JP2023128778A - Information processing device, control method, program and storage medium - Google Patents

Abstract

To suitably estimate the height of a reference surface.SOLUTION: A controller 13 of an information processing device 1 mainly includes acquisition means, measurement means, and estimation means. The acquisition means acquires measurement data outputted by a measurement device. The measurement means acquires position information including a height direction of an object existing within a visual field range of the measurement device on the basis of the measurement data. The estimation means estimates the height of a reference surface within the visual field range on the basis of the position information.SELECTED DRAWING: Figure 6

Description

The present disclosure relates to processing of measured data.

2. Description of the Related Art Conventionally, a laser radar device has been known that irradiates a detection space with a pulse of laser light and detects an object within the detection space based on the level of the reflected light. For example, Patent Document 1 discloses that by appropriately controlling the emission direction (scanning direction) of repeatedly emitted light pulses, the surrounding space is scanned, and by observing the returned light, information about objects existing in the surroundings is obtained. A lidar is disclosed that generates point cloud data representing information such as distance and reflectance. Further, Patent Document 2 discloses a technology that detects an object frame surrounding an object based on point cloud data and performs tracking processing of the object by associating the object frames in time series. Further, Patent Document 3 discloses a technique for interpolating background data by assuming a virtual ground for a position for which a distance value is not obtained when acquiring background data.

JP2018-009831A JP2020-042800A JP 2019-045199 Publication

When a measurement device such as a lidar is installed facing a reference surface such as the ground, the range in which measurement data on the reference surface can be obtained is limited to the area near the measurement device, and data on the reference surface beyond that cannot be obtained and the reference surface There was a problem in that it was not possible to determine the height of the

The present disclosure has been made to solve the above-mentioned problems, and its main purpose is to provide an information processing device that can suitably estimate the height of a reference plane.

The claimed invention is:
an acquisition means for acquiring measurement data output by the measurement device;
a measuring means for acquiring positional information including a height direction of an object existing within a visual field of the measuring device based on the measurement data;
Estimating means for estimating the height of the reference plane within the visual field based on the position information;
An information processing device comprising:

In addition, the claimed invention is
The computer is
Obtain the measurement data output by the measurement device,
Based on the measurement data, obtain position information including the height direction of an object existing within the field of view of the measurement device;
estimating the height of the reference plane within the field of view based on the position information;
This is a control method.

In addition, the claimed invention is
Obtain the measurement data output by the measurement device,
Based on the measurement data, obtain position information including the height direction of an object existing within the field of view of the measurement device;
This is a program that causes a computer to execute a process of estimating the height of a reference plane within the field of view based on the position information.

1 is a schematic configuration of a rider unit according to an embodiment. An example of installing lidar to measure vehicles on the road is shown. 1 is a block configuration diagram showing an example of a hardware configuration of an information processing device. FIG. A virtually set grid and a moving trajectory of a tracked object are shown in a two-dimensional space viewed from directly above the field of view. Shows the bottom height histogram for a certain grid mass. It is an example of a flowchart showing a processing procedure of an information processing device.

According to a preferred embodiment of the present invention, the information processing device includes an acquisition unit that acquires measurement data output from a measurement device, and a height of an object existing within a field of view of the measurement device based on the measurement data. The apparatus includes a measuring means for acquiring positional information including a direction, and an estimating means for estimating the height of the reference plane within the visual field based on the positional information. With this aspect, the information processing device can accurately estimate the height of the reference plane within the field of view of the measuring device.

In one aspect of the information processing device, the measuring means acquires the positional information indicating the lower end position of the object, and the estimating means calculates the height of the lower end position from the reference surface based on the height of the lower end position indicated by the positional information. Estimate the height of. According to this aspect, the information processing device can accurately estimate the height of the reference plane based on the measured height of the lower end position of the object.

In another aspect of the information processing device, the measuring means acquires the position information for the plurality of objects, and the estimating means sets a virtual grid on a horizontal plane within the field of view, and the estimating means sets a virtual grid on a horizontal plane within the field of view, and The height of the reference plane is estimated for each grid based on the position information distributed to each grid grid. With this aspect, the information processing device can accurately estimate the height of the reference plane at the passing position based on the position information of the passing object.

In another aspect of the information processing device, the estimation means generates a histogram in which the positions in the height direction indicated by the position information are aggregated for each square, and the lowest position in the height direction in the histogram. The height of the reference plane is estimated for each square based on at least one of the value and the mode. With this aspect, the information processing device can accurately estimate the height of the reference plane for each square.

In another aspect of the information processing device, the measuring means tracks the object and generates the position information of the tracked object in time series. In a preferred example, the estimating means selects the position information to be used for estimating the height of the reference plane based on the length of time during which the tracking was performed. With this aspect, the information processing device can accurately estimate the height of the reference plane based on the position information of the tracked object.

In another aspect of the information processing device, the estimating means selects the position information to be used for estimating the height of the reference plane based on the type of the object. With this aspect, the information processing device can select position information of an object that tends to be measured stably and accurately estimate the height of the reference plane.

In another aspect of the information processing device, the estimating means determines whether or not the position information of the object is used to estimate the height of the reference plane based on the presence or absence of another object that blocks the object. With this aspect, the information processing device can accurately estimate the height of the reference plane based on the position information of the object measured without being blocked by other objects.

In another aspect of the information processing device, the measuring means interpolates the measurement data representing the object based on model information of the object, and generates the position information based on the interpolated data of the object. do. With this aspect, the information processing device can generate accurate position information even for objects that are measured while being blocked by other objects, and can accurately estimate the height of the reference plane.

According to another preferred embodiment of the present invention, there is provided a control method executed by a computer, in which the computer acquires measurement data output from a measurement device, and based on the measurement data, determines whether the measurement data is within the field of view of the measurement device. positional information including a height direction of an object existing in the field of view is acquired, and the height of the reference plane within the field of view is estimated based on the positional information. By executing this control method, the computer can accurately estimate the height of the reference plane within the visual field of the measuring device.

According to another preferred embodiment of the present invention, measurement data output by a measurement device is acquired, and based on the measurement data, position information including the height direction of an object existing within the field of view of the measurement device is obtained. This is a program that causes a computer to execute a process of estimating the height of a reference plane within the field of view based on the acquired position information. By executing this program, the computer can accurately estimate the height of the reference plane within the field of view of the measuring device. Preferably, the program is stored on a storage medium.

Hereinafter, preferred embodiments of the present invention will be described with reference to the drawings.

(1) Overview of lidar unit
FIG. 1 shows a schematic configuration of a rider unit 100 according to an embodiment. The lidar unit 100 includes an information processing device 1 that performs processing related to data generated by the sensor group 2, and a sensor group 2 that includes at least a lidar (Light Detection and Ranging or Laser Illuminated Detection and Ranging) 3. . The lidar unit 100 is installed indoors or outdoors, and accurately estimates the height of the ground (also referred to as "ground height") based on the results of tracking objects in time series. FIG. 2 shows an example of how the rider 3 is installed. The lidar 3 shown in FIG. 2 has a viewing range "Rv" which is a range in which the lidar 3 can measure distances, and measures objects (vehicles and pedestrians in FIG. 2) existing within the viewing range Rv in chronological order. .

The information processing device 1 is electrically connected to the sensor group 2 by wire or wirelessly, and processes data output by various sensors included in the sensor group 2. In this embodiment, the information processing device 1 performs a process of tracking an object based on point cloud data output by the lidar 3 and a process of estimating the ground height based on the tracking result. The information processing device 1 is fixedly installed, for example, in a state where it is housed in a housing together with the rider 3. Note that the information processing device 1 may be provided integrally with the rider 3 as an electronic control device for the rider 3, or may be provided at a position remote from the rider 3 while being capable of data communication with the rider 3.

The lidar 3 discretely measures the distance to an object in the outside world by emitting a pulsed laser, which is an infrared laser, while changing the angle within a predetermined angular range in the horizontal and vertical directions. In this case, the lidar 3 includes an irradiation section that irradiates laser light while changing the irradiation direction (i.e., scanning direction), a light receiving section that receives reflected light (scattered light) of the irradiated laser light, and a light receiving section that receives the light output from the light receiving section. and an output section that outputs data based on the signal. Then, using the lidar 3 as a reference point, the lidar 3 measures the distance to the object irradiated with the pulsed laser (measurement distance) and the received intensity of the reflected light (reflection intensity value) for each measurement direction (that is, the emission direction of the pulsed laser). Generate the point cloud data shown in . In this case, the lidar 3 calculates the time from when the pulse laser is emitted until the light receiving unit detects the reflected light as the time of flight of light, and calculates the measured distance according to the calculated flight time. demand. Hereinafter, point cloud data obtained by one measurement over the entire viewing range Rv will be referred to as point cloud data for one frame.

Here, the point cloud data can be regarded as an image in which each measurement direction is a pixel, and the measurement distance and reflection intensity value in each measurement direction are pixel values. In this case, the emission direction of the pulsed laser at the elevation/depression angle differs in the vertical arrangement of the pixels, and the emission direction of the pulsed laser at the horizontal angle differs in the horizontal arrangement of the pixels. Then, for each pixel, a coordinate value in a three-dimensional coordinate system with the lidar 3 as a reference is determined based on the corresponding pair of emission direction and measurement distance. Hereinafter, these coordinate values will be expressed as (x, y, z), where the set of x and y coordinates represents the position in the horizontal direction, and the z coordinate represents the position in the height direction.

Note that the lidar 3 is not limited to the above-mentioned scan type lidar, but may be a flash type lidar that generates three-dimensional data by diffusing laser light into the field of view of a two-dimensional array of sensors. Hereinafter, the point measured by being irradiated with the pulsed laser emitted from the irradiation unit (and its measurement data) will also be referred to as a "point to be measured." The rider 3 is an example of a "measuring device" in the present invention.

In addition to the rider 3, the sensor group 2 may include various external sensors and/or internal sensors. For example, the sensor group 2 may include a GNSS (Global Navigation Satellite System) receiver, etc. necessary for generating position information.

(2) Configuration of information processing device
FIG. 3 is a block diagram showing an example of the hardware configuration of the information processing device 1. As shown in FIG. The information processing device 1 mainly includes an interface 11, a memory 12, and a controller 13. Each of these elements is interconnected via a bus line.

The interface 11 performs interface operations related to data exchange between the information processing device 1 and external devices. In this embodiment, the interface 11 acquires output data from the sensor group 2 such as the rider 3 and supplies it to the controller 13. The interface 11 may be a wireless interface such as a network adapter for wireless communication, or may be a hardware interface for connecting to an external device via a cable or the like. Further, the interface 11 may perform interface operations with various peripheral devices such as an input device, a display device, and a sound output device.

The memory 12 includes various types of volatile memory and nonvolatile memory, such as RAM (Random Access Memory), ROM (Read Only Memory), hard disk drive, and flash memory. The memory 12 stores programs for the controller 13 to execute predetermined processes. Note that the program executed by the controller 13 may be stored in a storage medium other than the memory 12.

Furthermore, the memory 12 stores information necessary for the controller 13 to execute predetermined processing. For example, in this embodiment, the memory 12 stores tracking history information ITR. The tracking history information ITR is information indicating the history of object tracking results, and is updated every frame period. For example, the tracking history information ITR includes time information indicating the detection time of an object to be tracked (also called a "tracked object"), identification information of the tracked object (also called a "tracked object ID"), A record is recorded that includes classification (class) information of the tracked object and position information indicating the detected position of the tracked object. Here, if the same tracked object exists at different processing times, the tracked objects are managed by the same tracked object ID. Further, the above-mentioned position information includes at least coordinate information of the lower end position of the tracked object (that is, x, y, and z coordinate values indicating the lower end position).

In addition, the memory 12 may include any information necessary for object tracking processing and ground height estimation processing. For example, memory 12 may include object model information indicating a model of an object that can be a tracked object. Generally, there are cases where only a portion of an object is detected by the lidar 3 due to occlusion by other objects. In this case, the controller 13 generates data representing the entire area of the tracking object by interpolating the point cloud data of the tracking object output by the lidar 3 based on the object model information, and based on the data representing the generated overall area. , generates detected position information indicating the lower end position of the tracked object.

The controller 13 includes one or more processors such as a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), and a TPU (Tensor Processing Unit), and controls the entire information processing device 1 . In this case, the controller 13 executes various processes described below by executing programs stored in the memory 12 and the like. Functionally, the controller 13 includes a segment detection section 15, a tracking section 16, and a ground height estimating section 17.

The segment detection unit 15 extracts groups of mutually adjacent data (also referred to as "segments") from the point cloud data generated in the frame period corresponding to the current processing time, and estimates that the extracted segments represent an object. (also referred to as "object segment").

In this case, the segment detection unit 15 extracts, as an object segment, a cluster of measured points specified from the point cloud data based on an arbitrary clustering technique such as Eugridian distance clustering. Then, the segment detection unit 15 determines the class of each extracted object segment by classifying each of the extracted object segments. The classes here are the types of objects that may be detected by the rider 3, such as vehicles (including bicycles), pedestrians, and signs. In this case, the segment detection unit 15 performs the above-mentioned classification based on, for example, the size and/or shape of the segments. In this case, prior information regarding the size and/or shape of each class is stored in advance in the memory 12 or the like, and the segment detection unit 15 determines the size and/or shape of each object segment by referring to the prior information. Performs classification based on shape. Note that the segment detection unit 15 may detect and classify object segments using an object detection model learned based on deep learning (neural network). In this case, the object detection model is, for example, a model based on instance segmentation, which outputs object segment detection results and classified classes when point cloud data expressed in a predetermined tensor format is input. is learned in advance.

The tracking unit 16 performs object tracking based on time-series object segments. In this case, the tracking unit 16 uses time-series object segments and an arbitrary object tracking model to determine whether or not object segments detected in consecutive frame periods represent the same object. . Specifically, the tracking unit 16 determines whether the tracked object detected at the past processing time and the object represented by the object segment detected at the current processing time are the same object based on the sameness of classification and , the determination is made based on the error between the predicted position of the tracked object predicted by the object tracking model from the past detected position and the detected position based on the object segment at the current processing time. Note that the object tracking model may be a model based on a Kalman filter or may be a model based on deep learning. In this case, the tracking unit 16 may determine a representative point from the object segment and perform tracking based on the representative point. The representative point in this case may be the center of gravity position of the object segment, or may be a measured point corresponding to a specific part of the object. In another example, the tracking unit 16 may set a bounding box for the object segment and perform tracking based on the bounding box.

Further, the tracking unit 16 updates the tracking history information ITR based on the tracking results at each processing time for each frame period. In this case, the tracking unit 16 creates a record including, for example, time information, a tracked object ID, classification (class) information, and position information for each object segment detected in the point cloud data at the current processing time. is added to the tracking history information ITR. In this case, the tracking unit 16 changes the tracked object ID of the tracked object that was successfully tracked in the time series (that is, the tracked object deemed to be the same as the tracked object detected at a past time) to the same ID in the time series. stipulate. Note that instead of generating a record of the tracking history information ITR for each processing time, one record may be generated for each tracked object. In this case, the record of the tracking history information ITR includes time-series data indicating the position of the tracked object (including the lower end position) as position information.

The ground height estimating unit 17 estimates the ground height within the visual field Rv based on the tracking history information ITR when it is time to estimate the ground height. In this case, the ground height estimating unit 17 regards the viewing range Rv as a two-dimensional space viewed from directly above, and defines a virtual grid in which the two-dimensional space is meshed at predetermined intervals in each dimension. Then, the ground height estimating unit 17 generates a histogram that totals the height of the lower limit position of the tracked object based on the tracking history information ITR for each square of the grid (also referred to as "grid mass"), and based on the histogram. Estimate the ground height for each grid mass. Details of the processing by the ground height estimation unit 17 will be described later.

The controller 13 functions as an "obtaining means", "measuring means", "estimating means", a computer that executes a program, and the like.

Note that the processing executed by the controller 13 is not limited to being implemented by software based on a program, but may be implemented by a combination of hardware, firmware, and software. Further, the processing executed by the controller 13 may be realized using a user-programmable integrated circuit such as an FPGA (Field-Programmable Gate Array) or a microcomputer. In this case, this integrated circuit may be used to realize the program that the controller 13 executes in this embodiment.

(3) Estimation of ground height
Next, a method for estimating the ground height by the ground height estimating section 17 will be specifically explained.

FIG. 4 shows a virtually set grid and the passing position of the tracked object at each processing time in a two-dimensional space viewed from directly above the visual field Rv. In FIG. 4, grid cells that are determined to have been passed by a certain tracking object are filled in. Here, the two-dimensional space is a space composed of x and y coordinates.

The ground height estimating unit 17 recognizes the grid mass in which the tracked object tracked by the tracking unit 16 existed in time series based on the position information included in the tracking history information ITR. Then, the ground height estimating unit 17 distributes the position information of the tracking history information ITR into grid cells and adds them to the histogram. This histogram is a histogram (also called a "lower end height histogram") that is a summation of the heights of the lower end positions of the tracked objects (also called "lower end heights") that have passed for each grid square. Note that the lower end height corresponds to the z coordinate. For example, in the example of FIG. 4, the position information of the tracking history information ITR corresponding to the tracked object shown in FIG. 4 is distributed for a total of six filled-in grid cells, and added to the lower end height histogram. Then, the ground height estimating unit 17 performs the above-described processing for each of all the tracked objects registered in the tracking history information ITR, and updates the lower end height histogram. For example, if a pedestrian passes through a 1 m square grid mass in 1 second, and the rider 3 is running at 24 fps, 24 pieces of position information will be associated with the target grid mass. In that case, 24 pieces of data may be added to the lower end height histogram, or 24 average values may be added to the lower end height histogram.

FIG. 5 shows a lower end height histogram for a certain grid mass, which is generated based on the tracking history information ITR. Here, the ground height estimating unit 17 sets a bin width of a predetermined width with respect to the lower end height, and totals the frequency of the lower end position of the tracked object passing through the target grid mass for each bin.

By summing up the lower end heights of a large number of tracked objects, the lower end height histogram represents a distribution that reflects the relationship between the lower end heights of the tracked objects and the ground height. Specifically, since the lower end of the tracked object is in contact with the ground in most cases, it is thought that the lower end height most frequently represents the ground height. On the other hand, the lidar 3 may not be able to measure the lower end of the tracked object due to occlusion of the tracked object by another object. In such cases, the lower end height will not be the ground height, but will be higher than the ground height. It also indicates a high position. Furthermore, except for errors, the lower end height representing a position lower than the ground level is not measured.

Taking the above into consideration, the ground height estimating unit 17 estimates the bottom height bin representing the ground height based on at least one of the mode or the lowest value in the bottom height histogram. In the example of FIG. 5, the bin with the lower end height "H0" (that is, the bin with H0 as the median value) is the mode and the lowest value. Therefore, the ground height estimating unit 17 estimates that the lower end height H0 is the ground height in the target grid mass. Thereby, the ground height estimating unit 17 can estimate the ground height for each grid mass with high accuracy.

Here, the effects of this embodiment will be supplementarily explained. Objects such as cars do not change in height and always move in contact with the ground. In general, there are many vehicles that can be detected farther than the range in which ground points can be acquired from the point cloud data of the lidar 3. Furthermore, compared to cars, pedestrians are difficult to track because their speed changes suddenly and people approach each other, and their shape may change as they crouch or jump. On the other hand, if a large amount of data is collected, there are many cases where the robot stands upright and moves while touching the ground. Taking the above into consideration, the ground height estimating unit 17 generates a lower end height histogram for each grid mass based on the tracking results of the plurality of objects by the tracking unit 16, and statistically estimates the ground height.

Here, variations regarding the ground height estimation method will be explained.

When estimating the ground height based on the lowest value in the lower end height histogram, the ground height estimation unit 17 excludes a predetermined number of lower end heights having the lowest value from among the lower end heights to be aggregated in order to deal with errors etc. It's good to do that. In this case, the ground height estimating unit 17 estimates the lowest value among the lower end heights after exclusion to be the ground height.

Further, after estimating the ground height for each grid mass, the ground height estimating unit 17 calculates the average value of the estimated ground height of each grid mass with the estimated ground height of the adjacent grid mass as the corrected ground height. Correction processing may also be performed. This makes it possible to obtain a smoothed ground height.

In addition, as pre-processing to generate a lower end height histogram, the ground height estimation unit 17 averages the lower end position of the tracked object indicated by the time series position information in the time direction for each tracked object registered in the tracking history information ITR. It may be smoothed by calculating the moving average (for example, calculating a moving average). In this case, the ground height estimation unit 17 extracts position information from the tracking history information ITR for each tracked object, and smoothes the bottom position of the tracked object in the time series indicated by the extracted position information by statistical processing such as a moving average. , correct to the value after smoothing. Thereafter, the ground height estimation unit 17 sequentially identifies the grid square through which the tracked object passed, generates a bottom height histogram for each grid square, and estimates the ground height based on the smoothed bottom position of each tracked object. conduct.

Further, the ground height estimating unit 17 may select the position information of the tracking history information ITR used for estimating the ground height based on the tracking result of the tracked object by the tracking unit 16.

In the first example, the ground height estimation unit 17 estimates the ground height based on the tracking result of a tracked object belonging to a predetermined class. In this case, the ground height estimation unit 17 extracts position information of a tracked object belonging to a predetermined class (for example, a car) based on the classification information of the tracked object included in the tracking history information ITR. Then, the ground height estimating unit 17 generates a lower end height histogram for each grid cell from the extracted position information, and estimates the ground height. Generally, in the case of automobiles, stable tracking can be expected, but there is a problem in that tracking of pedestrians is unstable due to occlusion and the like. Therefore, the ground height estimating unit 17 can estimate the ground height with high accuracy by using the tracking results of a class (for example, a car) in which tracking is stable.

In the second example, the ground height estimating unit 17 may estimate the ground height based on the tracking results of the tracked object that has been tracked for a predetermined period of time or more instead of or in addition to using the classification information. good. In this case, the ground height estimation unit 17 calculates the tracking time length of each tracked object based on the tracking history information ITR, and extracts the position information of the tracked object whose tracking time length is equal to or longer than a predetermined time length. Then, the ground height estimating unit 17 generates a lower end height histogram for each grid cell based on the extracted position information, and estimates the ground height. This also allows the ground height estimating unit 17 to estimate the ground height with high accuracy using the tracking results of the tracked object that has been stably tracked.

In addition, in the case of a congested road, a pedestrian in a crowded downtown area, etc., the ground height estimating unit 17 uses the following as a shielding measure, based on the tracking results when there are no other objects in the distance of the tracked object. Ground height may also be estimated. In this case, for example, the ground height estimating unit 17 determines the presence or absence of another object between the tracked object and the rider 3 at each processing time based on the tracking history information ITR, and locates the tracked object where no other object exists. Extract information. Then, the ground height estimating unit 17 generates a lower end height histogram for each grid cell based on the extracted position information, and estimates the ground height.

In this way, the ground height estimating unit 17 determines whether or not the tracking result of the tracked object is used for estimating the ground height, based on the presence or absence of another object that blocks the tracked object. With this, even when the tracked object is a congested road, pedestrians in a crowded downtown area, etc., the ground height estimating unit 17 can use the tracking result of the tracked object that is not occluded to achieve high accuracy. It becomes possible to estimate the ground height.

Further, in updating the tracking history information ITR, the tracking unit 16 may generate position information representing a more accurate lower end position of the tracked object by interpolating object segments based on the object model information. In this case, the tracking unit 16 refers to object model information corresponding to the classification (class) of the target object segment, interpolates the object segment with the object model indicated by the object model information, and generates data representing the entire area of the tracked object. generate. Then, the tracking unit 16 registers position information representing the lower end position of the tracked object specified based on the data in the tracking history information ITR. Thereby, the tracking unit 16 can generate position information representing the accurate lower end position of the tracked object even if the tracked object is blocked by another object.

(4) Processing flow
FIG. 6 is an example of a flowchart showing the procedure of the ground height estimation process executed by the information processing device 1. The information processing device 1 repeatedly executes the process of this flowchart.

First, the controller 13 of the information processing device 1 acquires point cloud data measured by the lidar 3 via the interface 11 (step S11). Then, the controller 13 detects object segments and classifies objects represented by each object segment based on the point cloud data acquired in step S11 (step S12).

Next, the controller 13 tracks the object and updates the tracking history information ITR (step S13). As a result, the controller 13 generates a record including detection time information, tracked object ID, classification information, and position information indicating at least the bottom position for each object segment detected in step S12, and records the generated record. is added to the tracking history information ITR.

Next, the controller 13 determines whether it is time to estimate the ground height (step S14). For example, the controller 13 determines that it is time to estimate the ground height when a predetermined ground height estimation start condition is met. The estimation start condition may be, for example, a condition that stipulates that the number of location information included in the tracking history information ITR is greater than or equal to a predetermined number, or may be a condition that stipulates the date and time at which the estimation is to be performed. The condition may be such that estimation is started when there is a request to start estimation from 11 or the like. If it is not the timing to estimate the ground height (step S14; No), the controller 13 returns the process to step S11 and performs the update process of the tracking history information ITR.

On the other hand, if it is the timing to estimate the ground height (step S14; Yes), the controller 13 generates a lower end height histogram for each grid mass in the two-dimensional space based on the tracking history information ITR (step S15). In this case, for example, as described above, the controller 13 selects the position information of the tracking history information ITR to be aggregated in the lower end height histogram based on the tracking time, the class of the tracked object, the presence or absence of the object on the near side, etc. You may go.

Then, the controller 13 estimates the ground height for each grid mass (step S16). In this case, the controller 13 identifies the bin of the lower end height histogram representing the ground height, for example, based on at least one of the lowest value or the mode of the lower end height histogram. Further, for example, as described above, the controller 13 may perform a ground height smoothing process based on the ground heights of adjacent grid masses.

(5) Modification example
Instead of estimating the ground height, the information processing device 1 may estimate the height of an arbitrary surface (also referred to as a "reference surface") that serves as a path for people or objects.

The reference plane in this case also includes the surfaces of man-made objects such as stairs, escalators, and floors. Therefore, the rider 3 may be installed outdoors or indoors. Then, the information processing device 1 generates tracking history information ITR by performing object tracking processing based on the point cloud data generated by the lidar 3. Then, at the timing of estimating the height of the reference plane, the information processing device 1 generates a lower end height histogram for each grid mass on the horizontal plane based on the tracking history information ITR, and based on the lowest value or the mode of the lower end height histogram. , estimate the height of the reference plane for each grid mass. With this aspect as well, the information processing device 1 can suitably estimate the height of the reference plane within the visual field Rv of the rider 3.

As explained above, the controller 13 of the information processing device 1 according to the present embodiment mainly includes an acquisition means, a measurement means, and an estimation means. The acquisition means acquires measurement data output by the measurement device. The measuring means acquires position information including a height direction of an object existing within a visual field of the measuring device based on the measurement data. The estimating means estimates the height of the reference plane within the visual field based on the position information. With this aspect, the information processing device 1 can estimate the height of the reference plane within the visual field of the measuring device with high accuracy.

In the embodiments described above, the program can be stored using various types of non-transitory computer readable media and supplied to a computer, such as a controller. Non-transitory computer-readable media include various types of tangible storage media. Examples of non-transitory computer-readable media include magnetic storage media (eg, flexible disks, magnetic tape, hard disk drives), magneto-optical storage media (eg, magneto-optical disks), CD-ROMs (Read Only Memory), CD-Rs, Includes CD-R/W, semiconductor memory (for example, mask ROM, PROM (Programmable ROM), EPROM (Erasable PROM), flash ROM, and RAM (Random Access Memory).

Although the present invention has been described above with reference to Examples, the present invention is not limited to the above-mentioned Examples. The configuration and details of the present invention can be modified in various ways that can be understood by those skilled in the art within the scope of the present invention. That is, it goes without saying that the present invention includes the entire disclosure including the claims and various modifications and modifications that a person skilled in the art would be able to make in accordance with the technical idea. Furthermore, the disclosures of the cited patent documents, non-patent documents, etc. mentioned above are incorporated into this document by reference.

1 Information processing device 2 Sensor group 3 Lidar 100 Lidar unit

Claims (12)

an acquisition means for acquiring measurement data output by the measurement device;
a measuring means for acquiring positional information including a height direction of an object existing within a visual field of the measuring device based on the measurement data;
Estimating means for estimating the height of the reference plane within the visual field based on the position information;
An information processing device comprising: The measuring means acquires the position information indicating the lower end position of the object,
The information processing apparatus according to claim 1, wherein the estimating means estimates the height of the reference plane based on the height of the lower end position indicated by the position information. The measuring means acquires the position information with respect to the plurality of objects,
The estimating means sets a virtual grid on a horizontal plane within the field of view, and estimates the height of the reference plane for each square based on the position information distributed to each square of the grid. The information processing device according to item 1 or 2. The estimating means generates a histogram that aggregates the positions in the height direction indicated by the position information for each square, and based on at least one of the lowest value or the mode of the positions in the height direction in the histogram, The information processing device according to claim 3, wherein the height of the reference plane is estimated for each of the squares. The information processing device according to claim 1, wherein the measuring means tracks the object and generates the position information of the tracked object in time series. The information processing apparatus according to claim 5, wherein the estimating means selects the position information to be used for estimating the height of the reference plane based on the length of time during which the tracking was performed. 7. The information processing apparatus according to claim 1, wherein the estimating means selects the position information to be used for estimating the height of the reference plane based on the type of the object. According to any one of claims 1 to 7, the estimating means determines whether or not the position information of the object is used to estimate the height of the reference plane based on the presence or absence of another object that blocks the object. The information processing device described. The measuring means interpolates the measurement data representing the object based on model information of the object, and generates the position information based on the interpolated data of the object. The information processing device described in section. The computer is
Obtain the measurement data output by the measurement device,
Based on the measurement data, obtain position information including the height direction of an object existing within the field of view of the measurement device;
estimating the height of the reference plane within the field of view based on the position information;
Control method. Obtain the measurement data output by the measurement device,
Based on the measurement data, obtain position information including the height direction of an object existing within the field of view of the measurement device;
A program that causes a computer to execute a process of estimating the height of a reference plane within the field of view based on the position information. A storage medium storing the program according to claim 11.