WO2021075112A1 - Information processing device, information processing method, and program - Google Patents

Abstract

A region extraction unit 35 performs region recognition on map information and extracts region information indicating regions having different possibilities of presence of a moving object. In the region information, a moving object presence score indicating the ease of presence of a moving object in each recognized region is set. A characteristic shape extraction unit 36 extracts characteristic shapes of objects present in sensing data acquired by an external sensor 312. A moving object candidate extraction unit 37 sets, for the objects, moving object scores indicating moving object-likenesses on the basis of the extraction results of the characteristic shapes, and extracts a moving object candidate on the basis of the moving object scores. A determination unit 38 determines whether the moving object candidate is a moving object on the basis of at least either of the moving object score indicated by the extraction result of the moving object candidate extraction unit 37 and the moving object presence score of a region where the moving object candidate is located in the region information extracted by the region extraction unit 35. A map information processing unit 39 deletes the moving object from map information. Thus, the map information from which the moving object is removed can be rapidly generated.

Description

Information processing equipment and information processing methods and programs

This technology makes it possible to quickly generate map information excluding animals with respect to information processing devices, information processing methods and programs.

Conventionally, map information has been updated as well as self-position estimation and map information generation using distance measurement sensors and image sensors. For example, in Patent Document 1, an object is extracted and tracked from a time-series image, an optical flow is calculated, an animal body is discriminated based on the calculated optical flow, and the discriminated animal body is removed to generate map information. Is being done. Further, in Patent Document 2, the number of measurements is recorded for each point on the map information indicating the driving environment, and the position data of the point where the number of measurements is equal to or less than the threshold value is due to a measurement error or in an animal body. Since there is a high possibility that it is, it is deleted and the map information is updated.

Japanese Unexamined Patent Publication No. 2016-126662 Japanese Unexamined Patent Publication No. 2004-326264

By the way, in Patent Document 1 and Patent Document 2, a time-series image or a plurality of measurement results are required to update the map information. Therefore, if the images and measurement results for a predetermined time cannot be obtained, the animal body cannot be deleted from the map information, and therefore the map information in which the animal body is deleted cannot be promptly generated. In addition, these methods, which deal with changes in the environment, update the map information based on the self-position estimation result and the measurement result for each measurement, but it is difficult to estimate the self-position when the spatial structure changes significantly. And the animal body cannot be deleted.

Therefore, the purpose of this technology is to provide an information processing device, an information processing method, and a program that can quickly generate map information excluding animal bodies.

The first aspect of this technology is
An extraction unit that extracts the characteristic shape of an object existing in the sensing data or the area information indicating the spatial area in which the object is located in the map information.
A determination unit that determines whether the object is an animal body based on the extraction result by the extraction unit, and
An information processing device including a map information processing unit that deletes the object from the map information when the determination unit determines that the object is an animal body.

In this technology, the extraction unit extracts the characteristic shape of the object existing in the sensing data acquired by using the external sensor, or the area information indicating the spatial area in which the object is located in the map information. The extraction unit extracts the feature shape of the object using, for example, a feature shape recognition model generated in advance, and sets an animal body score indicating the animal body likeness to the object based on the extraction result. The extraction unit extracts animal body candidates based on the animal body score. In addition, the extraction unit recognizes the spatial region indicated by the sensing data using the region recognition model generated in advance, and the animal body existence score indicating the ease of existence of the animal body for each recognized region. To set.

The determination unit determines, for example, whether the animal body candidate is an animal body based on the extraction result by the extraction unit. The determination unit determines whether or not the animal body candidate is an animal body, for example, based on at least one of the animal body score of the animal body candidate and the animal body presence score of the region where the animal body candidate is located.

The map information processing unit deletes the object from the map information when the determination unit determines that the object is an animal body. Further, the map information processing unit adds the area label information regarding the recognized area to the map information based on the area recognition result of the extraction unit. The map information is generated by the map information generation unit based on the sensing data acquired by the external sensor.

In addition, the star reckoning unit detects its own position using the sensing data and the map information in which the animal body has been deleted by the map information processing unit. Further, a weighting processing unit for weighting the sensing data is provided, and the star reckoning unit detects the self-position using the sensing data weighted by the weighting processing unit. The weighting processing unit weights the sensing data indicating the animal body candidates extracted by the extraction unit. Further, the map information is provided with area label information indicating the animal body existence score indicating the ease of existence of the animal body for each area, and the weighting processing unit weights the sensing data based on the area label information. You may. For example, the weighting processing unit weights the sensing data according to at least one of the animal body presence score and the animal body score indicating the animal-likeness set as the animal body candidate based on the extraction result of the feature shape in the extraction unit. As the animal body becomes easier to exist or becomes more animal-like, the weight is reduced.

Furthermore, the dead reckoning unit that detects the self-position based on the sensing data acquired by the internal sensor, the self-position detected by the dead reckoning unit, and the self-position integration that integrates the self-position detected by the star reckoning unit. It may be provided with a part.

The second aspect of this technology is
The extraction unit extracts the characteristic shape of the object existing in the sensing data or the area information indicating the area where the object is located in the map information.
Based on the extraction result by the extraction unit, the determination unit determines whether the object is an animal body.
When the determination unit determines that the object is an animal body, the information processing method includes deleting the object from the map information by the map information processing unit.

The third aspect of this technology is
A program that causes a computer to generate map information.
The procedure for extracting the characteristic shape of an object existing in the sensing data or the area information indicating the area where the object is located in the map information, and the procedure for extracting the area information.
A procedure for determining whether the object is an animal body based on the feature shape or the extraction result of the region information, and
There is a program in which the computer executes a procedure for deleting an object determined to be an animal body from the map information.

The program of the present technology provides, for example, a storage medium, a communication medium, for example, a storage medium such as an optical disk, a magnetic disk, or a semiconductor memory, which is provided in a computer-readable format to a general-purpose computer capable of executing various program codes. Or, it is a program that can be provided by a communication medium such as a network. By providing such a program in a computer-readable format, processing according to the program can be realized on the computer.

It is a figure which illustrated the structure of the information processing system using the information processing apparatus of this technology. It is a figure which illustrated the structure of the learning block. It is a flowchart which exemplifies the operation of the learning block. It is a figure which illustrated the feature shape to recognize. It is a figure which illustrated the structure of the map generation block. It is a flowchart exemplifying the operation of the map generation block. It is a figure for demonstrating the extraction of an animal body candidate. It is a figure which illustrated the map information. It is a figure which illustrated the structure of the map utilization block. It is a flowchart which exemplifies a part of the operation of a map utilization block. It is a figure which illustrated the weighting processing result. It is a block diagram which shows an example of the schematic structure of a vehicle control system. It is explanatory drawing which shows an example of the installation position of the vehicle exterior information detection unit and the image pickup unit.

Hereinafter, modes for implementing the present technology will be described. The explanation will be given in the following order.
1. 1. Information processing system configuration 2. Structure and operation of learning block 3. Configuration and operation of map generation block 4. Configuration and operation of map utilization block 5. Application example

<1. Information processing system configuration>
FIG. 1 illustrates the configuration of an information processing system using the information processing device of the present technology. The information processing system 10 has a learning block 20, a map generation block 30, and a map utilization block 40.

The learning block 20 performs learning using a learning image data group and a feature shape table showing which region in the image corresponds to what kind of feature shape, and generates a feature shape recognition model. .. Further, the learning block 20 performs learning using a learning map data group and a region label table showing regions in which the possibility of existence of an animal body differs from a spatial region in the map, and generates a region recognition model. ..

The map generation block 30 relates to the characteristic shape of the object existing in the sensing data acquired by the internal world sensor or the external world sensor, or the spatial region in which the object is located in the map information, using the model generated by the learning block 20. The area information is extracted, and it is determined whether the object is an animal body based on the extraction result. Further, when the object is determined to be an animal body, the map generation block 30 deletes the animal body from the map information generated based on the sensing data or the map information generated in advance.

The map utilization block 40 estimates its own position based on the map generated by the map generation block 30 and the sensing data acquired by the outside world sensor or the inside world sensor and the outside world sensor.

The learning block 20, the map generation block 30, and the map utilization block 40 may be provided independently, or a plurality of blocks may be provided integrally. For example, a map generation block 30 and a map utilization block 40 are provided on a moving body such as a robot or a vehicle, and the moving body can generate map information in which an animal body is deleted based on sensing data, a feature shape recognition model, and a region recognition model. Update to the map information with the animal body deleted. In addition, the moving body performs self-position estimation based on the sensing data and map information in which the animal body is deleted, and moves according to the action plan based on the self-position estimation result. Further, the map information generated by the map generation block 30 of the moving body may be subsequently used in the map utilization block 40 of the same moving body or another moving body to perform self-position estimation. In addition, a learning block 20 and a map generation block 30 are provided on a server or the like to generate map information, and the generated map information is used by the map utilization block 40 of a moving body as a client to perform self-position estimation and the like. You may.

<2. Learning block structure and operation>
FIG. 2 illustrates the configuration of the learning block. The learning block 20 has a data storage unit 21, a feature shape learner 22, and a region learner 23.

The data storage unit 21 stores, for example, a learning image data group such as a color image or a color image and a depth image, and a feature shape table indicating which region in the image corresponds to what kind of feature shape. ing. Further, in the data storage unit 21, for example, a map data group for learning a movement route of a robot or the like, or a space area in the map in which the possibility of existence of an animal body is different, for example, a passage, a room, a doorway, etc. The area label table indicating whether or not it corresponds to the area of is stored.

The feature shape learner 22 performs learning using the learning image data group and the feature shape table stored in the data storage unit 21, and recognizes the feature shape of an object existing in the sensing data. Generate a model. Further, the area learner 23 performs learning using the learning map data group and the area label table stored in the data storage unit 21, and the possibility of existence of an animal body in the spatial area indicated by the map information (map data). Generates a region recognition model for recognizing different regions.

FIG. 3 is a flowchart illustrating the operation of the learning block. In step ST1, the learning block 20 reads out the data group and the table. The learning block 20 reads out the learning image data group and the feature shape table, and the learning map data group and the area label table stored in the data storage unit 21, and proceeds to step ST2.

In step ST2, the learning block generates a feature shape recognition model. The learning block 20 performs learning with the feature shape learner 22 using the learning image data group and the feature shape table, and generates a feature shape recognition model.

FIG. 4 illustrates the feature shape to be recognized. For example, wheels are used in chairs, desks, luggage carriers, movable whiteboards, etc. in offices. Wheels are used for chairs, desks, TV stands, side tables, etc. in the home environment. Further, the wheels are used outdoors for cars, bicycles, strollers and the like.

The handle is used for drawers and sliding doors in the office. In addition, handles are used for drawers, sliding doors, buckets, etc. in the home environment. Further, the handle is used outdoors for a push cart, a front door, and the like.

Rails are used for sliding doors, movable bookshelves, etc. in the office. In addition, rails are used for sliding doors and the like in a home environment. Further, rails are used outdoors for gates, sliding doors, and the like.

By using such a feature shape, it is possible to handle a large number of animal bodies with a small number of classes (number of feature shapes), and it is possible to easily and efficiently extract animal body candidates described later.

The learning block 20 generates a feature shape recognition model for recognizing feature shapes related to the animal body, for example, wheels, handles, rails, etc. shown in FIG. 4, and proceeds to step ST3.

In step ST3, the learning block generates an area recognition model. The learning block 20 performs learning with the area learner 23 using the learning map data group and the area label table, and recognizes an area such as a passage, a room, a doorway, etc., in which the possibility of existence of an animal body is different. Generate a model.

Note that the processes of step ST2 and step ST3 may be performed first, or may be processed in parallel.

<3. Configuration and operation of map generation block>
FIG. 5 illustrates the configuration of the map generation block. The map generation block 30 includes a sensor unit 31, a self-position estimation unit 32, a map generation unit 34, a region extraction unit 35, a feature shape extraction unit 36, an animal body candidate extraction unit 37, a determination unit 38, and a map information processing unit 39. doing. Further, the map generation block 30 may have an animal body deletion filter 33 as described later. The region extraction unit 35, the feature shape extraction unit 36, and the animal body candidate extraction unit 37 correspond to the extraction unit in the claims.

The sensor unit 31 has an inner world sensor 311 and an outer world sensor 312. The internal sensor 311 acquires information about the mobile body itself on which the information processing device is provided (for example, information indicating the position and posture of the moving body and its change). As the internal world sensor 311, for example, a position sensor, an angle sensor, an acceleration sensor, a gyro sensor, or the like is used. The internal world sensor 311 outputs the generated sensing data (also referred to as “internal world sensing data”) to the self-position estimation unit 32. The external sensor 312 acquires information on the surrounding environment of the moving body in which the information processing device is provided (for example, information on surrounding objects and the like). As the external world sensor 312, for example, a range finder (LIDAR (Light Detection and Ringing, Laser Imaging Detection and Ringing), TOF (Time Of Flight), stereo camera, etc.) 312a, an image sensor 312b for acquiring an captured image, and the like are used. .. The distance measuring sensor 312a of the outside world sensor 312 generates sensing data (also referred to as “distance measuring data”) indicating the distance to a surrounding object and sends it to the self-position estimation unit 32, the map generation unit 34, and the feature shape extraction unit 36. Output. Further, the image sensor 312b of the external world sensor 312 generates sensing data (also referred to as “image image data”) indicating an image captured in the peripheral region and outputs the sensing data (also referred to as “image data”) to the feature shape extraction unit 36.

The self-position estimation unit 32 estimates and estimates the self-position based on the inside world sensing data or the inside world sensing data generated by the inside world sensor 311 of the sensor unit 31 and the distance measurement data generated by the outside world sensor 312. The self-position information indicating the self-position is output to the determination unit 38 and the map information processing unit 39.

The map generation unit 34 generates map information based on the distance measurement data generated by the outside world sensor 312 of the sensor unit 31. The map generation unit 34 generates an occupied grid map by, for example, dividing a two-dimensional plane into grids and assigning the AF points indicated by the distance measurement data to the corresponding grids. The map generation unit 34 outputs the generated map information to the map information processing unit 39.

The region extraction unit 35 uses the region recognition model generated by the learning block 20 to perform region recognition for the map information generated by the map generation unit 34, and extracts regions having different possibilities of existence of the animal body. Further, the region extraction unit 35 assigns region label information including region identification information and an animal body existence score according to the existence possibility of the animal body for each region in which the existence possibility of the animal body is different. For example, the region extraction unit 35 performs region recognition using the region recognition model, and in the map information generated by the map generation unit 34, each region indicates a passage, a room, a doorway, etc., in which the possibility of existence of an animal body is different. Judgment is made, and area identification information indicating which area is set is set. In general, animals are often located in the passages, and there are many fixed objects near the walls in the room. Furthermore, animals are less likely to be located in the doorway than in the passage. Therefore, the region extraction unit 35 raises the animal presence score in the region corresponding to the passage and lowers the animal presence score in the region corresponding to the wall in the room. In addition, the animal presence score in the area corresponding to the doorway or the like is lower than the animal presence score in the area indicating the passage and higher than the animal presence score in the area indicating the wall of the room.

The feature shape extraction unit 36 uses the feature shape recognition model generated by the learning block 20 based on the captured image indicated by the captured image data output from the external world sensor 312 or the depth image indicated by the captured image and the distance measurement data. Feature shape extraction is performed. For example, the feature shape extraction unit 36 extracts the feature shape using the feature shape recognition model, and extracts the feature shape, for example, a wheel, a handle, a rail, or the like from the captured image or the depth image. The feature shape extraction unit 36 generates a feature shape list showing the extracted feature shape and outputs it to the animal body candidate extraction unit 37.

The animal body candidate extraction unit 37 extracts an object having the feature shape shown in the feature shape list as an animal body candidate based on the feature shape list output from the feature shape extraction unit 36. The animal body candidate extraction unit 37 may, for example, separate an object based on sensing data and extract an object including a feature shape as an animal body candidate, and may extract the distance to the feature shape and the distance or shape from the feature shape. Animal candidates may be extracted based on continuity or the like. The animal body candidate extraction unit 37 generates the animal body candidate identification information indicating the extracted animal body candidate and the feature shape label information including the animal body score set based on the feature shape of the animal body candidate. The animal body candidate identification information includes label information and object position information for individually identifying animal body candidates. The animal body score is calculated using a score set in advance for each feature shape. For example, an object with wheels often has a high wheel score because it is often an animal body. Further, since the object provided with the handle may be used not only for the animal body but also for other objects, the score of the handle is made smaller than the score of the wheel. In addition, the score is increased when other characteristic shapes are often used in animals. The animal body candidate extraction unit 37 sets the cumulative value of the score of the feature shape in the animal body candidate as the animal body score indicating the animal body-likeness. In addition, the animal body candidate extraction unit 37 may extract animal body candidates using the animal body score. For example, the animal body candidate extraction unit 37 extracts an object whose animal body score is higher than a preset threshold value as an animal body candidate. The animal body candidate extraction unit 37 outputs the generated feature shape label information to the determination unit 38.

The determination unit 38 determines whether the object is an animal body based on the extraction results of the region extraction unit 35, the feature shape extraction unit 36, and the animal body candidate extraction unit 37. The determination unit 38 uses the self-position estimated by the self-position estimation unit 32 as a reference, the region indicated by the region label information generated by the region extraction unit 35, and the feature shape label generated by the animal candidate extraction unit 37. Correspond to the positions of animal candidates indicated in the information. Further, the determination unit 38 determines whether the animal candidate is an animal body based on the animal body score of the animal body candidate and the animal body existence score of the region where the animal body candidate is located, and determines whether the animal body candidate is an animal body, and determines the determination result in the map information processing unit 39. Output to.

The map information processing unit 39 adds the area label information generated by the area extraction unit 35 to the map information generated by the map generation unit 34, and maps based on the self-position estimated by the self-position estimation unit 32. Store as information. Further, the map information processing unit 39 updates the stored map information by using the newly generated map information with the area label information. Therefore, by repeating the movement of the sensor unit 31 and the generation of map information, the spatial area indicated by the map information stored in the map information processing unit 39 can be expanded. Further, the map information processing unit 39 deletes the animal body candidate determined to be the deletion target in the determination result of the determination unit 38 from the map information. The map information to which the area label information is added is used in the map utilization block 40 as advance map information.

FIG. 6 is a flowchart illustrating the operation of the map generation block. In step ST11, the map generation block acquires sensing data. The map generation block 30 acquires the sensing data generated by the inner world sensor 311 and the outer world sensor 312 of the sensor unit 31 and proceeds to step ST12 and step ST14.

In step ST12, the map generation block performs feature shape extraction processing. The map generation block 30 extracts the feature shape using the feature shape recognition model generated in the learning block 20 based on the sensing data (for example, ranging data and the captured image data) acquired in step ST11, and for example, the extracted feature. A feature shape list indicating the type of shape and the detection position is generated, and the process proceeds to step ST13.

In step ST13, the map generation block performs animal body candidate extraction processing. The map generation block 30 detects an object that moves based on the feature shape extracted in step ST12, and sets the detected object as an animal body candidate. Further, the map generation block 30 generates the feature shape label information including the animal body candidate identification information and the animal body score, and proceeds to step ST16.

FIG. 7 is a diagram for explaining the extraction of animal body candidates. FIG. 7A exemplifies an object included in the sensing range of the external sensor, and it is assumed that, for example, a luggage carrier OB1, a chair OB2, and a box OB3 are included. In FIG. 7B, the wheel FS1 and the handles FS2a and FS2b are extracted by extracting the characteristic shape. Therefore, as shown in FIG. 7 (c), the luggage carrier OB1 shown in the object region which is substantially equal in distance from the wheel FS1 and the handle FS2a and is continuous with the wheel FS1 and the handle FS2a is used as an animal body candidate. Further, the box OB3 indicated by the object region continuous with the handle FS2b at a distance substantially equal to that of the handle FS2b is set as an animal body candidate. In addition, the scores for each of the extracted feature shapes are accumulated and the cumulative value is set as the animal body score SL.

In step ST14, the map generation block performs map generation processing. The map generation block 30 generates map information such as an occupied grid map based on the sensing data (for example, depth image) acquired in step ST11, and proceeds to step ST15.

In step ST15, the map generation block performs area recognition processing. The map generation block 30 performs area recognition for the map information generated in step ST14 using the area recognition model generated in the learning block 20, and based on the area recognition result, the area label including the area identification information and the animal presence score. Information is generated and the process proceeds to step ST16.

FIG. 8 illustrates map information. FIG. 8A exemplifies the map information generated by the map generation process, and FIG. 8B exemplifies the map information to which the area label information is added. For example, FIG. 8B illustrates a case where the area identification information LB1 is assigned to the area indicating the passage, the area identification information LB2 is provided to the area indicating the doorway, and the area identification information LB3 is provided to the area indicating the room. Further, the animal presence score SR1 in the area showing the passage, the animal presence score SR2 in the area showing the doorway, and the animal presence score SR3 in the area showing the room are set to "SR1> SR2> SR3".

In step ST16, the map generation block performs the animal body deletion process. The map generation block 30 is based on the feature shape label information generated in step ST13 and the area label information generated in step ST15, and the animal body score SL of the animal body candidate and the animal body presence score of the area where the animal body candidate is located. Based on the SR, it is determined whether or not each animal candidate is to be deleted. For example, the evaluation value VE is calculated by performing the calculation of the equation (1) using the animal body score SL and the animal body existence score SR. The coefficients α and β are weights for the animal body score SL and the animal body existence score SR, and are set in advance. Further, one of the coefficients α and β may be “0”.
VE = α × SR + β × SL ・ ・ ・ (1)

The map generation block 30 determines that the animal candidate whose evaluation value VE is larger than the preset threshold Th is the deletion target, and deletes the animal candidate to be deleted from the map information. For example, in the case of (b) of FIG. 8, the luggage carrier OB1 at which the passage is located is deleted from the map information, and the box OB3 placed in the room is included in the map information.

Note that, in FIG. 6, a case where the process related to the detection of the animal body candidate and the process related to the area recognition are performed in parallel is illustrated, but one of the processes may be performed first and the other process may be performed later.

In this way, the map generation block is based on the animal body score of the animal body candidate based on the feature shape extracted by the feature shape extraction process and the animal body score of the area recognized by the area recognition process, and the animal body from the map information. To delete. Therefore, even if the sensing data for a predetermined time is not obtained, the map information in which the animal body is deleted can be quickly obtained. In addition, the animal body can be deleted even if the spatial structure changes significantly with the passage of time. Furthermore, since the animal body can be determined based on the characteristic shape of the object or the area information indicating the spatial area where the object is located in the map information, even if the object cannot be recognized by semantic segmentation, for example, from the map information. You will be able to delete the animal body.

By the way, when the animal body deletion filter 33 is provided in the map generation block 30, the animal body deletion filter 33 deletes the distance measurement data of the animal body candidate to be deleted from the distance measurement data based on the determination result of the animal body. .. By deleting the distance measurement data related to the animal body candidate in this way, the map generation unit 34 can generate map information that does not include the animal body.

<4. Configuration and operation of map utilization block>
FIG. 9 illustrates the configuration of the map utilization block. The map utilization block 40 includes a sensor unit 41, a dead reckoning unit 42, a feature shape extraction unit 43, an animal body candidate extraction unit 44, a map information storage unit 45, a weighting processing unit 46, a star reckoning unit 47, and a self-position integration unit 48. Have.

The sensor unit 41 is configured in the same manner as the map generation block 30, and has an inner world sensor 411 and an outer world sensor 412. The internal world sensor 411 is configured by using a position sensor, an angle sensor, an acceleration sensor, a gyro sensor, and the like, and acquires sensing data related to the moving body itself and outputs the sensing data to the dead reckoning unit 42. The external world sensor 412 is configured by using the distance measuring sensor 412a, the image sensor 412b, and the like, and acquires sensing data regarding the surrounding environment of the moving body. The distance measuring sensor 412a of the outside world sensor 412 outputs sensing data (distance measuring data) indicating the distance to a surrounding object to the feature shape extraction unit 43 and the weighting processing unit 46. Further, the image sensor 412b of the external world sensor 412 outputs sensing data (imaging image data) to the feature shape extraction unit 43.

The dead reckoning unit 42 estimates the self-position by determining in which direction and how much the moving body has moved based on the sensing data output from the internal sensor 411, and position information indicating the estimated current self-position. Is output to the self-position integration unit 48.

The feature shape extraction unit 43 uses the feature shape recognition model generated by the learning block 20 based on the captured image indicated by the captured image data output from the external world sensor 412 or the depth image indicated by the captured image and the distance measurement data. Feature shape extraction is performed. The feature shape extraction unit 43 generates a feature shape list showing the extracted feature shape and outputs it to the animal body candidate extraction unit 44.

The animal body candidate extraction unit 44 extracts an object including the feature shape shown in the feature shape list as an animal body candidate based on the feature shape list output from the feature shape extraction unit 43. The animal body candidate extraction unit 44 outputs information indicating the extracted animal body candidate to the weighting processing unit 46.

The map information storage unit 45 stores the map information (preliminary map information) generated by the map generation block 30. The map information storage unit 45 outputs the prior map information to which the area label information is added to the weighting processing unit 46. Further, the map information storage unit 45 outputs the advance map information to the star reckoning unit 47.

The weighting processing unit 46 weights the sensing data supplied from the external sensor 412. For example, the weighting processing unit 46 weights the depth image indicated by the distance measurement data based on at least one of the information indicating the animal body candidate or the area label information of the prior map information, and the area of the animal body candidate. And reduce the weight of the area where the animal body is likely to exist. The weighting processing unit 46 weights the depth image according to at least one of the animal body score of the animal body candidate and the animal body presence score of the region where the animal body candidate is located, so that the animal body is likely to exist. The weight is reduced as the animal-like appearance increases. The weighting processing unit 46 outputs the depth image after the weighting processing to the star reckoning unit 47. Further, the weighting processing unit 46 determines the animal body in the same manner as the determination unit 38 of the map generation block 30 based on the information indicating the animal body candidate and the area label information of the prior map information, and determines that the animal body is an animal body in the depth image. The weight of the created area may be reduced.

The star reckoning unit 47 estimates the self-position by matching the depth image output from the weighting processing unit 46 with the prior map information stored in the map information storage unit 45, and indicates the estimated current self-position. The position information is output to the self-position integration unit 48. Here, in the depth image used by the star reckoning unit 47, the weights of the animal body region and the region where the animal body is likely to exist are reduced, so that the position information generated by the star reckoning unit 47 is influenced by the animal body. It becomes less information. The star reckoning unit 47 outputs the generated position information to the self-position integration unit 48.

The self-position integration unit 48 integrates the position information output from the dead reckoning unit 42 and the position information output from the star reckoning unit 47. For example, the self-position integration unit 48 integrates the two position information obtained by the dead reckoning unit 42 and the star reckoning unit 47 by using an extended Kalman filter, a grid point observer, or the like, and is obtained by, for example, the dead reckoning unit 42. It generates and outputs self-position information that has less error than the position information and is more accurate than the position information obtained by the star reckoning unit 47.

FIG. 10 is a flowchart illustrating a part of the operation of the map utilization block. In step ST21, the map utilization block 40 acquires sensing data. The map utilization block 40 acquires the sensing data generated by the inner world sensor 411 and the outer world sensor 412 of the sensor unit 41, and proceeds to step ST22.

In step ST22, the map utilization block performs feature shape extraction processing. The map utilization block 40 extracts the feature shape using the feature shape recognition model generated in the learning block 20 based on the sensing data (for example, ranging data and the captured image data) acquired in step ST21, and the extracted feature shape. A feature shape list indicating the above is generated, and the process proceeds to step ST23.

In step ST23, the map generation block performs animal candidate extraction processing. The map utilization block 40 determines an animal body based on the feature shape extracted in step ST22, and sets the determined animal body as an animal body candidate. Further, the map utilization block 40 assigns an animal body score indicating the animal body-likeness to the animal body candidate, and proceeds to step ST23.

In step ST24, the map utilization block is weighted. The map utilization block 40 calculates the weight for each animal body candidate extracted in step ST23 based on the animal body score of the animal body candidate and the animal body presence score of the area on the map where the animal body candidate is located. For example, the calculation of the equation (2) is performed using the animal body score SL and the animal body existence score SR to calculate the weight WE of the distance measurement data indicating the animal body candidate. The coefficients α and β are weights for the animal body score SL and the animal body existence score SR, and are set in advance.
WE = 1 / (α × SR + β × SL) ・ ・ ・ (2)

FIG. 11 illustrates the weighting processing result. For example, when the luggage carrier is placed in the aisle indicated by the area identification information LB1, the animal presence score SR1 of the area identification information LB1 and the animal score SL1 of the luggage carrier have score values as described above. As the height increases, the weight WEob1 of the grid Gob1 on which the luggage carrier is located becomes smaller.

In this way, in the map utilization block 40, the grid is weighted based on the animal body score SL of the animal body candidate and the animal body existence score SR of the area where the animal body candidate is located, and the estimation level with the animal body area is performed. The weight is reduced as the value increases. Therefore, for example, when performing star reckoning, it is possible to estimate the self-position by reducing the influence of the animal body.

According to this technology, the animal body is quickly deleted from the pre-map information used in the map utilization block 40, so that the star reckoning unit 47 can estimate the self-position with less influence of the animal body. .. Further, since the weight of the grid indicating the animal body candidate is reduced, the star reckoning unit 47 can estimate the self-position with less influence of the animal body. Further, since the self-position is estimated by integrating the position information output from the dead reckoning unit 42 and the position information output from the star reckoning unit 47, for example, the position information based on the sensing data from the internal sensor is used. Even if the accuracy decreases, the self-position can be estimated accurately. Furthermore, the self-position can be estimated even when the spatial structure changes significantly with the passage of time.

In addition, since animal body candidates can be extracted based on the characteristic shape of the object, the influence of the animal body that cannot be recognized by, for example, semantic segmentation can be reduced. Further, the animal body may be detected by semantic segmentation, and the object which is not detected as an animal body by the semantic segmentation may be determined to be an animal body by using this technique.

<5. Application example>
The technology according to the present disclosure can be applied to various products. For example, the technology according to the present disclosure includes any type of movement such as automobiles, electric vehicles, hybrid electric vehicles, motorcycles, bicycles, personal mobility, airplanes, drones, ships, robots, construction machines, agricultural machines (tractors), and the like. It may be realized as a device mounted on the body.

FIG. 12 is a block diagram showing a schematic configuration example of a vehicle control system 7000, which is an example of a mobile control system to which the technique according to the present disclosure can be applied. The vehicle control system 7000 includes a plurality of electronic control units connected via the communication network 7010. In the example shown in FIG. 12, the vehicle control system 7000 includes a drive system control unit 7100, a body system control unit 7200, a battery control unit 7300, an external information detection unit 7400, an in-vehicle information detection unit 7500, and an integrated control unit 7600. .. The communication network 7010 connecting these plurality of control units conforms to any standard such as CAN (Controller Area Network), LIN (Local Interconnect Network), LAN (Local Area Network) or FlexRay (registered trademark). It may be an in-vehicle communication network.

Each control unit includes a microcomputer that performs arithmetic processing according to various programs, a storage unit that stores a program executed by the microcomputer or parameters used for various arithmetics, and a drive circuit that drives various control target devices. To be equipped. Each control unit is provided with a network I / F for communicating with other control units via the communication network 7010, and is connected to devices or sensors inside or outside the vehicle by wired communication or wireless communication. A communication I / F for performing communication is provided. In FIG. 12, as the functional configuration of the integrated control unit 7600, the microcomputer 7610, the general-purpose communication I / F 7620, the dedicated communication I / F 7630, the positioning unit 7640, the beacon receiving unit 7650, the in-vehicle device I / F 7660, the audio image output unit 7670, The vehicle-mounted network I / F 7680 and the storage unit 7690 are shown. Other control units also include a microcomputer, a communication I / F, a storage unit, and the like.

The drive system control unit 7100 controls the operation of the device related to the drive system of the vehicle according to various programs. For example, the drive system control unit 7100 provides a driving force generator for generating the driving force of the vehicle such as an internal combustion engine or a driving motor, a driving force transmission mechanism for transmitting the driving force to the wheels, and a steering angle of the vehicle. It functions as a control device such as a steering mechanism for adjusting and a braking device for generating a braking force of a vehicle. The drive system control unit 7100 may have a function as a control device such as ABS (Antilock Brake System) or ESC (Electronic Stability Control).

The vehicle condition detection unit 7110 is connected to the drive system control unit 7100. The vehicle state detection unit 7110 may include, for example, a gyro sensor that detects the angular velocity of the axial rotation motion of the vehicle body, an acceleration sensor that detects the acceleration of the vehicle, an accelerator pedal operation amount, a brake pedal operation amount, or steering wheel steering. Includes at least one of the sensors for detecting angular velocity, engine speed, wheel speed, and the like. The drive system control unit 7100 performs arithmetic processing using a signal input from the vehicle state detection unit 7110 to control an internal combustion engine, a drive motor, an electric power steering device, a braking device, and the like.

The body system control unit 7200 controls the operation of various devices mounted on the vehicle body according to various programs. For example, the body system control unit 7200 functions as a keyless entry system, a smart key system, a power window device, or a control device for various lamps such as headlamps, back lamps, brake lamps, blinkers or fog lamps. In this case, the body system control unit 7200 may be input with radio waves transmitted from a portable device that substitutes for the key or signals of various switches. The body system control unit 7200 receives inputs of these radio waves or signals and controls a vehicle door lock device, a power window device, a lamp, and the like.

The battery control unit 7300 controls the secondary battery 7310, which is the power supply source of the drive motor, according to various programs. For example, information such as the battery temperature, the battery output voltage, or the remaining capacity of the battery is input to the battery control unit 7300 from the battery device including the secondary battery 7310. The battery control unit 7300 performs arithmetic processing using these signals, and controls the temperature control of the secondary battery 7310 or the cooling device provided in the battery device.

The vehicle outside information detection unit 7400 detects information outside the vehicle equipped with the vehicle control system 7000. For example, at least one of the image pickup unit 7410 and the vehicle exterior information detection unit 7420 is connected to the vehicle exterior information detection unit 7400. The imaging unit 7410 includes at least one of a ToF (Time Of Flight) camera, a stereo camera, a monocular camera, an infrared camera, and other cameras. The vehicle exterior information detection unit 7420 is used to detect, for example, the current weather or an environmental sensor for detecting the weather, or other vehicles, obstacles, pedestrians, etc. around the vehicle equipped with the vehicle control system 7000. At least one of the ambient information detection sensors is included.

The environmental sensor may be, for example, at least one of a raindrop sensor that detects rainy weather, a fog sensor that detects fog, a sunshine sensor that detects the degree of sunshine, and a snow sensor that detects snowfall. The ambient information detection sensor may be at least one of an ultrasonic sensor, a radar device, and a LIDAR (Light Detection and Ranging, Laser Imaging Detection and Ranging) device. The image pickup unit 7410 and the vehicle exterior information detection unit 7420 may be provided as independent sensors or devices, or may be provided as a device in which a plurality of sensors or devices are integrated.

Here, FIG. 13 shows an example of the installation positions of the image pickup unit 7410 and the vehicle exterior information detection unit 7420. The imaging units 7910, 7912, 7914, 7916, 7918 are provided, for example, at at least one of the front nose, side mirrors, rear bumpers, back door, and upper part of the windshield of the vehicle interior of the vehicle 7900. The image pickup unit 7910 provided on the front nose and the image pickup section 7918 provided on the upper part of the windshield in the vehicle interior mainly acquire an image in front of the vehicle 7900. The imaging units 7912 and 7914 provided in the side mirrors mainly acquire images of the side of the vehicle 7900. The image pickup unit 7916 provided on the rear bumper or the back door mainly acquires an image of the rear of the vehicle 7900. The imaging unit 7918 provided on the upper part of the windshield in the vehicle interior is mainly used for detecting a preceding vehicle, a pedestrian, an obstacle, a traffic light, a traffic sign, a lane, or the like.

Note that FIG. 13 shows an example of the shooting range of each of the imaging units 7910, 7912, 7914, 7916. The imaging range a indicates the imaging range of the imaging unit 7910 provided on the front nose, the imaging ranges b and c indicate the imaging ranges of the imaging units 7912 and 7914 provided on the side mirrors, respectively, and the imaging range d indicates the imaging range d. The imaging range of the imaging unit 7916 provided on the rear bumper or the back door is shown. For example, by superimposing the image data captured by the imaging units 7910, 7912, 7914, 7916, a bird's-eye view image of the vehicle 7900 as viewed from above can be obtained.

The vehicle exterior information detection units 7920, 7922, 7924, 7926, 7928, 7930 provided on the front, rear, side, corners and the upper part of the windshield in the vehicle interior of the vehicle 7900 may be, for example, an ultrasonic sensor or a radar device. The vehicle exterior information detection units 7920, 7926, 7930 provided on the front nose, rear bumper, back door, and upper part of the windshield in the vehicle interior of the vehicle 7900 may be, for example, a lidar device. These out-of-vehicle information detection units 7920 to 7930 are mainly used for detecting a preceding vehicle, a pedestrian, an obstacle, or the like.

Return to FIG. 12 and continue the explanation. The vehicle outside information detection unit 7400 causes the image pickup unit 7410 to capture an image of the outside of the vehicle and receives the captured image data. Further, the vehicle exterior information detection unit 7400 receives detection information from the connected vehicle exterior information detection unit 7420. When the vehicle exterior information detection unit 7420 is an ultrasonic sensor, a radar device, or a lidar device, the vehicle exterior information detection unit 7400 transmits ultrasonic waves, electromagnetic waves, or the like, and receives received reflected wave information. The vehicle exterior information detection unit 7400 may perform object detection processing or distance detection processing such as a person, a vehicle, an obstacle, a sign, or a character on a road surface based on the received information. The vehicle exterior information detection unit 7400 may perform an environment recognition process for recognizing rainfall, fog, road surface conditions, etc., based on the received information. The vehicle outside information detection unit 7400 may calculate the distance to an object outside the vehicle based on the received information.

Further, the vehicle exterior information detection unit 7400 may perform image recognition processing or distance detection processing for recognizing a person, a vehicle, an obstacle, a sign, a character on the road surface, or the like based on the received image data. The vehicle exterior information detection unit 7400 performs processing such as distortion correction or alignment on the received image data, and synthesizes the image data captured by different imaging units 7410 to generate a bird's-eye view image or a panoramic image. May be good. The vehicle exterior information detection unit 7400 may perform the viewpoint conversion process using the image data captured by different imaging units 7410.

The in-vehicle information detection unit 7500 detects the in-vehicle information. For example, a driver state detection unit 7510 that detects the driver's state is connected to the in-vehicle information detection unit 7500. The driver state detection unit 7510 may include a camera that captures the driver, a biosensor that detects the driver's biological information, a microphone that collects sound in the vehicle interior, and the like. The biosensor is provided on, for example, the seat surface or the steering wheel, and detects the biometric information of the passenger sitting on the seat or the driver holding the steering wheel. The in-vehicle information detection unit 7500 may calculate the degree of fatigue or concentration of the driver based on the detection information input from the driver state detection unit 7510, and may determine whether the driver is dozing or not. You may. The in-vehicle information detection unit 7500 may perform processing such as noise canceling processing on the collected audio signal.

The integrated control unit 7600 controls the overall operation in the vehicle control system 7000 according to various programs. An input unit 7800 is connected to the integrated control unit 7600. The input unit 7800 is realized by a device such as a touch panel, a button, a microphone, a switch or a lever, which can be input-operated by a passenger. Data obtained by recognizing the voice input by the microphone may be input to the integrated control unit 7600. The input unit 7800 may be, for example, a remote control device using infrared rays or other radio waves, or an externally connected device such as a mobile phone or a PDA (Personal Digital Assistant) that supports the operation of the vehicle control system 7000. You may. The input unit 7800 may be, for example, a camera, in which case the passenger can input information by gesture. Alternatively, data obtained by detecting the movement of the wearable device worn by the passenger may be input. Further, the input unit 7800 may include, for example, an input control circuit that generates an input signal based on the information input by the passenger or the like using the input unit 7800 and outputs the input signal to the integrated control unit 7600. By operating the input unit 7800, the passenger or the like inputs various data to the vehicle control system 7000 and instructs the processing operation.

The storage unit 7690 may include a ROM (Read Only Memory) for storing various programs executed by the microcomputer, and a RAM (Random Access Memory) for storing various parameters, calculation results, sensor values, and the like. Further, the storage unit 7690 may be realized by a magnetic storage device such as an HDD (Hard Disc Drive), a semiconductor storage device, an optical storage device, an optical magnetic storage device, or the like.

The general-purpose communication I / F 7620 is a general-purpose communication I / F that mediates communication with various devices existing in the external environment 7750. General-purpose communication I / F7620 is a cellular communication protocol such as GSM (registered trademark) (Global System of Mobile communications), WiMAX (registered trademark), LTE (registered trademark) (Long Term Evolution) or LTE-A (LTE-Advanced). , Or other wireless communication protocols such as wireless LAN (also referred to as Wi-Fi®), Bluetooth® may be implemented. The general-purpose communication I / F 7620 connects to a device (for example, an application server or a control server) existing on an external network (for example, the Internet, a cloud network, or a business-specific network) via, for example, a base station or an access point. You may. Further, the general-purpose communication I / F7620 uses, for example, P2P (Peer To Peer) technology, and is a terminal existing in the vicinity of the vehicle (for example, a terminal of a driver, a pedestrian, or a store, or an MTC (Machine Type Communication) terminal). May be connected with.

The dedicated communication I / F 7630 is a communication I / F that supports a communication protocol formulated for use in a vehicle. The dedicated communication I / F7630 uses a standard protocol such as WAVE (Wireless Access in Vehicle Environment), DSRC (Dedicated Short Range Communications), or a cellular communication protocol, which is a combination of the lower layer IEEE802.11p and the upper layer IEEE1609. May be implemented. The dedicated communication I / F7630 typically includes vehicle-to-vehicle (Vehicle to Vehicle) communication, road-to-vehicle (Vehicle to Infrastructure) communication, vehicle-to-home (Vehicle to Home) communication, and pedestrian-to-pedestrian (Vehicle to Pedestrian) communication. ) Carry out V2X communication, which is a concept that includes one or more of communications.

The positioning unit 7640 receives, for example, a GNSS signal from a GNSS (Global Navigation Satellite System) satellite (for example, a GPS signal from a GPS (Global Positioning System) satellite), executes positioning, and executes positioning, and the latitude, longitude, and altitude of the vehicle. Generate location information including. The positioning unit 7640 may specify the current position by exchanging signals with the wireless access point, or may acquire position information from a terminal such as a mobile phone, PHS, or smartphone having a positioning function.

The beacon receiving unit 7650 receives radio waves or electromagnetic waves transmitted from a radio station or the like installed on the road, and acquires information such as the current position, traffic jam, road closure, or required time. The function of the beacon receiving unit 7650 may be included in the above-mentioned dedicated communication I / F 7630.

The in-vehicle device I / F 7660 is a communication interface that mediates the connection between the microprocessor 7610 and various in-vehicle devices 7760 existing in the vehicle. The in-vehicle device I / F7660 may establish a wireless connection using a wireless communication protocol such as wireless LAN, Bluetooth (registered trademark), NFC (Near Field Communication) or WUSB (Wireless USB). In addition, the in-vehicle device I / F7660 is connected via a connection terminal (and a cable if necessary) (not shown), USB (Universal Serial Bus), HDMI (registered trademark) (High-Definition Multimedia Interface, or MHL (Mobile High)). A wired connection such as -definition Link) may be established. The in-vehicle device 7760 includes, for example, at least one of a mobile device or a wearable device owned by a passenger, or an information device carried in or attached to a vehicle. The in-vehicle device 7760 may include a navigation device that searches for a route to an arbitrary destination. The in-vehicle device I / F 7660 is a control signal to and from these in-vehicle devices 7760. Or exchange the data signal.

The in-vehicle network I / F7680 is an interface that mediates communication between the microcomputer 7610 and the communication network 7010. The vehicle-mounted network I / F7680 transmits and receives signals and the like according to a predetermined protocol supported by the communication network 7010.

The microcomputer 7610 of the integrated control unit 7600 is via at least one of general-purpose communication I / F7620, dedicated communication I / F7630, positioning unit 7640, beacon receiving unit 7650, in-vehicle device I / F7660, and in-vehicle network I / F7680. Based on the information acquired in the above, the vehicle control system 7000 is controlled according to various programs. For example, the microcomputer 7610 calculates the control target value of the driving force generator, the steering mechanism, or the braking device based on the acquired information inside and outside the vehicle, and outputs a control command to the drive system control unit 7100. May be good. For example, the microcomputer 7610 realizes ADAS (Advanced Driver Assistance System) functions including vehicle collision avoidance or impact mitigation, follow-up driving based on inter-vehicle distance, vehicle speed maintenance driving, vehicle collision warning, vehicle lane deviation warning, and the like. Cooperative control may be performed for the purpose of. In addition, the microcomputer 7610 automatically travels autonomously without relying on the driver's operation by controlling the driving force generator, steering mechanism, braking device, etc. based on the acquired information on the surroundings of the vehicle. Coordinated control for the purpose of driving or the like may be performed.

The microcomputer 7610 has information acquired via at least one of a general-purpose communication I / F7620, a dedicated communication I / F7630, a positioning unit 7640, a beacon receiving unit 7650, an in-vehicle device I / F7660, and an in-vehicle network I / F7680. Based on the above, the three-dimensional distance information between the vehicle and an object such as a surrounding structure or a person may be generated, and local map information including the peripheral information of the current position of the vehicle may be generated. Further, the microprocessor 7610 may predict a danger such as a vehicle collision, a pedestrian or the like approaching or entering a closed road based on the acquired information, and may generate a warning signal. The warning signal may be, for example, a signal for generating a warning sound or turning on a warning lamp.

The audio image output unit 7670 transmits the output signal of at least one of the audio and the image to the output device capable of visually or audibly notifying the passenger or the outside of the vehicle of the information. In the example of FIG. 12, an audio speaker 7710, a display unit 7720, and an instrument panel 7730 are exemplified as output devices. The display unit 7720 may include, for example, at least one of an onboard display and a heads-up display. The display unit 7720 may have an AR (Augmented Reality) display function. The output device may be a wearable device such as a headphone or a spectacle-type display worn by a passenger, or another device such as a projector or a lamp other than these devices. When the output device is a display device, the display device displays the results obtained by various processes performed by the microcomputer 7610 or the information received from other control units in various formats such as texts, images, tables, and graphs. Display visually. When the output device is an audio output device, the audio output device converts an audio signal composed of reproduced audio data, acoustic data, or the like into an analog signal and outputs the audio signal audibly.

In the example shown in FIG. 12, at least two control units connected via the communication network 7010 may be integrated as one control unit. Alternatively, each control unit may be composed of a plurality of control units. Further, the vehicle control system 7000 may include another control unit (not shown). Further, in the above description, the other control unit may have a part or all of the functions carried out by any of the control units. That is, as long as information is transmitted and received via the communication network 7010, predetermined arithmetic processing may be performed by any control unit. Similarly, a sensor or device connected to one of the control units may be connected to the other control unit, and the plurality of control units may send and receive detection information to and from each other via the communication network 7010. ..

A computer program for realizing the function of the information processing system 10 according to the present embodiment shown in FIG. 1 can be implemented in any control unit or the like. It is also possible to provide a computer-readable recording medium in which such a computer program is stored. The recording medium is, for example, a magnetic disk, an optical disk, a magneto-optical disk, a flash memory, or the like. Further, the above computer program may be distributed via, for example, a network without using a recording medium.

In the vehicle control system 7000 described above, if the map generation block 30 and the map utilization block 40 according to the present embodiment are applied to the integrated control unit 7600 shown in FIG. 12, the animal body can be quickly deleted from the map information. , It becomes possible to detect the self-position accurately by reducing the influence of the animal body. Therefore, it is possible to accurately perform coordinated control for the purpose of automatic driving and the like.

The series of processes described in the specification can be executed by hardware, software, or a composite configuration of both. When executing processing by software, the program that records the processing sequence is installed in the memory in the computer embedded in the dedicated hardware and executed. Alternatively, the program can be installed and executed on a general-purpose computer capable of executing various processes.

For example, the program can be recorded in advance on a hard disk as a recording medium, an SSD (Solid State Drive), or a ROM (Read Only Memory). Alternatively, the program is a flexible disc, CD-ROM (Compact Disc Read Only Memory), MO (Magneto optical) disc, DVD (Digital Versatile Disc), BD (Blu-Ray Disc (registered trademark)), magnetic disc, semiconductor memory card. It can be temporarily or permanently stored (recorded) on a removable recording medium such as an optical disc. Such a removable recording medium can be provided as so-called package software.

In addition to installing the program on the computer from the removable recording medium, the program may be transferred from the download site to the computer wirelessly or by wire via a network such as LAN (Local Area Network) or the Internet. The computer can receive the program transferred in this way and install it on a recording medium such as a built-in hard disk.

Note that the effects described in this specification are merely examples and are not limited, and there may be additional effects not described. In addition, the present technology should not be construed as being limited to the embodiments of the above-mentioned technology. The embodiment of this technique discloses the present technology in the form of an example, and it is obvious that a person skilled in the art can modify or substitute the embodiment without departing from the gist of the present technique. That is, in order to judge the gist of this technology, the claims should be taken into consideration.

In addition, the information processing device of the present technology can have the following configurations.
(1) An extraction unit that extracts characteristic shape of an object existing in sensing data or area information indicating a spatial area in which the object is located in map information.
A determination unit that determines whether the object is an animal body based on the extraction result by the extraction unit, and
An information processing device including a map information processing unit that deletes the object from the map information when the determination unit determines that the object is an animal body.
(2) The extraction unit extracts animal body candidates based on the characteristic shape of the extracted object.
The information processing device according to (1), wherein the determination unit determines whether the animal body candidate extracted by the extraction unit is an animal body.
(3) The information processing apparatus according to (2), wherein the extraction unit extracts a feature shape of the object using a feature shape recognition model generated in advance.
(4) The information processing apparatus according to (2) or (3), wherein the extraction unit sets an animal body score indicating animal-likeness on the object based on the extraction result of the characteristic shape.
(5) The information processing apparatus according to (4), wherein the extraction unit extracts the animal body candidate based on the animal body score.
(6) The extraction unit recognizes a space region indicated by the sensing data using a region recognition model generated in advance, and indicates the susceptibility of an animal body to each recognized region. Set the body presence score,
The determination unit determines whether or not the animal body candidate is an animal body based on at least one of the animal body score of the animal body candidate and the animal body presence score of the region where the animal body candidate is located ( The information processing apparatus according to 5).
(7) The extraction unit performs region recognition on the spatial region indicated by the sensing data using a region recognition model generated in advance, and then performs region recognition.
The information processing apparatus according to any one of (1) to (6), wherein the map information processing unit adds area label information relating to the recognized area to the map information based on the area recognition result of the extraction unit.
(8) The sensing data is data acquired by an external sensor, and is
The information processing apparatus according to any one of (1) to (7), comprising a map information generation unit that generates the map information based on the sensing data.
(9) The information processing according to any one of (1) to (8), comprising a star reckoning unit that detects a self-position using the sensing data and the map information in which the animal body is deleted by the map information processing unit. apparatus.
(10) A weighting processing unit for weighting the sensing data is further provided.
The information processing apparatus according to (9), wherein the star reckoning unit detects a self-position using the sensing data weighted by the weighting processing unit.
(11) The extraction unit extracts animal body candidates based on the characteristic shape of the extracted object.
The information processing apparatus according to (10), wherein the weighting processing unit weights the sensing data indicating the animal body candidate extracted by the extraction unit.
(12) The map information is provided with area label information indicating the animal body existence score indicating the ease of existence of the animal body for each area.
The information processing device according to (10) or (11), wherein the weighting processing unit weights the sensing data based on the area label information.
(13) The weighting processing unit corresponds to at least one of the animal body presence score and the animal body score indicating the animal body-likeness set as the animal body candidate based on the extraction result of the characteristic shape by the extraction unit. , Weighting the sensing data
The information processing apparatus according to (12), wherein the weight is reduced as the animal body becomes more likely to exist or the animal body becomes more likely to exist.
(14) A dead reckoning unit that detects its own position based on the sensing data acquired by the internal sensor, and
The information processing apparatus according to any one of (9) to (13), further comprising a self-position integrating unit that integrates the self-position detected by the dead reckoning and the self-position detected by the star reckoning unit.

10 ... Information processing system 20 ... Learning block 21 ... Data storage unit 22 ... Feature shape learner 23 ... Area learner 30 ... Map generation block 31 ... Sensor unit 32 ...・ ・ Self-position estimation unit 33 ・ ・ ・ Animal body deletion filter 34 ・ ・ ・ Map generation unit 35 ・ ・ ・ Area extraction unit 36 ・ ・ ・ Feature shape extraction unit 37 ・ ・ ・ Animal body candidate extraction unit 38 ・ ・ ・ Judgment Part 39 ・ ・ ・ Map information processing part 40 ・ ・ ・ Map utilization block 41 ・ ・ ・ Sensor part 42 ・ ・ ・ Dead reckoning part 43 ・ ・ ・ Characteristic shape extraction part 44 ・ ・ ・ Animal body candidate extraction part 45 ・ ・ ・Map information storage unit 46 ・ ・ ・ Weight processing unit 47 ・ ・ ・ Star reckoning unit 48 ・ ・ ・ Self- position integration unit 311, 411 ・ ・ ・ Internal world sensor 312, 412 ・ ・ ・ External world sensor 312a, 411a ・ ・ ・ Measurement Distance sensor 312b, 412b ... Image sensor

Claims (16)

1. An extraction unit that extracts the characteristic shape of an object existing in the sensing data or the area information indicating the spatial area in which the object is located in the map information.
   A determination unit that determines whether the object is an animal body based on the extraction result by the extraction unit, and
   An information processing device including a map information processing unit that deletes the object from the map information when the determination unit determines that the object is an animal body.
2. The extraction unit extracts animal body candidates based on the characteristic shape of the extracted object, and then
   The information processing device according to claim 1, wherein the determination unit determines whether the animal body candidate extracted by the extraction unit is an animal body.
3. The information processing device according to claim 2, wherein the extraction unit extracts a feature shape of the object by using a feature shape recognition model generated in advance.
4. The information processing device according to claim 2, wherein the extraction unit sets an animal body score indicating animal-likeness on the object based on the extraction result of the characteristic shape.
5. The information processing device according to claim 4, wherein the extraction unit extracts the animal body candidate based on the animal body score.
6. The extraction unit performs region recognition for the spatial region indicated by the sensing data using a region recognition model generated in advance, and an animal body existence score indicating the ease of existence of the animal body for each recognized region. Set,
   The determination unit determines whether or not the animal body candidate is an animal body based on at least one of the animal body score of the animal body candidate and the animal body presence score of the region where the animal body candidate is located. Item 5. The information processing apparatus according to Item 5.
7. The extraction unit performs region recognition on the spatial region indicated by the sensing data using a region recognition model generated in advance, and then performs region recognition.
   The information processing device according to claim 1, wherein the map information processing unit adds area label information relating to the recognized area to the map information based on the area recognition result of the extraction unit.
8. The sensing data is data acquired by an external sensor, and is
   The information processing apparatus according to claim 1, further comprising a map information generation unit that generates the map information based on the sensing data.
9. The information processing apparatus according to claim 1, further comprising a star reckoning unit that detects a self-position using the sensing data and the map information in which an animal body has been deleted by the map information processing unit.
10. A weighting processing unit that weights the sensing data is further provided.
    The information processing device according to claim 9, wherein the star reckoning unit detects a self-position using the sensing data weighted by the weighting processing unit.
11. The extraction unit extracts animal body candidates based on the characteristic shape of the extracted object, and then
    The information processing device according to claim 10, wherein the weighting processing unit weights the sensing data indicating the animal body candidate extracted by the extraction unit.
12. Area label information indicating the animal body existence score indicating the ease of existence of the animal body for each area is added to the map information.
    The information processing device according to claim 10, wherein the weighting processing unit weights the sensing data based on the area label information.
13. The weighting processing unit responds to at least one of the animal body presence score and the animal body score indicating the animal-likeness set as the animal body candidate based on the extraction result of the characteristic shape by the extraction unit. Weight the data and
    The information processing device according to claim 12, wherein the weight is reduced as the animal body becomes more likely to exist or the animal body becomes more likely to exist.
14. A dead reckoning unit that detects its own position based on the sensing data acquired by the internal sensor,
    The information processing apparatus according to claim 9, further comprising a self-position integration unit that integrates the self-position detected by the dead reckoning and the self-position detected by the star reckoning unit.
15. The extraction unit extracts the characteristic shape of the object existing in the sensing data or the area information indicating the area where the object is located in the map information.
    Based on the extraction result by the extraction unit, the determination unit determines whether the object is an animal body.
    An information processing method including deleting the object from the map information by the map information processing unit when the determination unit determines that the object is an animal body.
16. A program that causes a computer to generate map information.
    The procedure for extracting the characteristic shape of an object existing in the sensing data or the area information indicating the area where the object is located in the map information, and the procedure for extracting the area information.
    A procedure for determining whether the object is an animal body based on the feature shape or the extraction result of the region information, and
    A program that causes the computer to execute a procedure for deleting an object determined to be an animal body from the map information.

Images