a means for reading, from the measurement point storage means, the attributes of the original measurement points corresponding to the image measurement points associated with the positions designated by the point designation means. The mapping means arranges multiple original measurement points (hereinafter referred to as "image measurement point set") acquired during one rotation of the laser scanner as image measurement points in the first axis direction so that they are arranged in the order of measurement time, and also arranges this image measurement point set in the second axis direction perpendicular to the first axis direction so that they are arranged in the order of measurement time.

[0014] The point cloud image manipulation system of the present invention may further include a point-to-point distance calculation means for calculating the point-to-point distance (distance in three-dimensional space) between two original measurement points designated by the point designation means, based on the three-dimensional coordinates of the original measurement points read by the measurement point attribute acquisition means.

[0015] The point cloud image manipulation system of the present invention may further include a three-dimensional neighboring point extraction means. This three-dimensional neighboring point extraction means generates line segments in three-dimensional space (hereinafter referred to as "three-dimensional line segments") based on the three-dimensional coordinates of the original measurement points and extracts original measurement points located near these three-dimensional line segments. In this case, when the operator specifies two desired image measurement points ("start point" and "end point") in the point cloud image displayed on the display means using the point specification means, the attribute acquisition means reads the three-dimensional coordinates of the original measurement points corresponding to these start point and end point from the measurement point storage means. Based on the read three-dimensional coordinates of these original measurement points, the three-dimensional neighboring point extraction means generates a three-dimensional line segment and extracts original measurement points near the three-dimensional line segment. Then, the mapping means represents the two-dimensional shape between the start point and end point in the point cloud image based on the image measurement points corresponding to the original measurement points extracted by the three-dimensional neighboring point extraction means.

[0016] The point cloud image manipulation system of the present invention may further include a two-dimensional neighboring point extraction means and a feature data generation means. The two-dimensional neighboring point extraction means generates line segments (hereinafter referred to as "two-dimensional line segments") on the point cloud image based on image measurement points and extracts image measurement points located near these two-dimensional line segments. The feature data generation means generates feature data arranged in three-dimensional space. In this case, when the operator specifies a desired start point and end point in the point cloud image displayed on the display means using the point specification means, the two-dimensional neighboring point extraction means generates a two-dimensional line segment based on these start point and end point and extracts image measurement points near the two-dimensional line segment. The feature data generation means then generates feature data based on the original measurement points corresponding to the start point and end point and the original measurement points corresponding to the nearby image measurement points extracted by the two-dimensional neighboring point extraction means. [Effects of the Invention]

[0017] The point cloud image manipulation system of the present invention has the following advantages. (1) The operator can visually view the point cloud image while processing the feature data, which makes it easy to recognize features, and the operator can easily grasp the depth-wise arrangement of the display. In addition, the operator can easily specify the desired measurement point. (2) Since the point cloud image at the image measurement point is associated with the original measurement point having three-dimensional coordinates, spatial operations in the three-dimensional coordinate system (e.g., distance calculation) can be performed simply by operating on the point cloud image. (3) The original measurement points that make up the image measurement point set are arranged, for example, horizontally so that they are arranged in the order of measurement time, and the image measurement point set is arranged, for example, vertically so that they are arranged in the order of measurement time, to create a point cloud image. Therefore, point cloud images arranged vertically are likely to represent similar features, meaning that a point cloud image can be created in which features are easy to recognize. [Brief explanation of the drawings]

[0018] [Figure 1] FIG. 1 is a block diagram showing the main configuration of a point cloud image manipulation system according to the present invention. [Figure 2] (a) is a side view showing a schematic diagram of the laser scan line in MMS measurement, and (b) is a plan view showing a schematic diagram of the laser scan line as seen from above. [Figure 3] FIG. 1A is a model diagram showing a set of image measurement points acquired in the same laser irradiation cycle, and FIG. 1B is a plan view showing a point cloud image according to the present invention. [Figure 4] (a) is a 3D point cloud image diagram representing a 3D point cloud, and (b) is a point cloud image diagram based on this 3D point cloud. [Figure 5] FIG. 1 is a flowchart showing the main processing flow of the point cloud image manipulation system 100 of the present invention. [Figure 6] FIG. 1 is a flow diagram showing the main processing flow up to displaying the 2D shape of a feature using a 3D neighboring point extraction means in the point cloud image manipulation system. [Figure 7] FIG. 1 is a flowchart showing the main processing flow up to generation of feature data using a two-dimensional nearby point extraction means and a feature data generation means. DETAILED DESCRIPTION OF THE INVENTION

[0019] An example of a point cloud image manipulation system according to the present invention will be described with reference to the drawings.

[0020] 1 is a block diagram showing the main components of a point cloud image manipulation system 100 of the present invention. As shown in this figure, the point cloud image manipulation system 100 comprises mapping means 101, display means 102, point designation means 103, attribute acquisition means 104, and measurement point storage means 109, and can also comprise inter-point distance calculation means 105, three-dimensional neighboring point extraction means 106, two-dimensional neighboring point extraction means 107, and feature data generation means 108.

[0021] The mapping means 101, attribute acquisition means 104, point-to-point distance calculation means 105, 3D neighboring point extraction means 106, 2D neighboring point extraction means 107, and feature data generation means 108 that make up the point cloud image manipulation system 100 can be manufactured as dedicated units, or a general-purpose computer device can be used. This computer device includes a processor such as a CPU, memories such as ROM and RAM, and some also include input means such as a mouse and keyboard, and a display, and can be configured, for example, as a personal computer (PC) or server.

[0022] The measurement point storage means 109 can be a storage device of a general-purpose computer (for example, a personal computer) or can be built in a database server. When built in a database server, it can be placed on a local network (LAN: Local Area Network) or a cloud server that stores data via the Internet can be used.

[0023] Below, each of the main elements that make up the point cloud image manipulation system 100 of the present invention will be described in detail.

[0024] (Measuring point storage means) The measurement point storage means 109 stores attribute information about a large number of 3D measurement points (i.e., 3D point clouds) acquired by laser measurement (particularly laser measurement using an MMS). Here, attribute information includes at least 3D coordinates, and can also include other information about the measurement, information about features, and image information captured along with the laser measurement. Each 3D measurement point is stored in association with (linked to) the 3D coordinates and the time of measurement (hereinafter simply referred to as the "measurement time"). It is also advisable to store the "number of times the laser scanner rotated (hereinafter referred to as the "laser irradiation period")," which will be described later. However, if the laser irradiation period can be determined from the measurement time, it is not necessary to store the laser irradiation period.

[0025] (Mapping method) The mapping means 101 reads out the three-dimensional point cloud data from the measurement point storage means 109 and creates a "point cloud image" using a plurality of three-dimensional measurement points. The procedure by which the mapping means 101 creates a point cloud image will be described below.

[0026] As shown in Figure 2(a), in MMS measurements, a moving object (such as a car) irradiates the laser while moving and the laser scanner LS rotates. Figure 2(a) is a side view that schematically illustrates the laser scan line (the trajectory of the laser irradiation) in MMS measurements. As mentioned above, the number of rotations of the laser scanner LS (i.e., the first rotation, the second rotation, ..., the nth rotation, ...) is referred to as the laser irradiation period, and is used in the sense of an ordinal number, such as "first laser irradiation period," "second laser irradiation period," or "nth laser irradiation period." For example, in the case of Figure 2(b), the period from 3D measurement point meter P01 to 3D measurement point meter P02 via 3D measurement point meter P04 is the "first laser irradiation period." Similarly, the period from 3D measurement point meter P02 to 3D measurement point meter P05 via 3D measurement point meter P03 is the "second laser irradiation period." FIG. 2(b) is a plan view seen from above, which schematically shows the laser scan line.

[0027] The rotation of the laser scanner LS allows it to acquire 3D measurement points over a wide area, but naturally, once the laser scanner LS rotates once, the laser returns to its original direction. Because the laser scanner rotates at a significantly higher speed than the moving objects that make up the MMS, 3D measurement points with the same irradiation direction (hereafter referred to as "laser irradiation phase" for convenience) in consecutive laser irradiation cycles (e.g., the first laser irradiation cycle and the second laser irradiation cycle) measure roughly the same location. In other words, 3D measurement points with the same laser irradiation phase in consecutive laser irradiation cycles can be considered to be measuring the same feature. For example, in the case of Figure 2(b), 3D measurement point meters P01 and P02 measure the same feature, and 3D measurement point meters P07 and P08 measure the same feature.

[0028] Therefore, in the present invention, a point cloud image is generated using each laser irradiation cycle as a single unit. That is, multiple 3D measurement points acquired during the same laser irradiation cycle are treated as a single group (hereinafter referred to as an "image measurement point set"), and a point cloud image is generated by arranging this image measurement point set on a plane. For example, in FIG. 3(a), the image measurement point set acquired during the nth laser irradiation cycle is represented by a white circle, and the image measurement point set acquired during the (n+1)th laser irradiation cycle is represented by a black circle. A point cloud image is generated by arranging these image measurement point sets on a plane, as shown in FIG. 3(b). More specifically, as shown in FIG. 3(b), the multiple (19 in the figure) 3D measurement points constituting the image measurement point set are arranged along the first axis (horizontal in the figure) and in the order of measurement time, and the image measurement point set is arranged along the second axis (vertical in the figure) perpendicular to the first axis, in the order of measurement time. For convenience, in order to distinguish between the original 3D measurement points and the 3D measurement points arranged as a point cloud image, the original 3D measurement points will be referred to as "original measurement points" and the 3D measurement points in the point cloud image will be referred to as "image measurement points."

[0029] The image measurement point placement process described above is executed by the mapping means 101. The mapping means 101 also reads attribute information of the original measurement points corresponding to the image measurement points from the measurement point storage means 109 and generates a point cloud image by assigning pixel values ​​to each image measurement point according to the attribute information. In other words, an image measurement point is assigned to each pixel constituting the point cloud image, and then a pixel value is assigned. Examples of the attribute information of the original measurement points include the reflection intensity of the laser received in the MMS measurement, height information (elevation value), and values ​​calculated from the height information (for example, the inclination angle with respect to an adjacent point). Furthermore, if an image captured in the MMS measurement is available, RGB or other attributes can be used as the attribute information of the original measurement points by associating the original measurement points with the image (by creating a so-called colored point cloud). The pixel values ​​assigned to each pixel can be grayscale, which represents shades of gray, color information such as RGB, or a combination of shades of gray and color information.

[0030] As shown in FIG. 3(b), the point cloud image of the present invention has image measurement point sets for successive laser irradiation cycles (the first and second laser irradiation cycles in the figure) arranged vertically. Furthermore, image measurement points for the same laser irradiation phase are adjacent to each other vertically, so that image measurement points relating to the same feature are arranged in a relatively concentrated manner. This makes it easy to grasp features even in a two-dimensional point cloud image. FIG. 4(a) shows an image of a three-dimensional point cloud (i.e., the original measurement points) (a so-called three-dimensional point cloud image), and FIG. 4(b) shows a point cloud image based on this three-dimensional point cloud. As can be seen from FIG. 4(b), buildings, roads, white lines, utility poles, power lines, etc. can be easily identified.

[0031] (Display means and point designation means) The display means 102 displays the point cloud image generated by the mapping means 101, and may be, for example, a display of a personal computer. The point designation means 103 allows an operator to designate a desired position in the point cloud image while checking the point cloud image displayed on the display means 102, and may be a pointing device (such as a mouse, touch panel, pen tablet, touchpad, trackpad, or trackball), a keyboard, or the like.

[0032] (Attribute acquisition means) The attribute acquisition means 104 reads out attribute information (especially, three-dimensional coordinates) of the original measurement point corresponding to the image measurement point. When the operator specifies a desired position in the point cloud image using the point specification means 103, the attribute acquisition means 104 identifies the image measurement point at that position, selects the original measurement point corresponding to that image measurement point, and further queries the measurement point storage means 109 with that original measurement point to read out the attribute information (three-dimensional coordinates) related to that original measurement point.

[0033] (Means for calculating distance between points) The point-to-point distance calculation means 105 calculates the distance between two points based on the three-dimensional coordinates read out by the attribute acquisition means 104. When the operator uses the point designation means 103 to designate two desired points (i.e., a "starting point" and an "ending point") in the point cloud image, the attribute acquisition means 104 reads out the three-dimensional coordinates of the starting point and the ending point, and the point-to-point distance calculation means 105 calculates the distance between the starting point and the ending point using these three-dimensional coordinates. Therefore, the point-to-point distance calculated by the point-to-point distance calculation means 105 is a distance in three-dimensional space (i.e., real space).

[0034] (3D neighboring point extraction method) The three-dimensional neighboring point extraction means 106 generates three-dimensional line segments and extracts original measurement points (hereinafter referred to as "nearby original measurement points") located near the three-dimensional line segments. When an operator specifies a desired start point and end point in the point cloud image using the point specification means 103, the attribute acquisition means 104 reads out the three-dimensional coordinates of the start point and end point, and the three-dimensional neighboring point extraction means 106 uses these three-dimensional coordinates to generate a three-dimensional line segment connecting the start point and end point. Therefore, the three-dimensional line segment generated by the three-dimensional neighboring point extraction means 106 is a line segment in three-dimensional space (i.e., real space). In addition, the three-dimensional neighboring point extraction means 106 reads out original measurement points located near the generated three-dimensional line segment (for example, within a predetermined distance from the three-dimensional line segment) from the measurement point storage means 109 as near original measurement points. When the near original measurement points are read out by the three-dimensional neighboring point extraction means 106, the mapping means 101 assigns a special pixel value to the image measurement point corresponding to the near original measurement point. For example, when an operator identifies a utility pole in the point cloud image and specifies two points (starting point and ending point) above and below the utility pole using the point specifying means 103, the 3D nearby point extraction means 106 generates a 3D line segment representing the utility pole and extracts nearby original measurement points (i.e., the original measurement points that make up the utility pole).Then, the mapping means 101 assigns a red pixel value to the image measurement points that correspond to the nearby original measurement points, so that the utility pole is displayed in red in the point cloud image, i.e., the shape of the utility pole is clearly shown on a plane (2D).

[0035] (Method for extracting 2D nearby points and generating feature data) The two-dimensional neighboring point extraction means 107 generates two-dimensional line segments and extracts image measurement points located near the two-dimensional line segments (hereinafter referred to as "neighboring image measurement points"). When an operator specifies a desired start point and end point in the point cloud image using the point specification means 103, the two-dimensional neighboring point extraction means 107 identifies the two-dimensional coordinates of the start point and end point in the point cloud image and generates a two-dimensional line segment connecting the start point and end point using these two-dimensional coordinates. Therefore, the two-dimensional line segment generated by the two-dimensional neighboring point extraction means 107 is a line segment on the point cloud image (i.e., in two-dimensional space). The two-dimensional neighboring point extraction means 107 also extracts image measurement points located near the generated two-dimensional line segment (for example, within a predetermined distance from the two-dimensional line segment) as neighboring image measurement points.

[0036] The feature data generation means 108 generates feature data to be arranged in three-dimensional space based on the nearby image measurement points extracted by the two-dimensional nearby point extraction means 107. When the start point and end point are specified by the operator and the nearby image measurement points are extracted by the two-dimensional nearby point extraction means 107, the feature data generation means 108 reads out the original measurement points corresponding to the start point, end point and nearby image measurement points from the measurement point storage means 109. The feature data generation means 108 then generates data (hereinafter referred to as "feature data") in which the read-out original measurement points are grouped.

[0037] For example, when an operator identifies a sign in a point cloud image and specifies two points (starting point and ending point) above and below this sign using the point specifying means 103, the two-dimensional neighboring point extraction means 107 generates a two-dimensional line segment representing the sign and extracts nearby image measurement points (i.e., image measurement points that make up the utility pole).Then, the feature data generation means 108 reads out original measurement points corresponding to the starting point, ending point, and nearby image measurement points from the measurement point storage means 109, and further generates feature data using these read-out original measurement points.In this way, the operator can generate feature data while checking the point cloud image, which means that the operator can easily grasp the features before performing the feature data conversion process.

[0038] (Processing flow) The main processing of point cloud image manipulation system 100 will be described in detail below with reference to Figures 5 to 7. Figure 5 is a flow diagram showing the main processing flow of point cloud image manipulation system 100 of the present invention, Figure 6 is a flow diagram showing the main processing flow up to displaying the 2D shape of a feature using 3D nearby point extraction means 106 in point cloud image manipulation system 100, and Figure 7 is a flow diagram showing the main processing flow up to generating feature data using 2D nearby point extraction means 107 and feature data generation means 108. In these flow diagrams, the actions to be performed are shown in the center column, what is necessary for each action is shown in the left column, and what results from each action is shown in the right column.

[0039] As shown in Fig. 5, first, a three-dimensional point cloud (original measurement points) is read from the measurement point storage means 109 (Step 201). Once the original measurement points have been read, a point cloud image is generated by the mapping means 101, and the generated point cloud image is displayed on the display means 102 (Step 202). Then, when the operator specifies a desired position in the point cloud image using the point specification means 103 (Step 203), the attribute acquisition means 104 selects an original measurement point corresponding to the image measurement point at that position, and furthermore, by querying the measurement point storage means 109 with the original measurement point, reads attribute information (three-dimensional coordinates) related to the original measurement point (Step 204). Furthermore, when the operator specifies a start point and an end point, the attribute acquisition means 104 reads the three-dimensional coordinates of the start point and the end point, and the point-to-point distance calculation means 105 calculates the distance between the start point and the end point using the three-dimensional coordinates (Step 205).

[0040] 6, when the operator specifies a start point and an end point (Step 203), the attribute acquisition means 104 reads out the three-dimensional coordinates of the start point and the end point (Step 204), and the three-dimensional neighboring point extraction means 106 uses these three-dimensional coordinates to generate a three-dimensional line segment connecting the start point and the end point (Step 206). The three-dimensional neighboring point extraction means 106 also reads out original measurement points that are in the vicinity of the generated three-dimensional line segment from the measurement point storage means 109 as neighboring original measurement points (Step 207). Then, when the neighboring original measurement points are read out by the three-dimensional neighboring point extraction means 106, the mapping means 101 assigns special pixel values ​​to the image measurement points that correspond to these neighboring original measurement points, thereby clearly showing the shape of the feature on a plane (two-dimensional) (Step 208).

[0041] As shown in Fig. 7, when the operator specifies a start point and an end point (Step 203), the two-dimensional neighboring point extraction means 107 identifies the two-dimensional coordinates of the start point and end point in the point cloud image and generates a two-dimensional line segment connecting the start point and end point using these two-dimensional coordinates (Step 209). The two-dimensional neighboring point extraction means 107 also extracts image measurement points near the generated two-dimensional line segment as neighboring image measurement points (Step 210). Once the neighboring image measurement points have been extracted by the two-dimensional neighboring point extraction means 107, the feature data generation means 108 reads out original measurement points corresponding to the start point, end point, and neighboring image measurement points from the measurement point storage means 109 (Step 211). The feature data generation means 108 then generates feature data by grouping the read original measurement points (Step 212). [Industrial Applicability]

[0042] The point cloud image manipulation system of the present invention is particularly suitable for use in managing various facilities, including road facilities, and as map information for automated driving. Furthermore, the present invention can provide highly accurate step information that is useful for elderly people and wheelchair users, and can also be effectively used in disaster prevention planning. Therefore, the point cloud image manipulation system of the present invention is not only applicable to industry, but is also expected to make a significant contribution to society. [Explanation of symbols]

[0043] 100 Point cloud image manipulation system of the present invention 101 Mapping method (point cloud image manipulation system) 102 Display means (of point cloud image manipulation system) 103 Point designation means (for point cloud image manipulation system) 104 Attribute Acquisition Method (for Point Cloud Image Manipulation System) 105 (Point cloud image manipulation system) Point distance calculation means 106 (Point Cloud Image Manipulation System) 3D Nearby Point Extraction Method 107 (Point Cloud Image Manipulation System) 2D Nearby Point Extraction Method 108 (Point Cloud Image Manipulation System) Feature Data Generation Method 109 (Point Cloud Image Manipulation System) Measurement Point Storage Means LS laser scanner

Claims

1. a measurement point storage means for storing three-dimensional original measurement points acquired by the laser scanner while moving; a mapping means for arranging the original measurement points read from the measurement point storage means on a plane as image measurement points and generating a two-dimensional point cloud image based on attributes of the original measurement points; a display means for displaying the point cloud image; a point designation means for allowing an operator to designate a desired position in the point cloud image displayed on the display means; an attribute acquisition means for reading, from the measurement point storage means, the attribute of the original measurement point corresponding to the image measurement point related to the position designated by the point designation means; the mapping means arranges the original measurement points acquired during one rotation of the laser scanner as the image measurement points in a first axis direction so as to be arranged in order of measurement time, and arranges an image measurement point set consisting of the image measurement points arranged in the first axis direction in a second axis direction perpendicular to the first axis direction so as to be arranged in order of measurement time. A point cloud image manipulation system characterized by:

2. further comprising a point-to-point distance calculation means for calculating a point-to-point distance between the two original measurement points designated by the point designation means based on the three-dimensional coordinates of the original measurement points read by the attribute acquisition means; 2. The point cloud image manipulation system according to claim 1.

3. a three-dimensional neighboring point extracting means for generating a three-dimensional line segment in a three-dimensional space based on the three-dimensional coordinates of the original measurement point and extracting the original measurement point located in the vicinity of the three-dimensional line segment; When an operator specifies a desired start point and an end point in the point cloud image displayed on the display means by the point specifying means, the attribute acquiring means reads out the three-dimensional coordinates of the original measurement points corresponding to the start point and the end point from the measurement point storage means, and the three-dimensional neighboring point extracting means generates the three-dimensional line segment based on the read three-dimensional coordinates of the original measurement points and extracts the neighboring original measurement points, Furthermore, the mapping means represents a two-dimensional shape between the starting point and the ending point in the point cloud image based on the image measurement points corresponding to the original measurement points extracted by the three-dimensional neighboring point extraction means.

3. The point cloud image manipulation system according to claim 1 or 2.

4. a two-dimensional neighboring point extraction means for generating two-dimensional line segments on the point cloud image based on the image measurement points and extracting the image measurement points located in the vicinity of the two-dimensional line segments; and a feature data generating means for generating feature data to be arranged in a three-dimensional space, When an operator specifies a desired start point and an end point in the point cloud image displayed on the display means by the point specifying means, the two-dimensional neighboring point extracting means generates the two-dimensional line segment based on the start point and the end point and extracts the neighboring image measurement points; further, the feature data generating means generates the feature data based on the original measurement points corresponding to the start point and the end point, and the original measurement points corresponding to the image measurement points extracted by the two-dimensional neighboring point extracting means.

4. The point cloud image manipulation system according to claim 1, wherein the point cloud image manipulation system is a point cloud manipulation system.

Images