Information processing system, information processing method, and information processing device

Abstract

In this information processing system, a first acquisition unit acquires target topography data indicating the terrain of a target region generated on the basis of sensing data of the target region. A second acquisition unit acquires post-removal topography data in which processing for removing data of a target object to be removed has been performed on the acquired target topography data. A distance information calculation unit calculates first distance information relating to the distance to data in the vicinity of each piece of data included in the target topography data, and calculates second distance information relating to the distance to data in the vicinity of each piece of data included in the post-removal topography data. A determination unit determines, on the basis of the calculated first distance information and second distance information, whether or not each piece of data included in the post-removal topography data is to be removed.

Description

Information processing system, information processing method, and information processing device

[0001] The present technology relates to an information processing system, an information processing method, and an information processing device that can be applied to generating topographical data of a surveying site or the like.

[0002] Patent Literature 1 discloses an information processing system that enables accurate extraction of specific regions from images such as aerial photographs and satellite photographs. In this information processing system, specific region candidates in the image are extracted by color analysis in which a specific color range is set for the image. Then, based on surface height information and reference plane height information of the image, noise regions that are within the specific color range but are not specific regions are detected and removed from the extracted specific region candidates.

[0003] Japanese Patent Application Laid-Open No. 2008-242508

[0004] At surveying sites, it is often necessary to obtain topographical data of the ground surface, excluding unnecessary objects such as vegetation and buildings. In such cases, it is important to accurately remove data on objects that are unnecessary in the target area from the topographical data of the target area.

[0005] In view of the above circumstances, an object of the present technology is to provide an information processing system, an information processing method, and an information processing device that are capable of removing data of removal target objects with high accuracy.

[0006] To achieve the above object, an information processing system according to one embodiment of the present technology includes a first acquisition unit, a second acquisition unit, a distance information calculation unit, and a determination unit. The first acquisition unit acquires target terrain data indicating the terrain of a target area, which is generated based on sensing data of the target area. The second acquisition unit acquires removed terrain data in which a process for removing data of objects to be removed is performed on the acquired target terrain data. The distance information calculation unit calculates first distance information regarding the distance between each piece of data included in the target terrain data and neighboring data, and calculates second distance information regarding the distance between each piece of data included in the removed terrain data and neighboring data. The determination unit determines whether each piece of data included in the removed terrain data is a removal target based on the calculated first distance information and second distance information.

[0007] In this information processing device, first distance information is calculated for each data item included in target terrain data indicating the terrain of a target area, relating to the distance to neighboring data. Second distance information is calculated for each data item included in post-removal terrain data, in which processing to remove data of objects to be removed from the target terrain data has been performed. Then, based on the first distance information and the second distance information, a determination is made as to whether each data item included in the post-removal terrain data is data to be removed. Based on the result of this determination, it becomes possible to remove data of objects to be removed with high accuracy.

[0008] The determination unit may perform the determination for each piece of data included in the removed topographical data based on a difference between the first distance information and the second distance information.

[0009] The determination unit may perform at least one of determining whether or not an object is a removal target and determining the degree to which the object is a removal target.

[0010] The distance information calculation unit may calculate the average of the distances for a predetermined number of nearby data for each data included in the target terrain data as the first distance information, and may calculate the average of the distances for the predetermined number of nearby data for each data included in the removed terrain data as the second distance information.

[0011] The determination unit may determine that each piece of data included in the removed topographical data is to be removed when a difference between the first distance information and the second distance information is greater than a predetermined threshold.

[0012] The determination unit may determine that, for each piece of data included in the removal-processed terrain data, the greater the difference between the first distance information and the second distance information, the greater the likelihood that the piece of data is to be removed.

[0013] The distance information calculation section may set the predetermined number based on the density of at least one of the target terrain data and the removal-processed terrain data.

[0014] The second acquisition unit may acquire attribute-assigned terrain data in which attribute information of objects present in the target area is assigned to each piece of data included in the target terrain data, and may acquire the removed terrain data by removing data included in the attribute-assigned terrain data to which attribute information of the object to be removed is assigned.

[0015] The information processing system may further include a data updating unit that updates at least one of the attribute-added terrain data and the removed terrain data based on the determination result by the determining unit.

[0016] The distance information calculation unit may calculate distance information regarding the distance in a predetermined direction to nearby data as the first distance information, and may calculate distance information regarding the distance in the predetermined direction to nearby data as the second distance information.

[0017] The predetermined direction may be a vertical direction.

[0018] The distance information calculation unit may calculate the first distance information and the second distance information by weighting distances in three directions that are orthogonal to each other.

[0019] The three mutually orthogonal directions may be a vertical direction, a first direction orthogonal to the vertical direction, and a second direction orthogonal to each of the vertical direction and the first direction.

[0020] The distance information calculation unit may generate a plurality of division target terrain data and a plurality of division-removed terrain data by dividing the target terrain data and the removed terrain data in a similar manner, calculate the first distance information for each of the plurality of division target terrain data, and calculate the second distance information for each of the plurality of division-removed terrain data. In this case, the determination unit may perform the determination based on the first distance information and the second distance information for each of the plurality of division-removed terrain data.

[0021] The target terrain data may be composed of point cloud data or image data.

[0022] An information processing method according to one aspect of the present technology is an information processing method executed by a computer system, and includes acquiring target terrain data indicating a terrain of a target area generated based on sensing data of the target area. A process for removing data of objects to be removed from the acquired target terrain data is performed to acquire post-removal-processed terrain data. First distance information regarding a distance between each piece of data included in the target terrain data and neighboring data is calculated, and second distance information regarding a distance between each piece of data included in the post-removal-processed terrain data and neighboring data is calculated. A determination is made as to whether each piece of data included in the post-removal-processed terrain data is a target for removal based on the calculated first distance information and second distance information.

[0023] An information processing device according to an embodiment of the present technology includes the first acquisition unit, the second acquisition unit, the distance information calculation unit, and the determination unit.

[0024] FIG. 1 is a schematic diagram showing an example of the configuration of a terrain data generation system according to a first embodiment. FIG. 2 is a flowchart showing the basic operation of a server device. FIG. 3 is a schematic diagram showing an example of a target area. FIG. 4 is a schematic diagram showing point cloud data generated as target terrain data. FIG. 5 is a schematic diagram showing point cloud data to which attribute information of an object has been assigned. FIG. 6 is a schematic diagram showing an example of the structure of attribute-assigned point cloud data. FIG. 7 is a schematic diagram showing removal-processed terrain data. FIG. 8 is a flowchart showing an example of processing by a distance information calculation unit and a determination unit. FIG. 9 is a schematic diagram for explaining an example of calculation of first distance information and second distance information. FIG. 10 is a flowchart showing another example of processing by the distance information calculation unit and a determination unit. FIG. 11 is a schematic diagram for explaining an example of data updating (adding an unnecessary flag). FIG. 12 is a schematic diagram for explaining an example of data updating (adding an erroneous recognition degree). FIG. 13 is a schematic diagram showing a DTM from which erroneous recognition point data has been removed. FIG. 14 is a schematic diagram showing attribute-assigned point cloud data from which erroneous recognition has been improved. FIG. 15 is a schematic diagram showing an example of the structure of attribute-assigned point cloud data according to a second embodiment. FIG. 16 is a block diagram showing an example of the hardware configuration of a computer that can be used as a server device and a user terminal.

[0025] Hereinafter, embodiments of the present technology will be described with reference to the drawings.

[0026] 1 is a schematic diagram showing a configuration example of a terrain data generation system according to a first embodiment of the present technology. The terrain data generation system 1 includes a server device 2, a database (DB) 3, a user terminal 4, a sensor 5, and a mobile object 6.

[0027] The server device 2 and the user terminal 4 have hardware necessary for a computer, such as a processor such as a CPU, a GPU, or a DSP, a memory such as a ROM or a RAM, and a storage device such as an HDD (see FIG. 16 ). The processor loads a program according to the present technology stored in the storage unit or the memory into the RAM and executes the program, thereby executing the information processing method (topography data generation method) according to the present technology.

[0028] The server device 2 and the user terminal 4 are connected to each other so that they can communicate with each other. The connection between the two devices is not limited, and may be achieved, for example, by wireless LAN communication such as Wi-Fi or short-range wireless communication such as Bluetooth (registered trademark). Of course, a wired connection may also be adopted.

[0029] Any sensor capable of generating topographical data of the target area TA or outputting a signal including the data is used as the sensor 5. In this embodiment, a digital camera equipped with an image sensor is used as the sensor 5. For example, a CMOS (Complementary Metal-Oxide Semiconductor) sensor or a CCD (Charge Coupled Device) sensor is used as the image sensor.

[0030] A distance measurement sensor can also be used as the sensor 5. As the distance measurement sensor, various types of distance measurement sensors can be used, such as an optical laser type distance measurement sensor, an ultrasonic type distance measurement sensor, a stereo camera, a ToF (Time of Flight) sensor, a LiDAR (Light Detection and Ranging, Laser Imaging Detection and Ranging), or a structured light type distance measurement sensor.

[0031] Both a digital camera and a distance measurement sensor may be used as the sensor 5. Alternatively, a sensor having the functions of both an image sensor and a distance measurement sensor may be used as the sensor 5. For example, a ToF sensor capable of detecting distance information for each pixel may be used.

[0032] 1 , in this embodiment, the sensor 5 is mounted on a drone, which is a mobile object 6. The drone is moved above a target area TA, and an image of the target area TA (hereinafter referred to as an aerial image) is captured by the sensor 5 mounted on the drone.

[0033] The application of this technology is not limited to a specific type of moving body 6. The technology can be applied to any moving body 6, including other flying bodies such as airplanes, helicopters, and satellites, as well as vehicles, robots, and operators (humans).

[0034] DB 3 stores various information and data related to the present topographical data generation system 1. For example, various information is stored, such as information on the target area TA, the generated topographical data, information on the user terminal 4, and information on the user 7 who uses the present topographical data generation system 1.

[0035] 1, DB3 is configured by a storage device or the like separate from the server device 2 and is connected to the server device 2. The configuration is not limited to this, and DB3 may be configured by a storage unit (see FIG. 16) of the server device 2. Also, a configuration may be adopted in which DB3 is constructed on a network and is accessible via the network.

[0036] The user terminal 4 is used by a user 7 who uses the terrain data generation system 1. For example, the user 7 downloads an application (application program) for using the terrain data generation system 1 to the user terminal 4. For example, the user 7 inputs information such as an ID and a password to create an account for using the terrain data generation system 1. Of course, creating an account may not be necessary.

[0037] The specific configuration of the user terminal 4 is not limited, and any computer may be used, such as a PC, a smartphone, a tablet terminal, or an HMD (Head Mounted Display) such as AR glasses or VR glasses.

[0038] The topographical data generation system 1 shown in Fig. 1 functions as an embodiment of an information processing system according to the present technology. The server device 2 shown in Fig. 1 functions as an embodiment of an information processing device according to the present technology.

[0039] 1, in this embodiment, the processor of the server device 2 executes a predetermined program to configure functional blocks including a first acquisition unit 8, a second acquisition unit 9, a distance information calculation unit 10, a determination unit 11, and a data update unit 12. Of course, dedicated hardware such as an IC (integrated circuit) may be used to realize each functional block.

[0040] The program is installed on the server device 2 via, for example, various recording media. Alternatively, the program may be installed via the Internet or the like. There are no limitations on the type of recording media on which the program is recorded, and any computer-readable recording media may be used. For example, any computer-readable non-transitory storage medium may be used.

[0041] 2 is a flowchart showing the basic operation of the server device 2. The first acquisition unit 8 acquires target terrain data that indicates the terrain of the target area TA and is generated based on sensing data of the target area TA (step 101).

[0042] In the present disclosure, acquiring data (information) includes not only receiving the data (information) from another device, but also generating the data (information) by executing a predetermined algorithm. Also, in the present disclosure, "device A acquires data C generated based on data B" includes not only "device A receives data C generated based on data B by a device other than device A," but also "device A itself generates data C based on data B."

[0043] That is, step 101 shown in FIG. 2 includes not only the first acquisition unit 8 receiving target terrain data from another device, but also the first acquisition unit 8 generating target terrain data.

[0044] 1 , a plurality of aerial images captured by a sensor 5 mounted on a moving object 6 are input as sensing data to a server device 2. For example, a storage medium such as an SD card is set in the sensor 5, and the captured aerial images are stored in the storage medium. The storage medium is set in the server device 2, and the plurality of aerial images are input to the server device 2.

[0045] Of course, the present invention is not limited to this configuration, and the sensor 5 and the server device 2 may be communicatively connected via wireless LAN communication, short-range wireless communication, etc. Then, a plurality of aerial images taken by the sensor 5 may be transmitted to the server device 2.

[0046] Fig. 3 is a schematic diagram showing an example of a target area TA. In the example shown in Fig. 3, the target area TA includes a ground surface (including roads) 14, buildings 15, vegetation 16, and vehicles 17. The buildings 15, vegetation 16, and vehicles 17 are located on the ground surface 14.

[0047] 3 shows a view of the target area TA from almost directly above, but it is possible to photograph the target area TA from various angles by controlling the flight of the mobile object 6 shown in FIG. 1. The multiple aerial images thus photographed are input to the server device 2.

[0048] In this embodiment, the first acquisition unit 8 generates point cloud data indicating the three-dimensional structure of the target area TA as target terrain data based on a plurality of aerial photograph images input as sensing data.

[0049] Fig. 4 is a diagram showing a schematic diagram of point cloud data generated as target terrain data. In the target terrain data (point cloud data) 19 shown in Fig. 4, each circle represents one point data 20. Note that the size, arrangement, etc. of the point data 20 shown in Fig. 4 are shown only as a schematic diagram.

[0050] Each point data 20 included in the point cloud data includes coordinate information (X, Y, Z) as position information in three-dimensional space. As a coordinate system for defining the coordinate information (X, Y, Z), an absolute coordinate system (world coordinate system) may be used, or a relative coordinate system with a predetermined point as the reference (origin) may be used.

[0051] Typically, the Z direction is the height direction, and coordinate values ​​in the vertical direction are stored as Z coordinate values. Furthermore, coordinate values ​​in two mutually orthogonal directions in the horizontal direction perpendicular to the vertical direction are stored as X and Y coordinates. Of course, this is not a limitation, and the present technology can also be applied by arbitrarily setting three mutually orthogonal directions.

[0052] The specific technique (algorithm) for generating point cloud data of the target area TA, which is the target terrain data 19, is not limited, and any technique (algorithm) may be used. For example, point cloud data can be generated using SfM (Structure from Motion). As is well known, SfM is a technique for determining the three-dimensional structure of an object and the camera position from multiple captured images taken while changing the camera's viewpoint. In SfM, point cloud data indicating the three-dimensional structure of the object is generated based on the results of detecting corresponding points (identical feature points) from multiple captured images taken while changing the camera's viewpoint.

[0053] To generate point cloud data, any machine learning algorithm may be used, such as a deep neural network (DNN), a recurrent neural network (RNN), or a convolutional neural network (CNN). For example, by using AI (artificial intelligence) that performs deep learning, it is possible to generate highly accurate point cloud data.

[0054] The application of a machine learning algorithm may be performed for any process within the present disclosure, i.e., any process described within the present disclosure may be subjected to a process using machine learning.

[0055] The second acquisition unit 9 acquires post-removal processing terrain data in which processing for removing data of objects to be removed is performed on the target terrain data 19 acquired in step 101 (step 102). In this embodiment, the second acquisition unit 9 executes processing for removing data of objects to be removed, and generates post-removal processing terrain data.

[0056] In this embodiment, the building 15, the vegetation 16, and the vehicle 17 shown in Fig. 3 are examples of objects to be removed. Of course, the present technology can be applied without limiting the specific type of object to be removed. For example, in the example shown in Fig. 3, the present technology can be applied to the "ground" as an object to be removed.

[0057] In this embodiment, first, the second acquisition unit 9 analyzes each of the multiple aerial images and evaluates the features of the subject at multiple positions within the target area TA. Specifically, the second acquisition unit 9 performs image recognition as semantic segmentation on the multiple aerial images using a machine-learned artificial intelligence model (for example, an artificial intelligence model based on CNN or the like).

[0058] Based on the results of this image recognition, object class information (object attribute information) is determined for each point data 20 in the point cloud data. In SfM, corresponding point detection is performed for multiple aerial images taken from different viewpoints, and points in three-dimensional space are recognized based on the positional relationships of corresponding points in each aerial image, so it is known which pixel in the aerial image corresponds to which point data in the point cloud data. Therefore, when image recognition is performed on the aerial image, it is possible to know which point data 20 in the point cloud data corresponds to the image area in which an object is recognized.

[0059] Fig. 5 is a diagram schematically illustrating point cloud data to which attribute information of an object has been assigned. Hereinafter, the point cloud data shown in Fig. 5 will be referred to as attribute-assigned point cloud data 22. Fig. 6 is a diagram schematically illustrating an example of the structure of the attribute-assigned point cloud data 22.

[0060] For example, as the target terrain data 19 shown in FIG. 4, a number ("No.") is assigned to each point data 20, and coordinate information (X, Y, Z) and RGB brightness values ​​are stored for each point data 20.

[0061] As shown in Fig. 6, attribute information determined by image recognition is added as metadata to the target terrain data 19. This generates attribute-assigned point cloud data 22. Of course, the specific data structures of the target terrain data 19 and the attribute-assigned point cloud data 22 are not limited and may be designed arbitrarily.

[0062] As shown in FIG. 5 , in this embodiment, one of the following information is added as attribute information: "building," "vegetation," "vehicle," or "ground." As shown in FIG. 5 , a white circle represents point data 23 to which the attribute information of "ground" has been added. A black circle represents point data 24 to which the attribute information of "building" has been added. A dark gray circle represents point data 25 to which the attribute information of "vegetation" has been added. A light gray circle represents point data 26 to which the attribute information of "vehicle" has been added.

[0063] In this embodiment, the attribute-assigned point cloud data 22 corresponds to an embodiment of attribute-assigned terrain data in which attribute information of objects existing in the target area is assigned to each data included in the target terrain data according to the present technology.

[0064] As shown in Fig. 5, classification by image recognition can result in erroneous recognition (missed recognition). For example, Fig. 5 shows erroneously recognized point data 27, in which an object that should have been recognized as either a "building," "vegetation," or "vehicle" was mistakenly recognized as "ground" (point data with an "x" inside a white circle).

[0065] The second acquisition unit 9 removes the point data 24 to 26 to which the attribute information of "buildings," "vegetation," and "vehicles" has been assigned from the attribute-assigned point cloud data 22. As a result, post-removal processing terrain data is generated in which processing for removing the data of the objects to be removed has been performed. Note that the specific algorithm for generating post-removal processing terrain data from the target terrain data is not limited.

[0066] 7 is a diagram showing a model of the removed terrain data 29. Point data 24 to 26 to which attribute information of "buildings," "vegetation," and "vehicles" has been assigned are removed from the attribute-assigned point cloud data 22.

[0067] 7, in the removed topographical data 29, erroneous recognition point data 27 that should have been recognized as either a "building," "vegetation," or "vehicle" but was mistakenly recognized as "ground" remains as data left to be removed. Such forgotten-to-be-removed erroneous recognition point data 27 becomes noise in ground extraction, and can cause a decrease in calculation accuracy when calculating, for example, altitude.

[0068] In the present topographical data generating system 1, it is possible to determine the erroneous recognition point data 27 shown in FIG. 7 and improve the accuracy of removing data of objects to be removed.

[0069] As shown in Fig. 2, the distance information calculation unit 10 calculates first distance information and second distance information (step 103). The first distance information is information relating to the distance between each point data 20 included in the target terrain data 19 shown in Fig. 4 and its neighboring point data 20. The second distance information is information relating to the distance between each point data 20 included in the removed terrain data 29 shown in Fig. 7 and its neighboring data 20.

[0070] In this embodiment, each point data 20 included in the target terrain data 19 corresponds to one embodiment of each data included in the target terrain data. Also, each point data 20 included in the removed terrain data 29 corresponds to one embodiment of each data included in the removed terrain data.

[0071] The determination unit 11 determines whether each point data 20 included in the removed terrain data 29 is a target for removal based on the first distance information and second distance information calculated in step 103 (step 104).

[0072] Fig. 8 is a flowchart showing an example of processing by the distance information calculation unit 10 and the determination unit 11. Fig. 9 is a schematic diagram for explaining an example of calculation of the first distance information and the second distance information.

[0073] In this embodiment, the first distance information is calculated as the average distance (referred to as the neighborhood distance average in FIG. 8) for a predetermined number of neighboring data points 20 included in the target terrain data 19 (step 201).

[0074] For example, as shown in Fig. 9A, taking the erroneous recognition point data 27 as an example, a predetermined n number of point data 20 are selected that exist in the vicinity of the erroneous recognition point data 27. Specifically, the n number of point data 20 are selected in order of proximity to the erroneous recognition point data 27 (15 number of point data 20 are selected in Fig. 9).

[0075] The distance between the point data 20 can be calculated using the following formula, where the coordinate values ​​of one point data 20 are (X1, Y1, Z1) and the coordinate values ​​of the other point data 20 are (X2, Y2, Z2).

[0076]

[0077] The average value of the distances to the selected n pieces of point data 20 is calculated as the first distance information. That is, the distances shown in equation (1) are calculated between the selected n pieces of point data 20, and the average value is calculated as the first distance information. Of course, the equation for calculating the distance is not limited to equation (1).

[0078] The first distance information is calculated for each point data 20 included in the target terrain data 19. Therefore, the first distance information is calculated for each point data 20 included in the target terrain data 19.

[0079] In this embodiment, the second distance information is calculated as the average distance (average neighborhood distance) for the same number (predetermined number) of neighboring data for each point data 20 included in the removed terrain data 29 (step 202).

[0080] 9B , taking the same erroneous recognition point data 27 as an example, n pieces of point data 20, the same number of which are present in the vicinity of the erroneous recognition point data 27, are selected. Specifically, the n pieces of point data 20 are selected in order of closest distance from the erroneous recognition point data 27.

[0081] The average value of the distances to the selected n pieces of point data 20 is calculated as the second distance information. That is, the distances shown in equation (1) are calculated between the selected n pieces of point data 20, and the average value is calculated as the second distance information.

[0082] The second distance information is calculated for each point data 20 included in the removed terrain data 29. Therefore, the second distance information is calculated for each point data 20 included in the removed terrain data 29.

[0083] In addition, the average distance for the n neighboring point data 20 calculated as the first distance information and the second distance information can also be called the "n neighbor point group distance average" or the "n neighbor distance average."

[0084] By removing the data of the object to be removed (point data (building) 24) from the target terrain data 19 shown in Fig. 9A, the removed terrain data 29 shown in Fig. 9B is generated. Therefore, in the removed terrain data 29, most of the point data 20 near the erroneously recognized point data 27 has been removed.

[0085] Therefore, in the erroneous recognition point data 27, there is a large difference between the first distance information, which is the average of the neighborhood distances in the target terrain data 19, and the second distance information, which is the average of the neighborhood distances in the removed terrain data 29. It is possible to detect the erroneous recognition point data 27 based on the difference between the first distance information and the second distance information.

[0086] The specific value of n, which is the number of point data 20 for calculating the neighborhood distance average, may be set as appropriate. If the value of n is too small, the data is more susceptible to the influence of the density of the target terrain data 19 and the removed terrain data 29. If the value of n is too large, the amount of calculation required to calculate the neighborhood distance average increases, increasing the processing load.

[0087] For example, the specific value of n can be set based on the density of at least one of the target terrain data 19 and the removed terrain data 29. For example, by setting the specific value of n to a value greater than the density of each of the target terrain data 19 and the removed terrain data 29, it becomes possible to detect the erroneous recognition point data 27 with high accuracy.

[0088] Furthermore, as a result of the inventor's investigation, it was found that by adopting 100 as the specific value of n, it was possible to accurately detect erroneous recognition point data 27 for various target terrain data 19 and removed terrain data 29. In other words, adopting a value of about 100 as the value of n is also an effective method. Of course, the value of n is not limited to 100.

[0089] The determination unit 11 calculates the difference between the first distance information and the second distance information for each point data 20 included in the removed terrain data 29 (step 203). Then, the determination unit 11 determines whether each point data 20 included in the removed terrain data 29 is a target for removal based on the difference between the first distance information and the second distance information.

[0090] Note that the distance information calculation unit 10 may calculate the difference between the first distance information and the second distance information in step 203 and output the calculation result to the determination unit 11 .

[0091] 8, the determination unit 11 performs threshold processing on the difference between the first distance information and the second distance information (step 204), and then determines whether or not the target point data 20 is to be removed based on the result of the threshold processing (step 205).

[0092] Specifically, the determination unit 11 determines that each point data 20 included in the removal-processed terrain data 29 is to be removed if the difference between the first distance information and the second distance information is greater than a predetermined threshold. In other words, the point data 20 for which the difference between the first distance information and the second distance information is greater than a predetermined threshold is determined to be erroneously recognized point data 27.

[0093] The specified threshold value can be set, for example, using the average or standard deviation of the average neighborhood distances of each point data 20 included in the target terrain data 19, or the average or standard deviation of the average neighborhood distances of each point data 20 included in the removed terrain data 29.

[0094] For example, the threshold value can be calculated by performing the following calculation on the point group of the removed topographical data 29.

[0095]

[0096] In equation 2, a is a coefficient. The value of a can be adjusted appropriately to set the threshold. For example, a value less than 1 or a value equal to or greater than 1 may be used. Of course, any other method (algorithm) may be used to set the threshold.

[0097] In this way, in the example shown in FIG. 8, the determination as to whether or not the data is to be removed is made by determining whether or not the data is to be removed (determining whether or not the data is the erroneously recognized point data 27).

[0098] 10 is a flowchart showing another example of processing by the distance information calculation unit 10 and the determination unit 11. Steps 301 to 303 shown in FIG.

[0099] In the example shown in Figure 10, the judgment unit 11 judges the degree of misrecognition for each point data 20 included in the removed terrain data 29 according to the difference between the first distance information and the second distance information (step 304).

[0100] Specifically, the determining unit 11 determines that the greater the difference between the first distance information and the second distance information for each point data 20 included in the removed topographical data 29, the greater the degree of erroneous recognition.

[0101] The degree of erroneous recognition is information indicating the degree of erroneous recognition, and is information indicating the degree to which the point data 20 is erroneously recognized point data 27 (i.e., information indicating to what extent the point data 20 is erroneously recognized point data 27). The degree of erroneous recognition can also be said to be information indicating to what extent the point data 20 is a target for removal, and corresponds to one embodiment of the degree of target for removal related to the present technology.

[0102] For example, the difference between the first distance information and the second distance information can be used as the recognition error rate. Alternatively, any algorithm may be used as a method for calculating the recognition error rate according to the difference between the first distance information and the second distance information.

[0103] In this way, in the example shown in FIG. 10, the determination as to whether or not a data item is to be removed is made by determining the degree to which the data item is to be removed (determining the degree to which the data item is the erroneous recognition point data 27).

[0104] In this embodiment, the data update unit 12 can update the attribute-assigned point cloud data 22 and the removed terrain data 29 based on the determination result by the determination unit 11. At least one of these data may be updateable.

[0105] In the present disclosure, data updating includes, for example, changing (overwriting) information of each item already stored in DB 3, deleting information of each item already stored in DB 3, and adding information of a new item to information already stored in DB 3. For example, adding metadata is also included in data updating.

[0106] 11 and 12 are schematic diagrams for explaining an example of data updating. For example, when the process shown in Fig. 8 is executed, as shown in Fig. 11, the data updating unit 12 adds an unnecessary object flag as metadata to the attribute-assigned point cloud data 22 shown in Fig. 6.

[0107] The unnecessary object flag is information indicating whether or not a piece of point data 20 is a removal target. Flag information (e.g., "1") indicating that the piece of point data 20 is a removal target is added to the point data 20 that the determination unit 11 determines to be a removal target. Flag information (e.g., "0") indicating that the piece of point data 20 is not a removal target is added to the point data 20 that the determination unit 11 determines to be not a removal target. Of course, the data format, etc. for adding the information indicating whether or not the piece of point data 20 is a removal target is not limited.

[0108] When the process shown in FIG. 10 is executed, as shown in FIG. 12, the data update unit 12 adds the degree of misrecognition as metadata to the attribute-added point cloud data 22 shown in FIG.

[0109] The attribute-assigned point cloud data 22 shown in Fig. 6 may be changed (overwritten) based on the determination result by the determination unit 11. For example, suppose that point data 20 (misrecognition point data 27) to be removed is detected by executing the process shown in Fig. 8. The attribute information of the misrecognition point data 27 is "ground" in the attribute-assigned point cloud data 22 shown in Fig. 6. The data update unit 12 changes the attribute information of the detected misrecognition point data 27 from "ground" to one of "building," "vegetation," or "vehicle."

[0110] For example, in the attribute-assigned point cloud data 22 shown in Fig. 5, point data 20 of the object to be removed (point data 24 to 26 having attribute information of either "building," "vegetation," or "vehicle") is selected, which is the closest point data 20 to the detected erroneous recognition point data 27. The attribute information of this point data 20 is overwritten as the attribute information of the detected erroneous recognition point data 27. This makes it possible to change the attribute information of the erroneously recognized erroneous recognition point data 27 to the original attribute information.

[0111] 10 is executed, it is possible to change the attribute information in the attribute-assigned point cloud data 22 for point data 20 whose degree of erroneous recognition is greater than a predetermined threshold. In this case, too, it is possible to change the attribute information of the erroneously recognized erroneous recognition point data 27 to the original attribute information.

[0112] The attribute-assigned point cloud data updated by the data update unit 12 is sent to the user terminal 4. On the user terminal 4, the user 7 starts an application for using the terrain data generation system 1. Within the application, it is possible to display, for example, aerial images taken by the mobile body 6 and target terrain data 19 acquired by the first acquisition unit 8.

[0113] The user 7 can also display, within the application, a DTM (Digital Terrain Model) in which only the ground surface portion of the environment has been extracted. For example, when the updated attribute-assigned point cloud data shown in Fig. 11 is generated, not only the point data 24 to 26 to which the attribute information of "buildings," "vegetation," and "vehicles" has been assigned, but also the point data 20 to which the attribute information of "ground" has been assigned but which has an unnecessary flag stored to indicate that it is to be removed, is removed.

[0114] This makes it possible to generate and display a DTM from which erroneous recognition point data 27 has been removed, as shown in Fig. 13. In other words, by using this topographical data generation system 1, highly accurate noise filtering is achieved, and it becomes possible to remove data of objects to be removed with high accuracy.

[0115] When the updated attribute-assigned point cloud data shown in Fig. 12 is generated, not only point data 24 to 26 to which attribute information of "buildings," "vegetation," and "vehicles" is assigned, but also point data 20 to which attribute information of "ground" is assigned but whose degree of erroneous recognition is greater than a predetermined threshold is removed. This makes it possible to generate and display a DTM from which erroneous recognition point data 27 has been removed, as shown in Fig. 13.

[0116] 12 has been generated, the user 7 can operate a predetermined control bar to adjust the threshold value for the degree of misrecognition, which is the criterion for determining whether or not a point is to be removed. By adjusting the threshold value, it is possible to adjust, for example, the number of point data 20 to be removed as removal targets (i.e., the number of point data 20 included in the DTM).

[0117] Furthermore, if the attribute-assigned point cloud data has been changed (overwritten) based on the determination result by the determination unit 11, that is, if the attribute information of the erroneously recognized point data 27 has been changed (overwritten), the point data 24 to 26 to which the attribute information of "building," "vegetation," and "vehicle" has been assigned are removed. This makes it possible to generate and display a DTM from which the erroneously recognized point data 27 has been removed, as shown in FIG.

[0118] In addition, if the attribute-assigned point cloud data is changed (overwritten) based on the judgment result by the judgment unit 11, it is also possible to display the attribute-assigned point cloud data 22 in which misrecognition (missing recognition) has been improved, as shown in Figure 14.

[0119] The data update unit 12 may update the removed terrain data 29 shown in Fig. 7. The data update unit 12 removes the erroneous recognition point data 27 included in the removed terrain data 29. The updated removed terrain data 29 is transmitted to the user terminal 4, and it becomes possible to generate and display a DTM from which the erroneous recognition point data 27 has been removed, based on the updated removed terrain data 29.

[0120] Second Embodiment A terrain data generation system according to a second embodiment of the present technology will be described. In the following description, the description of the same parts as those in the configuration and operation of the terrain data generation system 1 described in the above embodiment will be omitted or simplified.

[0121] 15 is a schematic diagram showing an example of the structure of attribute-assigned point cloud data according to this embodiment. In this embodiment, not only attribute information but also the degree of unnecessaryness of the attribute information is added as metadata to the attribute-assigned point cloud data 31.

[0122] In this example, the degree of unnecessary material is calculated as a weighted value according to the likelihood of image recognition (an evaluation index of the accuracy of recognition) in image recognition processing using an artificial intelligence model. The likelihood of image recognition is a parameter that can also be used as the reliability of attribute information.

[0123] For example, for point data recognized as an object to be removed ("building," "plant," "vehicle"), a basic unwanted substance degree of "1.0" is set. Then, the unwanted substance degree of the point data is calculated by multiplying this basic unwanted substance degree by a coefficient corresponding to the likelihood of image recognition for the point data. For example, if the likelihood is "0.8," then the unwanted substance degree of "0.8" is calculated for the basic unwanted substance degree = "1.0."

[0124] It is possible to uniformly set the unwanted matter degree of point data recognized as an object ("ground") that does not correspond to an object to be removed to, for example, "0." Alternatively, as in the case of an object to be removed, a value other than 0 (but smaller than the basic unwanted matter degree of the object to be removed) may be set as the basic unwanted matter degree, and a value obtained by weighting the basic unwanted matter degree according to the likelihood of image recognition may be calculated as the unwanted matter degree.

[0125] Since the unwanted object degree is added in this way, it becomes possible to perform various processes based on the unwanted object degree. For example, in the removal process, it is possible to remove point data that has been assigned attribute information of "buildings," "vegetation," and "vehicles" and has an unwanted object degree higher than a predetermined threshold. It is also possible to remove data of objects to be removed by only threshold processing of the unwanted object degree without referring to the attribute information.

[0126] Furthermore, by operating a predetermined control bar, the user 7 can adjust the threshold value for the degree of unnecessaryness, which is the criterion for determining whether or not something is to be removed. By adjusting the threshold value, it is possible to adjust, for example, the number of point data 20 to be removed as removal targets (i.e., the number of point data 20 included in the DTM).

[0127] The data update unit 12 updates the attribute-assigned point cloud data shown in Fig. 15 based on the determination result by the determination unit 11. For example, the unnecessary flag shown in Fig. 11 and the degree of misrecognition shown in Fig. 12 are added as metadata.

[0128] 15 may be changed (overwritten) based on the determination result by the determination unit 11. For example, point data 20 of the object to be removed (point data 24 to 26 having attribute information of "building," "plant," or "vehicle") that is closest to the detected erroneous recognition point data 27 is selected. The attribute information and unwanted object degree of the point data 20 are overwritten as the attribute information and unwanted object degree of the detected erroneous recognition point data 27.

[0129] Also, suppose that the attribute information of the point data 20 detected as data to be left unremoved (misrecognized point data 27) is one of "building," "vegetation," or "vehicle," and that the unwantedness degree is a low value (it was not removed because of its low unwantedness degree). In this case, it is possible to perform a data update in which only the unwantedness degree is overwritten while maintaining the attribute information. For example, the unwantedness degree of the point data 20 of the object to be removed that is closest to the misrecognized point data 27 is stored. Alternatively, the unwantedness degree may be increased so that it exceeds a threshold value.

[0130] By performing the data update as described above, it becomes possible to generate and display a DTM in which the data of the object to be removed has been removed with high accuracy, as shown in Fig. 13. It also becomes possible to generate and display attribute-assigned point cloud data in which misrecognition (recognition omissions) has been reduced, as shown in Fig. 14.

[0131] As described above, in the terrain data generation system described in each of the above embodiments, first distance information is calculated for each point data 20 included in target terrain data 19 indicating the terrain of the target area TA, relating to the distance to nearby point data 20. Also, second distance information is calculated for each point data 20 included in removal-processed terrain data 29, which has been processed to remove point data of objects to be removed from the target terrain data 19, relating to the distance to nearby data. Then, based on the first distance information and the second distance information, a determination is made as to whether each data item included in the removal-processed terrain data 29 is to be removed. Based on the result of this determination, it is possible to remove the point data 20 of objects to be removed with high accuracy.

[0132] For example, at a surveying site, etc., it is often necessary to create a DTM, which is a model that illustrates the elevation of the ground surface of the environment, excluding vegetation and buildings. To create a DTM, a process is performed to remove areas that recognize vegetation, buildings, construction machinery, etc., leaving only the ground surface. In this case, due to the accuracy of the area removal, some objects other than the ground surface may remain as noise.

[0133] For example, as disclosed in Patent Document 1, consider a case where a technology is implemented that performs different processing for each recognition target in order to determine whether the area in which an object to be removed, such as a building or construction machinery, is recognized includes an area different from the recognition target.

[0134] For example, if the recognition target is the ground or vegetation, height information for that area is obtained from the target terrain data 19, and an area outside the range of thresholds (upper limit / lower limit) previously set to correspond to the ground or vegetation is determined to be an area different from the recognition target. If the recognition target is a building, a contour extraction process is performed, and an area outside the range is determined to be an area different from the recognition target by determining whether it is inside or outside the building.

[0135] The problems with this method are as follows. First, it is difficult to set a height threshold (upper / lower limit) to accurately distinguish areas that differ from the recognition target. Second, the extraction system used in this method requires the use of different extraction and noise removal systems for each recognition target, such as the ground, vegetation, and buildings. While the ground and vegetation are identified using RGB values ​​and height information, buildings are identified by contour processing and straight line extraction from binary images. Therefore, it is difficult to extract unexpected objects.

[0136] This technology can detect misrecognized point data 27 based on the difference between the average neighborhood distances in the target terrain data 19 and the removed terrain data 29. In other words, there is no need to label or classify existing objects. The necessary parameters are a threshold value related to the difference between the first distance information and the second distance information, a threshold value related to the degree of misrecognition, and the like, and there is no need to set a threshold value for each individual object. Furthermore, this technology does not require detecting differences in color or height (or contour) between objects, and can be applied to complex objects.

[0137] By applying this technology, it is possible to detect areas outside the target recognition area in terrain data due to misrecognition or errors. This technology can reduce noise in ground extraction, making it possible to generate a highly accurate DTM without noise.

[0138] Other Embodiments The present technology is not limited to the above-described embodiments, and various other embodiments can be realized.

[0139] In the above embodiment, the target terrain data is configured from three-dimensional point cloud data. However, the present technology is not limited to this, and can be applied to both cases where the target terrain data is configured from arbitrary point cloud data and cases where the target terrain data is configured from arbitrary image data.

[0140] As long as each piece of data (point data or pixel data) included in the target terrain data specifies positional information corresponding to the terrain and the distance between data can be calculated, various forms of terrain data can be used as one embodiment of the target terrain data related to this technology.

[0141] Furthermore, the position information corresponding to the terrain is not necessarily limited to the case where three-dimensional position information is defined. For example, terrain data in which only two-dimensional position information such as longitude and latitude is defined can also be used as one embodiment of the target terrain data according to the present technology.

[0142] Examples of image data to which the present technology can be applied include orthophoto data and DSM (Digital Surface model) image data.

[0143] In the above, the distance between the point data 20 is calculated based on the distance in each of the three directions of X, Y, and Z, as shown in Equation 1. That is, the first distance information and the second distance information are calculated based on the distance in each of the three directions of X, Y, and Z.

[0144] Without being limited to this, the first distance information may be calculated as a distance in a predetermined direction to nearby data, and the second distance information may be calculated as a distance in the same predetermined direction to nearby data.

[0145] For example, the first distance information and the second distance information, which are the average of the neighborhood distances, may be calculated based on the distance in one of the X direction, Y direction, and Z direction. For example, the first distance information and the second distance information may be calculated based on the distance in the vertical direction (Z direction), and a determination as to whether or not the erroneous recognition point data 27 is to be removed may be performed.

[0146] For example, when the target terrain data is composed of three-dimensional point cloud data or the like, point data that becomes noise is often points that are far from the ground. Therefore, by calculating the first distance information and the second distance information based on the distance in the vertical direction (Z direction), it is possible to detect noise that is far from the ground with high accuracy. Furthermore, since it is only necessary to perform calculations using the distance in one direction, it is possible to reduce the amount of calculation.

[0147] Instead of the above equation (1), the distance between the point data 20 can also be calculated using the following equation.

[0148]

[0149] In equation 3, a, b, and c are weighting coefficients. By adjusting the values ​​of a, b, and c, it is possible to weight the distance in each of the three mutually orthogonal directions, that is, the X direction, the Y direction, and the Z direction. In other words, by weighting the distance in each of the X direction, the Y direction, and the Z direction, it is possible to calculate the first distance information and the second distance information.

[0150] The X direction, Y direction, and Z direction in the above embodiment correspond to an embodiment of the vertical direction, a first direction perpendicular to the vertical direction, and a second direction perpendicular to each of the vertical direction and the first direction according to the present technology.

[0151] The direction for calculating the distance and the magnitude of the weighting may be set appropriately based on, for example, the specific shape of the target area and the type of object to be removed from the target area.

[0152] The number of data items included in the target terrain data may be enormous. For example, if the target terrain data is composed of three-dimensional point cloud data, the target terrain data may be composed of an extremely large number of point data items.

[0153] In such a case, the target terrain data and the removed target terrain data may be divided in a similar manner, first distance information and second distance information may be calculated for each divided area, and a determination may be made based on the first distance information and second distance information.

[0154] That is, the distance information calculation unit divides each of the target terrain data and the removed terrain data in a similar manner to generate a plurality of divided target terrain data and a plurality of divided and removed terrain data. Then, calculation of first distance information is performed for each of the plurality of divided target terrain data, and calculation of second distance information is performed for each of the plurality of divided and removed terrain data.

[0155] The determination unit performs a determination based on the first distance information and the second distance information for each of the plurality of divided and removed topographical data. This makes it possible to reduce the amount of calculation required to calculate the first distance information and the second distance information, thereby suppressing the processing load. When performing data update or the like based on the determination results, it is possible to perform processing for the entire target area by integrating the determination results performed for each of the plurality of divided and removed topographical data.

[0156] The method of dividing the terrain data may use any algorithm, such as division using a kd tree. Furthermore, a data dividing unit may be configured separately from the distance information calculation unit, which divides the target terrain data and the removed terrain data in a similar manner to generate a plurality of divided target terrain data and a plurality of divided and removed terrain data.

[0157] In the above embodiment, the neighborhood distance average is calculated as the first distance information and the second distance information. However, the present invention is not limited to this, and other parameters such as the variance value or standard deviation of the distance may be used as the first distance information and the second distance information relating to the distance to the neighborhood data.

[0158] This technology can also be applied to a configuration in which the server device 2 shown in Figure 1 is located in a remote location away from the target area TA and the location of the user 7, and is connected to the sensor 5 and user terminal 4 via a network so that they can communicate with each other.

[0159] FIG. 16 is a block diagram showing an example of the hardware configuration of a computer 60 that can be used as the server device 2 and the user terminal 4.

[0160] The computer 60 includes a CPU 61, a ROM 62, a RAM 63, an input / output interface 65, and a bus 64 interconnecting these components. The input / output interface 65 is connected to a display unit 66, an input unit 67, a storage unit 68, a communication unit 69, a drive unit 70, and other components. The display unit 66 is a display device using, for example, an LCD or EL display. The input unit 67 is a keyboard, a pointing device, a touch panel, or other operating device. If the input unit 67 includes a touch panel, the touch panel may be integrated with the display unit 66. The storage unit 68 is a non-volatile storage device such as a HDD, flash memory, or other solid-state memory. The drive unit 70 is a device capable of driving a removable storage medium 71 such as an optical storage medium or magnetic recording tape. The communication unit 69 is a modem, router, or other communication device connectable to a LAN, WAN, or the like for communicating with other devices. The communication unit 69 may communicate via either a wired or wireless connection. The communication unit 69 is often used separately from the computer 60. Information processing by the computer 60 having the above-described hardware configuration is realized by cooperation between software stored in the storage unit 68 or the ROM 62, etc. and the hardware resources of the computer 60. Specifically, the information processing method according to the present technology is realized by loading a program constituting the software stored in the ROM 62, etc., into the RAM 63 and executing it. The program is installed in the computer 60 via, for example, the recording medium 71. Alternatively, the program may be installed in the computer 60 via a global network, etc. Alternatively, any computer-readable, non-transitory storage medium may be used.

[0161] The information processing method (topographical data generation method) and program according to the present technology may be executed by cooperation among multiple computers connected to each other via a network or the like, thereby constructing an information processing system or information processing device according to the present technology. In other words, the information processing method and program according to the present technology can be executed not only in a computer system composed of a single computer, but also in a computer system in which multiple computers operate in conjunction with each other. In this disclosure, a "system" refers to a collection of multiple components (devices, modules (parts), etc.), regardless of whether all the components are contained in the same housing. Therefore, both multiple devices housed in separate housings and connected via a network and a single device housed in a single housing with multiple modules are systems.

[0162] The execution of the information processing method and program according to the present technology by a computer system includes both cases where, for example, acquisition of target terrain data, acquisition of attribute-assigned terrain data, acquisition of removed terrain data, calculation of first distance information, calculation of second distance information, determination of whether or not a terrain is subject to removal, data update, etc. are performed by a single computer, and cases where each process is performed by a different computer. Furthermore, the execution of each process by a specific computer includes having another computer execute part or all of the process and obtaining the results. In other words, the information processing method and program according to the present technology can also be applied to a cloud computing configuration in which a single function is shared and processed jointly by multiple devices via a network.

[0163] The configurations of the terrain data generation system, server device, user terminal, sensor, and mobile body, and the processing flows of the generation of target terrain data, generation of attribute-assigned terrain data, generation of removed terrain data, calculation of first distance information, calculation of second distance information, determination of whether or not a terrain is subject to removal, data update, etc., which have been described with reference to the drawings, are merely one embodiment and can be modified as desired without departing from the spirit of the present technology. In other words, any other configurations, algorithms, etc. for implementing the present technology may be adopted.

[0164] In this disclosure, terms such as "about," "approximately," "almost," and "roughly" may be used as appropriate to facilitate understanding of the description. However, there is no clear difference between using and not using terms such as "about," "approximately," "almost," and "approximately." In other words, in this disclosure, concepts that define shape, size, positional relationship, state, etc., such as "center," "middle," "uniform," and "equal," are concepts that include "substantially center," "substantially central," "substantially uniform," and "substantially equal." For example, states that fall within a predetermined range (e.g., a range of ±10%) based on "completely centered," "completely central," "completely uniform," and "completely equal" are also included. Therefore, even if terms such as "approximately," "almost," and "approximately" are not used, concepts expressed by adding "approximately," "almost," and "approximately" may be included. Conversely, states expressed by adding terms such as "approximately," "almost," and "approximately" do not necessarily exclude perfect states.

[0165] In the present disclosure, expressions using "than", such as "greater than A" and "smaller than A", are expressions that comprehensively include both concepts that include the case where it is equivalent to A and concepts that do not include the case where it is equivalent to A. For example, "greater than A" is not limited to cases that do not include equivalent to A, but also includes "A or greater". Furthermore, "smaller than A" is not limited to "less than A" but also includes "A or less". When implementing the present technology, specific settings and the like can be appropriately adopted from the concepts included in "greater than A" and "smaller than A" so that the effects described above can be achieved.

[0166] It is also possible to combine at least two of the features of the present technology described above. That is, the various features described in each embodiment may be arbitrarily combined without distinguishing between the embodiments. Furthermore, the various effects described above are merely examples and are not intended to be limiting, and other effects may also be achieved.

[0167] The present technology can also be configured as follows: (1) An information processing system comprising: a first acquisition unit that acquires target terrain data indicating the terrain of a target area generated based on sensing data of the target area; a second acquisition unit that acquires removed terrain data in which a process of removing data of objects to be removed has been performed on the acquired target terrain data; a distance information calculation unit that calculates first distance information regarding the distance between each piece of data included in the target terrain data and neighboring data, and calculates second distance information regarding the distance between each piece of data included in the removed terrain data and neighboring data; and a determination unit that determines whether each piece of data included in the removed terrain data is a removal target based on the calculated first distance information and second distance information. (2) The information processing system according to (1), wherein the determination unit determines, for each piece of data included in the removed terrain data, based on a difference between the first distance information and the second distance information. (3) The information processing system according to (1) or (2), wherein the determination unit performs at least one of determining whether or not a data item is a removal target and determining the degree to which it is a removal target. (4) The information processing system according to any one of (1) to (3), wherein the distance information calculation unit calculates, for each data item included in the target terrain data, an average of distances to a predetermined number of nearby data items as the first distance information, and calculates, for each data item included in the removal-processed terrain data, an average of distances to the predetermined number of nearby data items as the second distance information. (5) The information processing system according to any one of (1) to (4), wherein the determination unit determines, for each data item included in the removal-processed terrain data, that it is a removal target when a difference between the first distance information and the second distance information is greater than a predetermined threshold.(6) The information processing system according to any one of (1) to (5), wherein the determination unit determines that, for each piece of data included in the removal-processed terrain data, the greater the difference between the first distance information and the second distance information, the greater the likelihood that the data is to be removed. (7) The information processing system according to (4), wherein the distance information calculation unit sets the predetermined number based on the density of at least one of the target terrain data and the removal-processed terrain data. (8) The information processing system according to any one of (1) to (7), wherein the second acquisition unit acquires attribute-assigned terrain data in which attribute information of an object existing in the target area is assigned to each piece of data included in the target terrain data, and acquires the removal-processed terrain data by removing data assigned attribute information of the object to be removed, which is included in the attribute-assigned terrain data. (9) The information processing system according to (8), further comprising: a data updating unit that updates at least one of the attribute-assigned terrain data and the removed terrain data based on the determination result by the determining unit. (10) The information processing system according to any one of (1) to (9), wherein the distance information calculation unit calculates distance information relating to the distance in a predetermined direction to nearby data as the first distance information, and calculates distance information relating to the distance in the predetermined direction to nearby data as the second distance information. (11) The information processing system according to (10), wherein the predetermined direction is a vertical direction. (12) The information processing system according to any one of (1) to (11), wherein the distance information calculation unit calculates the first distance information and the second distance information by weighting the distance in each of three mutually orthogonal directions. (13) The information processing system according to (12), wherein the three mutually orthogonal directions are a vertical direction, a first direction orthogonal to the vertical direction, and a second direction orthogonal to each of the vertical direction and the first direction.(14) An information processing system according to any one of (1) to (13), wherein the distance information calculation unit generates a plurality of pieces of divided terrain data and a plurality of pieces of divided and removed terrain data by dividing the target terrain data and the removed terrain data in a similar manner, calculates the first distance information for each of the plurality of pieces of divided terrain data, and calculates the second distance information for each of the plurality of pieces of divided and removed terrain data, and the determination unit performs the determination based on the first distance information and the second distance information for each of the plurality of pieces of divided and removed terrain data. (15) An information processing system according to any one of (1) to (14), wherein the target terrain data is composed of point cloud data or image data. (16) An information processing method executed by a computer system, which includes: acquiring target terrain data indicating the terrain of a target area generated based on sensing data of the target area; acquiring removed terrain data in which a process for removing data of objects to be removed has been performed on the acquired target terrain data; calculating first distance information regarding the distance between each piece of data included in the target terrain data and nearby data; calculating second distance information regarding the distance between each piece of data included in the removed terrain data and nearby data; and determining whether each piece of data included in the removed terrain data is a target for removal based on the calculated first distance information and second distance information.(17) An information processing device comprising: a first acquisition unit that acquires target terrain data indicating the terrain of a target area generated based on sensing data of the target area; a second acquisition unit that acquires removed terrain data in which a process of removing data of objects to be removed has been performed on the acquired target terrain data; a distance information calculation unit that calculates first distance information regarding the distance between each piece of data included in the target terrain data and nearby data, and calculates second distance information regarding the distance between each piece of data included in the removed terrain data and nearby data; and a determination unit that determines whether each piece of data included in the removed terrain data is a target for removal based on the calculated first distance information and second distance information.

[0168] TA...Target area 1...Terrain data generation system 2...Server device 4...User terminal 6...Mobile object 7...User 14...Ground surface 15...Building 16...Vegetation 17...Vehicle 19...Target terrain data 20...Point data 22...Attribute-assigned point cloud data 23...Point data assigned with ground attribute information 24...Point data assigned with building attribute information 25...Point data assigned with vegetation attribute information 26...Point data assigned with vehicle attribute information 27...Misrecognized point data 29...Terrain data after removal processing 31...Attribute-assigned point cloud data 60...Computer

Claims

1. An information processing system comprising: a first acquisition unit that acquires target terrain data that indicates the terrain of a target area, generated based on sensing data of the target area; a second acquisition unit that acquires removed terrain data in which a process for removing data of objects to be removed has been performed on the acquired target terrain data; a distance information calculation unit that calculates first distance information regarding the distance between each piece of data included in the target terrain data and nearby data, and calculates second distance information regarding the distance between each piece of data included in the removed terrain data and nearby data; and a determination unit that determines whether each piece of data included in the removed terrain data is a target for removal based on the calculated first distance information and second distance information.

2. An information processing system according to claim 1, wherein the determination unit performs the determination for each data included in the removed topographical data based on the difference between the first distance information and the second distance information.

3. An information processing system according to claim 1, wherein the determination unit executes at least one of determining whether or not an object is a removal target and determining the degree to which it is a removal target.

4. An information processing system as described in claim 1, wherein the distance information calculation unit calculates the average distance between each piece of data included in the target terrain data and a predetermined number of pieces of data nearby as the first distance information, and calculates the average distance between each piece of data included in the removed terrain data and the predetermined number of pieces of data nearby as the second distance information.

5. An information processing system as described in claim 1, wherein the determination unit determines that each data included in the removed terrain data is to be removed when the difference between the first distance information and the second distance information is greater than a predetermined threshold.

6. An information processing system according to claim 1, wherein the determination unit determines that for each data included in the removed terrain data, the greater the difference between the first distance information and the second distance information, the greater the likelihood that the data is to be removed.

7. An information processing system according to claim 4, wherein the distance information calculation unit sets the predetermined number based on the density of at least one of the target terrain data and the removed terrain data.

8. An information processing system as described in claim 1, wherein the second acquisition unit acquires attribute-assigned terrain data in which attribute information of objects existing in the target area is assigned to each piece of data included in the target terrain data, and acquires the removed terrain data by removing data that is assigned attribute information of the object to be removed and that is included in the attribute-assigned terrain data.

9. An information processing system according to claim 8, further comprising a data updating unit that updates at least one of the attribute-added terrain data and the removed terrain data based on the determination result by the determination unit.

10. An information processing system according to claim 1, wherein the distance information calculation unit calculates distance information relating to the distance in a predetermined direction to nearby data as the first distance information, and calculates distance information relating to the distance in the predetermined direction to nearby data as the second distance information.

11. An information processing system according to claim 10, wherein the predetermined direction is the vertical direction.

12. An information processing system according to claim 1, wherein the distance information calculation unit calculates the first distance information and the second distance information by weighting the distance in each of three mutually orthogonal directions.

13. An information processing system according to claim 12, wherein the three mutually orthogonal directions are a vertical direction, a first direction orthogonal to the vertical direction, and a second direction orthogonal to each of the vertical direction and the first direction.

14. An information processing system as described in claim 1, wherein the distance information calculation unit generates a plurality of divided target terrain data and a plurality of divided and removed terrain data by dividing each of the target terrain data and the removed terrain data in a similar manner, calculates the first distance information for each of the plurality of divided target terrain data, and calculates the second distance information for each of the plurality of divided and removed terrain data, and the determination unit performs the determination based on the first distance information and the second distance information for each of the plurality of divided and removed terrain data.

15. An information processing system according to claim 1, wherein the target topographical data is composed of point cloud data or image data.

16. An information processing method executed by a computer system, which comprises: acquiring target terrain data indicating the terrain of a target area generated based on sensing data of the target area; acquiring removed terrain data in which a process for removing data of objects to be removed has been performed on the acquired target terrain data; calculating first distance information regarding the distance between each piece of data included in the target terrain data and nearby data; calculating second distance information regarding the distance between each piece of data included in the removed terrain data and nearby data; and determining whether each piece of data included in the removed terrain data is subject to removal based on the calculated first distance information and second distance information.

17. An information processing device comprising: a first acquisition unit that acquires target terrain data indicating the terrain of a target area generated based on sensing data of the target area; a second acquisition unit that acquires removed terrain data in which a process of removing data of objects to be removed has been performed on the acquired target terrain data; a distance information calculation unit that calculates first distance information regarding the distance between each piece of data included in the target terrain data and nearby data, and calculates second distance information regarding the distance between each piece of data included in the removed terrain data and nearby data; and a determination unit that determines whether each piece of data included in the removed terrain data is a target for removal based on the calculated first distance information and second distance information.

Images