Apparatus and method for automatic inspection of a detailed map using object recognition

Abstract

The present invention relates to an automatic precision map inspection device and method using object recognition, comprising: a camera for photographing the exterior of a vehicle; a position measuring unit for measuring the position of a vehicle; a precision map DB storing a precision map containing attribute information for objects on a road; and a control unit that recognizes objects on the road from an image captured by the camera, calculates the relative position between the recognized objects and the vehicle, calculates the absolute position of the recognized objects based on the vehicle's position information obtained through the position measuring unit to produce object recognition information, and compares the calculated object recognition information with the precision map to determine whether there is an error in the precision map, wherein the object recognition information includes the type of object and the absolute position.

Description

Technology Field

[0001] The present invention relates to an automatic precision map inspection device and method using object recognition, and more specifically, to an automatic precision map inspection device and method using object recognition that can automatically inspect a stored precision map by comparing information of an object recognized through a camera with a precision map stored in a precision map DB. Background Technology

[0002] Generally, navigation devices are carried by the user or installed in means of transportation such as vehicles, and perform the function of providing the exact distance and travel time from the current location to the destination, as well as guiding the way to the destination.

[0003] To determine the precise location of a moving object, current navigation systems use a Global Navigation Satellite System (GNSS) that tracks the location of targets on the ground using a satellite network.

[0004] It is essential for such navigation systems to utilize map data for route guidance, and recently, there has been an increasing number of cases where route guidance is provided using high-precision maps that include not only road link and node information but also information such as road width, number of lanes, curvature, gradient, and facilities.

[0005] Furthermore, with the advancement of technology, interest in autonomous vehicles is increasing, and these autonomous vehicles are applying technology that uses precise maps to generate routes and perform driving.

[0006] However, in order to perform autonomous driving using such precise maps, information on lanes, road facilities, and signage facilities must be used, so it is important to provide this information with very high accuracy.

[0007] Therefore, it is required to identify and correct errors when facilities such as signs are added due to road maintenance, or when there are errors in the initially generated precision map.

[0008] Meanwhile, the background technology of the present invention is disclosed in Korean registered patent No. 10-0456377 (November 1, 2004). The problem to be solved

[0009] The purpose of the present invention is to provide an automatic precision map inspection device and method using object recognition, which compares information about an object recognized through a camera with a precision map stored in a precision map DB to automatically inspect a previously stored precision map and perform an update of the precision map. means of solving the problem

[0010] The automatic inspection device for precision maps using object recognition according to the present invention comprises: a camera that photographs the exterior of a vehicle; a position measuring unit that measures the position of the vehicle; a precision map DB that stores a precision map containing attribute information for objects on the road; and a control unit that recognizes objects on the road from an image captured by the camera, calculates the relative position between the recognized object and the vehicle, calculates the absolute position of the recognized object based on the vehicle's position information through the position measuring unit to calculate object recognition information, and compares the calculated object recognition information with the precision map to determine whether there is an error in the precision map, wherein the object recognition information includes the type of object and the absolute position.

[0011] In the present invention, the control unit is characterized by applying an error range according to the positioning method of the position measuring unit when comparing the recognition information of the calculated object with a precision map.

[0012] In the present invention, the control unit is characterized by comparing the recognition information of the calculated object with the precision map within an area centered on the coordinates of the recognized object and having a radius of the error range.

[0013] In the present invention, the control unit is characterized by updating the precision map stored in the precision map DB according to the object recognition information when it is determined that an error exists in the precision map.

[0014] In the present invention, the control unit is characterized by transmitting the recognition information of the object to a server when it is determined that an error exists in the precision map.

[0015] In the present invention, the server verifies the accuracy of the received recognition information of the object, and if it is determined that the recognition information of the object is accurate, it updates the precision map stored in the server according to the recognition information of the object.

[0016] In the present invention, the server is characterized by determining that the recognition information of the object is accurate when recognition information of the same object is received from multiple vehicles.

[0017] A method for automatically verifying a precision map using object recognition according to one aspect of the present invention comprises: a step in which a control unit recognizes an object on a road in an image captured through a camera; a step in which the control unit calculates a relative position between the recognized object and a vehicle; a step in which the control unit calculates the absolute position of the recognized object based on the vehicle's position information and the calculated relative position to calculate object recognition information; a step in which the control unit compares the calculated object recognition information with a precision map stored in a vehicle to determine whether there is an error in the precision map; and a step in which, if it is determined that an error exists in the precision map, the control unit updates the precision map according to the object recognition information.

[0018] In the present invention, the recognition information of the object is characterized by including the type and absolute position of the object.

[0019] In the step of determining whether there is an error in the precision map of the present invention, the control unit is characterized by applying an error range according to the vehicle's positioning method and comparing the recognition information of the object with the precision map.

[0020] A method for automatically verifying a precision map using object recognition according to another aspect of the present invention comprises: a step in which a control unit recognizes an object on a road in an image captured through a camera; a step in which the control unit calculates a relative position between the recognized object and a vehicle; a step in which the control unit calculates the absolute position of the recognized object based on the vehicle's position information and the calculated relative position to produce object recognition information; a step in which the control unit compares the calculated object recognition information with a precision map stored in a vehicle to determine whether there is an error in the precision map; and a step in which, if it is determined that an error exists in the precision map, the control unit transmits the object recognition information to a server.

[0021] A method for automatically verifying a precision map using object recognition according to another aspect of the present invention is characterized by further comprising: a step in which the server verifies the accuracy of the recognition information of the object received; and a step in which, if the recognition information of the object is determined to be accurate, the server updates a precision map stored in the server according to the recognition information of the object.

[0022] In the step of verifying the accuracy of the object recognition information of the present invention, the server determines that the object recognition information is accurate when the same object recognition information is received from multiple vehicles. Effects of the invention

[0023] The automatic precision map inspection device and method using object recognition according to the present invention has the effect of ensuring that the latest precision map is always maintained by determining whether there is an error in the precision map through a comparison between an object recognized by a vehicle and the precision map, and thereby automatically performing a precision map update accordingly. Brief explanation of the drawing

[0024] FIG. 1 is a block diagram showing the configuration of a precision map automatic inspection device using object recognition according to one embodiment of the present invention. FIG. 2 is a flowchart illustrating a method for automatically inspecting a precise map using object recognition according to an embodiment of the present invention. FIG. 3 is another flowchart illustrating a method for automatically inspecting a precision map using object recognition according to an embodiment of the present invention. Specific details for implementing the invention

[0025] Hereinafter, an embodiment of the precision map automatic inspection device and method using object recognition according to the present invention will be described with reference to the attached drawings. In this process, the thickness of lines or the size of components shown in the drawings may be exaggerated for clarity and convenience of explanation. Furthermore, the terms described below are defined considering their functions in the present invention, and these may vary depending on the intention or convention of the user or operator. Therefore, the definitions of these terms should be based on the content throughout this specification.

[0026] FIG. 1 is a block diagram showing the configuration of a precision map automatic inspection device using object recognition according to one embodiment of the present invention.

[0027] As illustrated in FIG. 1, an automatic precision map inspection device using object recognition according to an embodiment of the present invention includes a control unit (100), a camera (110), a location measurement unit (120), a sensor unit (130), and a precision map DB (140). Additionally, the control unit (100) can communicate with a server (200) through a communication module (not shown) and can be configured to transmit data for precision map updates to the server (200). Various methods already in use can be applied to the communication method between the control unit (100) and the server (200).

[0028] The automatic precision map inspection device using such object recognition is installed in each vehicle and can perform inspection on the precision maps stored in the vehicle.

[0029] The camera (110) is configured to photograph the exterior of the vehicle (front, side, rear, etc. of the vehicle), and multiple cameras may be provided.

[0030] The position measuring unit (120) can measure the position of the vehicle. The position measuring unit (120) may include a GNSS system commonly referred to as GPS (Global Positioning System), a Dead Reckoning system, an RTK (Real-Time Kinematic) GPS system, a DGPS (Differential GPS) system, etc., and can calculate the position information of the vehicle from these.

[0031] The sensor unit (130) may be equipped with various sensors to detect the external conditions of the vehicle. For example, various sensors may be provided, such as an altitude sensor for measuring the altitude of the vehicle, a gyroscope sensor for measuring the attitude of the vehicle, a lidar sensor or radar sensor for detecting distance information between the vehicle and an object outside the vehicle.

[0032] Meanwhile, in some embodiments of the present invention, it is also possible for a sensor included in the dead reckoning system of the position measuring unit (120) to function as a sensor included in the sensor unit (130).

[0033] The precision map DB (140) is composed of a storage device such as non-volatile memory and stores a precision map. This precision map stores more information than a general map, and specifically includes data corresponding to additional information (attribute information) about objects on the road, such as lane information, road gradient, signs, milestones, traffic lights, and speed enforcement cameras. Here, the attribute information may include the type of the object.

[0034] The control unit (100) can recognize objects outside the vehicle (lanes, signs, milestones, traffic lights, speed enforcement cameras, etc.) through the camera (110), and can calculate the relative positional relationship (distance) between the recognized objects and the vehicle.

[0035] For example, when detecting facilities using a camera, it is possible to determine the actual size of the facility and its relative distance to the vehicle based on its position and size in the image, and this technology is already widely used in vehicle image sensing.

[0036] In addition, even when using lidar or radar, determining the relative positional relationship between the detected object and the vehicle is a widely known technique, so the control unit (100) can determine the location information of the facility using various methods.

[0037] At this time, the control unit (100) can determine the current location of the vehicle through the location measuring unit (120), so it can calculate the absolute location of the recognized object based on the current location information of the vehicle and the relative location between the recognized object and the vehicle.

[0038] In addition, the control unit (100) can determine what kind of object (lane, sign, milestone, traffic light, speed enforcement camera, etc.) the object is by processing the image recognized through the camera.

[0039] Therefore, the control unit (100) can identify the type and absolute location of the object as recognition information of the object.

[0040] The control unit (100) can verify the precision map by comparing the information of the object whose coordinates were calculated in this way with the attribute information on the precision map of the corresponding coordinates. That is, the control unit (100) can determine that the actual road environment and the precision map match if the attribute information of the same object exists at a point corresponding to the location of the object information on the precision map, and can determine that the actual road environment and the precision map do not match if the attribute information of the same object does not exist.

[0041] At this time, the control unit (100) can select an error range for map verification by considering the error of the positioning system, that is, the position measurement unit (120). In other words, since there is an error in the vehicle's positioning system, verification must be performed by considering this error to reduce the occurrence of inconsistency judgments caused by the error. In addition, since the accuracy of the vehicle's position varies depending on the type of positioning system used for the vehicle's positioning, a variable error range can be applied by taking this into account.

[0042] For example, when using a high-precision positioning system such as RTK, the error range can be set to within 10 cm, when using a system with precision such as DGPS, the error range can be set to within 1 m, and when using a system with precision such as GPS, the error range can be set to within 5 m.

[0043] That is, when using DGPS, if attribute information of an object identical to the object exists within an area (sphere, circle) with a radius of 1m centered on the coordinates of the recognized object, it is determined to be a match, and if attribute information of an object identical to the object does not exist within the area, it is determined to be a mismatch.

[0044] If the control unit (100) determines that the actual road environment and the precision map do not match, it can update the precision map based on the recognition information of the recognized object. That is, it can store the attribute information of the object at the location of the recognized object on the precision map.

[0045] Alternatively, if the control unit (100) determines that the actual road environment and the precision map do not match, it may transmit the recognition information of the object to the server (200) so that the server (200) can perform verification of the recognized information and update of the precision map.

[0046] The server (200) receives recognition information of objects from multiple vehicles that are determined to differ from the precision map mounted on each vehicle, combines these to determine whether there is an error in the precision map data, and can update the precision map.

[0047] Specifically, based on the received object recognition information, the server (200) can determine that if recognition information for the same object is collected from multiple vehicles, the recognition information of the object is accurate (i.e., there is an error in the precision map).

[0048] If the server (200) determines that there is an error in the precision map, it can update the precision map based on the object recognition information. That is, it can store the object's attribute information at the object's location on the precision map.

[0049] At this time, the server (200) may use a method of determining the average position of multiple position values ​​received for the same object as the position of the object.

[0050] That is, the server (200) verifies the accuracy of the received object recognition information, and if it is determined that the object recognition information is accurate, it can update the precision map stored in the server according to the object recognition information.

[0051] Meanwhile, the control unit (100) may be configured to periodically perform object recognition and comparison with a precision map.

[0052] For example, the control unit (100) can compare the attribute information of a precision map of a linear line perpendicular to the direction of travel at the vehicle's location with the recognized object information at set intervals (e.g., 1m).

[0053] In other words, this comparison makes it possible to compare attribute information of the lane alignment at set intervals. In this case, the attribute information of the map alignment can be the number of lanes and the attribute information of each lane.

[0054] In addition, when the control unit (100) periodically compares the attribute information of the lane alignment, if an additional object is recognized, it can also perform a comparison for that object.

[0055] Meanwhile, the control unit (100) may determine whether there is an error in the attribute information of the precision map by using the information measured by the sensor unit (130).

[0056] For example, the inclination of the vehicle and the inclination on the precision map of the current location can be compared to determine whether the attribute information of the precision map matches the actual environment information. Even in this case, the control unit (100) can determine whether there is a match by setting an allowable range while considering the error of the sensor.

[0057] In addition, in the case of complex road forms such as elevated roads or interchange access roads, the control unit (100) can be configured to accurately determine which road the vehicle is currently traveling on by using the vehicle's attitude and altitude information and to perform precise map inspection.

[0058] FIG. 2 is a flowchart illustrating a method for automatically inspecting a precision map using object recognition according to one embodiment of the present invention, and FIG. 3 is another flowchart illustrating a method for automatically inspecting a precision map using object recognition according to one embodiment of the present invention.

[0059] As illustrated in FIG. 2, the control unit (100) first recognizes an object on the road and obtains the relative position of the recognized object with respect to the vehicle (S200). For example, when detecting a facility using a camera, it is possible to determine the actual size of the facility and the relative distance from the vehicle based on the position and size of the facility on the image.

[0060] Next, the control unit (100) calculates the absolute position of the recognized object based on the vehicle's location information (S210). That is, since the control unit (100) can determine the current position of the vehicle through the location measuring unit (120), it can calculate the absolute position of the recognized object based on the vehicle's current location information and the relative position between the recognized object and the vehicle.

[0061] Subsequently, the control unit (100) compares the recognized object with the precision map, taking into account the error according to the vehicle's positioning method (S220). That is, the control unit (100) determines that the actual road environment and the precision map match if the attribute information of the same object exists at a point corresponding to the location of the object information in the precision map, and determines that the actual road environment and the precision map do not match if the attribute information of the same object does not exist.

[0062] At this time, since the accuracy of the vehicle's position varies depending on the type of positioning system used for positioning the vehicle, the control unit (100) can apply a variable error range by taking this into account.

[0063] According to the comparison in the above step (S220), if there is a difference between the recognized object and the precision map, the control unit (100) transmits the recognition information of the object to the server (200) (S230). That is, the precision map automatic verification device mounted on the vehicle only determines whether the actual road environment matches the precision map, and subsequent updates can be performed on the server. By transmitting the recognition information of the object (location coordinates and object type) to the server only when there is a difference between the actual road environment and the precision map, the amount of data communication can be reduced.

[0064] Meanwhile, the control unit (100) may further transmit the image of the object captured along with the recognition information of the object described above to the server (200) so that the server (200) can also utilize the image information.

[0065] Subsequently, as illustrated in FIG. 3, the server (200) verifies the recognition information of the received object (S310). For example, based on the recognition information of the received object, if the recognition information of the object is collected from multiple vehicles at the same location, the server (200) can determine that the recognition information of the object is accurate (i.e., there is an error in the precision map).

[0066] If it is determined that there is an error in the precision map data based on the verification result of the above step (S320), the server (200) updates the precision map according to the object recognition information (S320). That is, the server (200) can update the precision map by storing the attribute information of the object at the location of the object on the precision map.

[0067] In addition, the server (200) may transmit the updated precision map to the control unit (100) of each vehicle so that the precision map update within the vehicle is performed.

[0068] As such, the automatic precision map inspection device and method using object recognition according to an embodiment of the present invention determine whether there are errors in the precision map by comparing an object recognized by a vehicle with the precision map, and thereby automatically perform a precision map update, so that the latest precision map can always be maintained.

[0069] Although the present invention has been described with reference to the embodiments illustrated in the drawings, this is merely illustrative, and those skilled in the art will understand that various modifications and equivalent alternative embodiments are possible therefrom. Accordingly, the technical scope of protection of the present invention should be determined by the following claims. Explanation of the symbols

[0070] 100: Control unit 110: Camera 120: Position measuring unit 130: Sensor section 140: Precision Map DB

Claims

Claim 1 A camera that photographs the exterior of a vehicle; a position measuring unit that measures the position of the vehicle; and a precision map DB that stores a precision map containing attribute information about objects on the road; The control unit includes a control unit that recognizes an object on the road in an image captured by the camera, calculates the relative position between the recognized object and the vehicle, calculates the absolute position of the recognized object based on the vehicle's position information through the position measuring unit to produce object recognition information, and determines whether there is an error in the precise map by comparing the calculated object recognition information with the precise map; the object recognition information includes the type of object and the absolute position; when the control unit compares the calculated object recognition information with the precise map, it applies an error range according to the positioning method of the position measuring unit, and different error ranges are set according to the precision of the positioning method of the position measuring unit; the positioning method of the position measuring unit is one of a GNSS system, a DGPS (Differential GPS) system, and an RTK (Real-Time Kinematic) GPS system, the DGPS system is set to have a smaller error range than the GNSS system, and the RTK GPS system is set to have a smaller error range than the DGPS system; and the control unit includes an object identical to the recognized object within an area centered on the absolute position of the recognized object of the precise map and having the error range according to the positioning method as a radius. An automatic precision map verification device using object recognition, characterized by determining that the road environment and the precision map match if attribute information of an object exists, and determining that the road environment and the precision map do not match if attribute information of an object identical to the recognized object does not exist within the area. Claim 2 delete Claim 3 delete Claim 4 An automatic precision map inspection device using object recognition, wherein, in claim 1, the control unit updates the precision map stored in the precision map DB according to the object recognition information when it is determined that an error exists in the precision map. Claim 5 An automatic precision map inspection device using object recognition, wherein, in claim 1, the control unit transmits object recognition information to a server when it is determined that an error exists in the precision map. Claim 6 An automatic precision map verification device using object recognition, characterized in that, in claim 5, the server verifies the accuracy of the received object recognition information, and if the object recognition information is determined to be accurate, updates the precision map stored in the server according to the object recognition information. Claim 7 A precision map automatic inspection device using object recognition, characterized in that, in claim 6, the server determines that the recognition information of the object is accurate when the same object recognition information is received from multiple vehicles. Claim 8 A step in which a control unit recognizes an object on the road in an image captured through a camera; a step in which the control unit calculates the relative position between the recognized object and the vehicle; a step in which the control unit calculates the absolute position of the recognized object based on the vehicle's position information and the calculated relative position to produce object recognition information; a step in which the control unit compares the calculated object recognition information with a precision map stored in the vehicle to determine whether there is an error in the precision map; The method includes a step in which, if it is determined that an error exists in the precision map, the control unit updates the precision map according to the recognition information of the object, and in the step of determining whether there is an error in the precision map, the control unit compares the recognition information of the object with the precision map by applying an error range according to the vehicle's positioning method, and different error ranges are set according to the precision of the vehicle's positioning method, and the vehicle's positioning method is one of a GNSS system, a DGPS (Differential GPS) system, and an RTK (Real-Time Kinematic) GPS system, and the error range of the DGPS system is set smaller than that of the GNSS system, and the error range of the RTK GPS system is set smaller than that of the DGPS system, and in the step of determining whether there is an error in the precision map, the control unit determines that the road environment and the precision map match if attribute information of an object identical to the recognized object exists within an area of ​​the precision map centered on the absolute position of the recognized object and with the error range according to the positioning method as a radius, and determines that the road environment and the precision map do not match if attribute information of an object identical to the recognized object does not exist within the area. Automatic precision map inspection method using object recognition characterized by Claim 9 delete Claim 10 delete Claim 11 delete Claim 12 A step in which a control unit recognizes an object on the road in an image captured through a camera; a step in which the control unit calculates the relative position between the recognized object and the vehicle; a step in which the control unit calculates the absolute position of the recognized object based on the vehicle's position information and the calculated relative position to produce object recognition information; a step in which the control unit compares the calculated object recognition information with a precision map stored in the vehicle to determine whether there is an error in the precision map; The object includes a step in which, if it is determined that an error exists in the precision map, the control unit transmits the recognition information of the object to a server; in the step of determining whether there is an error in the precision map, the control unit compares the recognition information of the object with the precision map by applying an error range according to the vehicle's positioning method, wherein different error ranges are set according to the precision of the vehicle's positioning method, wherein the vehicle's positioning method is one of a GNSS system, a DGPS (Differential GPS) system, and an RTK (Real-Time Kinematic) GPS system, wherein the error range of the DGPS system is set smaller than that of the GNSS system, and the error range of the RTK GPS system is set smaller than that of the DGPS system; and in the step of determining whether there is an error in the precision map, the control unit determines that the road environment and the precision map match if attribute information of an object identical to the recognized object exists within an area of ​​the precision map centered on the absolute position of the recognized object and with the error range according to the positioning method as a radius, and determines that the road environment and the precision map do not match if attribute information of an object identical to the recognized object does not exist within the area. Automatic precision map verification method using recognition. Claim 13 A method for automatically verifying a precision map using object recognition, characterized in that, in claim 12, the server further comprises the step of verifying the accuracy of the received object recognition information; and, if the object recognition information is determined to be accurate, the server updates a precision map stored in the server according to the object recognition information. Claim 14 A method for automatic inspection of a precision map using object recognition, characterized in that, in the step of verifying the accuracy of the recognition information of the object in claim 13, the server determines that the recognition information of the object is accurate when the same object recognition information is received from multiple vehicles.

Images