JP4459717B2 - Map update system - Google Patents

Description

  The present invention relates to a map update system that provides a three-dimensional map to a driver of a vehicle. In particular, the present invention relates to a map update system that updates a three-dimensional map of a car navigation system using an image captured by a camera.

  Conventionally, a car navigation system that displays a three-dimensional map to a driver is known (see, for example, Non-Patent Document 1). The driver determines a building that is a landmark around the actual road with reference to the three-dimensional map displayed in the car navigation, and drives to the destination.

“Mitsubishi Car Navigation System Real Map”, [online], [Search on April 1, 2004], Internet <URL: http://www.mitsubishielectric.co.jp/carele/carnavi/h8000/realmap_b.html>

  However, when a new building is created around the road, the information displayed by the car navigation system is different from the actual situation around the road. In such a case, there is a problem that a driver becomes confused because a building as a landmark is not found around an actual road.

  Then, an object of this invention is to provide the map update system which can solve said subject. This object is achieved by a combination of features described in the independent claims. The dependent claims define further advantageous specific examples of the present invention.

  In order to solve such a problem, a map update system according to a first embodiment of the present invention includes a camera mounted on a vehicle, a car navigation system having a three-dimensional map, and acquiring a position and a traveling direction of the vehicle. And a map updating unit that updates a three-dimensional map of the car navigation system using an image captured by the camera. Thereby, even if various buildings are built around the road, the latest three-dimensional map can be provided to the driver.

  In addition, the map update system according to the present embodiment extracts a region of a high frequency component whose size is larger than a reference value compared with the size of the low frequency component in the image captured by the camera, and the high frequency component in this region is extracted. A high frequency removing unit that removes the component may be further provided, and the map updating unit may update the three-dimensional map using the image from which the high frequency has been removed by the high frequency removing unit. As a result, it is possible to effectively reduce information such as dirt contained in the subject while maintaining the high frequency component of the boundary region of the subject. Further, since the three-dimensional map is updated using the image from which the high frequency has been removed, it is possible to generate a three-dimensional map that is easy for the driver to grasp and has a small amount of information.

  Further, the map update system according to the present embodiment includes a moving body detection unit that detects a moving body included in the captured image by comparing the positions of the subjects included in the images continuously captured by the camera, and the moving body. The map updating unit may further include a moving body removing unit that removes the moving body detected by the detecting unit, and may update the three-dimensional map using the image from which the moving body has been removed by the moving body removing unit. Thereby, it can prevent that a mobile body is stored in a three-dimensional map.

  The map update system according to the present embodiment further includes a vehicle speed detection unit that detects the speed of the vehicle, and an optical flow calculation unit that calculates the optical flow of the stationary subject using the vehicle speed detected by the vehicle speed detection unit. The body detection unit may detect a subject having a movement vector different from the optical flow calculated by the optical flow calculation unit as a moving body. Thereby, a moving body can be removed reliably.

  In addition, the map update system according to the present embodiment uses only the information on the road width and the position on the road acquired from the car navigation system, and uses only the subject around the road in the image captured by the camera. The map update unit further includes a surrounding subject detection unit to detect, and a surrounding subject selection unit that selects only the surrounding subject detected by the surrounding subject detection unit. The map update unit uses an image of the surrounding subject detected by the surrounding subject selection unit. The 3D map may be updated. Thereby, it is possible to avoid an unnecessary image such as an image on the road or an image of the road itself from being updated to the three-dimensional map.

  The map update system in this embodiment further includes a server storing the latest map, and the map update unit updates the latest three-dimensional map stored in the server using images received from a plurality of vehicles. . Thereby, the server can acquire the latest 3D map of each place and can update an old 3D map.

The map update system according to the present embodiment may further include a latest map acquisition unit that acquires the latest three-dimensional map from the server and stores it in the car navigation system. Thereby, the car navigation system mounted in each vehicle can display the latest three-dimensional map even if it is a place where it passes for the first time.

  When the route search is performed by the car navigation system, the latest map acquisition unit may acquire the latest three-dimensional map on the route from the server and store it in the car navigation system. Since only the necessary three-dimensional map is acquired in this way, the communication time with the server, the communication fee, and the update time can be reduced.

  The above summary of the invention does not enumerate all the necessary features of the present invention, and sub-combinations of these feature groups can also be the invention.

  Hereinafter, the present invention will be described through embodiments of the invention. However, the following embodiments do not limit the claimed invention, and all the combinations of features described in the embodiments are not included in the invention. It is not always essential for development means.

  FIG. 1 shows a configuration of a map update system 900 of the present invention. The map update system 900 includes a plurality of map update devices 100a to 100n, a server 200, a network 300, and a plurality of base stations 400a to 400m. Server 200 is connected to a plurality of base stations 400 a to 400 m via network 300. Each of the map update devices 100a to 100n includes a car navigation system and is mounted on each of the vehicles 10a to 10n.

  Here, conventionally, a car navigation system that displays a three-dimensional map to a driver is known. The driver refers to the three-dimensional map displayed on the car navigation and determines a building existing around the actual road as a landmark for driving. However, when a new building is made around the road or when an existing building is destroyed, the map information displayed by the car navigation system is different from the actual situation around the road. In such a case, there is a problem that the driver is confused because the landmark building is not found around the actual road.

  Therefore, the map update device 100 of the present invention is further provided with a camera and aims to provide the driver with the latest three-dimensional map by capturing an image of the vicinity of the road on which the vehicle 10 travels. Further, the map update device 100 mounted on each vehicle 10 communicates with the server 200 via base stations 400 provided in various places and a network 300 connecting the base stations 400. As a result, the map update system 900 of the present invention acquires images captured by the map update device 100 mounted on the vehicle 10 at various locations, updates the old 3D map to the latest 3D map, and from each vehicle 10. An object is to transmit the latest three-dimensional map to each vehicle 10 in response to the request.

  FIG. 2 is a diagram illustrating a three-dimensional map that the map update apparatus 100 has. FIG. 2A shows a three-dimensional map stored in the map update device 100 in advance. In the example of this figure, while the vehicle 10 is traveling on the road 606 toward the intersection 620, the car navigation system of the map updating apparatus 100 includes the traveling road 606, the intersection 620 ahead, and the road 606 and the intersection 620. A plurality of buildings 608, 610, 612, and 616 existing around are displayed to the driver.

  FIG. 2B shows an image captured by the camera included in the map update device 100. A new building 614 that is not stored in the map update device 100 exists around the actual road 606, and the camera of the map update device 100 captures an image including the new building 614. The image captured by the camera of the map update device 100 includes vehicles 602 and 604 traveling on the road 606, a pedestrian 600 walking on the sidewalk, and the like. Therefore, the map updating apparatus 100 of the present invention detects only the subject around the road by removing the subject on the road, updates the three-dimensional map to the latest three-dimensional map including the new building 614, and moves with respect to the ground. An object of the present invention is to update a three-dimensional map that does not include a moving person or vehicle by detecting a subject to be moved.

  Further, the image captured by the camera included in the map updating apparatus 100 includes information that is not necessarily required to display the course to the driver, such as windows 700 to 704, dirt 706, and the entrance 708 of each building. It is. Therefore, the map updating apparatus 100 is intended to display a three-dimensional map from which unnecessary information is removed to the driver. FIG. 2C shows an example of the updated 3D map.

  FIG. 3 shows an example of a detailed configuration of the map update system 900. The map update device 100 includes a 3D camera 42, a high frequency removing unit 44, an optical flow calculating unit 46, a vehicle speed detecting unit 48, a steering angle detecting unit 50, a turning speed calculating unit 51, a moving body detecting unit 52, a moving body removing unit 54, A surrounding subject detection unit 56, a surrounding subject selection unit 58, a map update unit 60, a car navigation system 62, a latest map acquisition unit 66, and a transmission unit 67 are provided. The car navigation system 62 has a three-dimensional map DB (database) 64. The map update device 100 acquires GPS information from the GPS satellite 500.

  The car navigation system 62 receives GPS information from the GPS satellite 500. Thereby, the car navigation system 62 acquires the current position of the vehicle 10. In addition, the car navigation system 62 of this example stores a history of GPS information received from GPS satellites. The car navigation system 62 identifies the traveling direction of the vehicle 10 from the stored history.

  The 3D camera 42 is arranged in advance in the vehicle 10 so that the optical axis coincides with the traveling direction of the vehicle 10 and continuously images the front of the vehicle 10. Here, the 3D camera 42 includes two cameras inside, and the distance to the subject with respect to the vehicle 10 is obtained using the shift of the position of the feature point of the subject imaged by each camera. The 3D camera 42 calculates the direction of the subject relative to the traveling direction of the vehicle 10 using the position of the subject in the captured image. As a result, the 3D camera 42 detects the relative position of the subject as viewed from the vehicle 10 in each of the continuously captured images.

  The high frequency removing unit 44 extracts a region of a high frequency component whose size is larger than a reference value compared with the size of the low frequency component in the image captured by the 3D camera 42, and the high frequency component in this region Remove. The high-frequency removing unit 44 of this example performs DCT transform (discrete cosine transform) on the obtained image, sets the coefficient of the spectrum having a frequency component larger than the reference value to zero in the spectrum obtained by the transform, and further reverses it. By applying DCT transform (inverse discrete cosine transform), high frequency components are removed. In this case, since the direct current component of the image remains in the image from which the high frequency component has been removed, the high frequency removal unit 44 obtains an image having the density of the direct current component as a pixel value as the image after the removal of the high frequency component.

  The vehicle speed detection unit 48 detects the moving speed of the vehicle 10, and the steering angle detection unit 50 detects the steering angle of the vehicle 10. The turning speed calculation unit 51 calculates the turning speed of the vehicle 10 based on the moving speed detected by the vehicle speed detection unit 48 and the steering angle detected by the steering angle detection unit 50. The turning speed calculation unit 51 of this example calculates the turning speed of the vehicle 10 by multiplying the moving speed detected by the vehicle speed detection unit 48 by the magnitude of the change in the steering angle detected by the steering angle detection unit 50.

  Here, as the vehicle 10 travels, movement occurs in each subject included in the captured image. The movement of the subject is small near the center of the image and large around. Further, the closer the distance from the vehicle 10 to the subject, the larger the distance. In this specification, the movement of each subject in the image determined by the traveling state of the vehicle 10 and the distance of the subject is called an optical flow, and the actual movement of the subject in the image is called a movement vector.

  When it is assumed that the subject is stationary, the optical flow calculation unit 46 calculates a movement vector generated in the subject due to the moving speed and the turning speed of the vehicle 10 as the optical flow of the stationary subject. When the vehicle 10 travels straight, the magnitude of the optical flow of each subject is determined from the moving speed of the vehicle 10, the distance from the vehicle 10 to the object, and the direction of the subject with respect to the vehicle 10. The direction of the optical flow is determined by the distance and direction from the vehicle 10 to the subject, that is, the relative position from the vehicle 10 to the subject. When the vehicle 10 turns within a minute time, the subject in the image has an optical flow in the left-right direction. The direction of the optical flow is opposite to the direction in which the turn is performed, and the magnitude is proportional to the turn speed. Therefore, the optical flow calculation unit 46 adds the optical flow of each object generated when the vehicle 10 turns to the optical flow generated when the vehicle 10 goes straight, so that the vehicle 10 advances while turning. The optical flow of a stationary subject is calculated. The optical flow calculation unit 46 of this example calculates a three-dimensional optical flow.

  The moving body detection unit 52 calculates a relative movement vector of the subject with respect to the vehicle 10 based on the position of the subject calculated by the 3D camera 42. In this case, the moving body detection unit 52 calculates the movement vector of the feature point by obtaining the difference in position between different frames of the same feature point. The moving body detection unit 52 detects, as a moving body, a subject having a movement vector different from the movement vector calculated by the moving body detection unit 52 in the optical flow of the stationary subject calculated by the optical flow calculation unit 46. . Here, the moving body detection unit 52 calculates a difference between the optical flow calculated by the optical flow calculation unit 46 and the calculated movement vector, and if the predetermined size is exceeded, the subject having the movement vector is calculated. Is determined to be a moving object. The moving body removing unit 54 removes the moving body detected by the moving body detecting unit 52. Thereby, a moving body can be removed reliably.

  The surrounding subject detection unit 56 is a subject around the road in the image captured by the 3D camera 42 using the information indicating the width of the running road and the position on the road acquired from the car navigation system 62. Detect only certain surrounding subjects. The car navigation system 62 of this example stores the width of the road in advance, and acquires GPS information from the GPS satellite 500 as information indicating the position of the vehicle 10 on the road. The surrounding subject may be a building or a sign that exists around the road on which the vehicle 10 travels. The surrounding subject detection unit 56 specifies a region indicating the road in the captured image from the current position of the vehicle 10 on the road and the width of the road, and detects a subject in a region other than the road as a surrounding subject. . Then, the surrounding subject selection unit 58 selects the surrounding subject detected by the surrounding subject detection unit 56.

  The map update unit 60 updates the three-dimensional map of the car navigation system 62 using the image captured by the 3D camera 42. The map updating unit 60 updates the three-dimensional map using the image from which the high frequency has been removed by the high frequency removing unit 44. As a result, it is possible to effectively reduce information such as dirt contained in the subject while maintaining the high frequency component of the boundary region of the subject. Further, since the three-dimensional map is updated using the image from which the high frequency has been removed, it is possible to generate a three-dimensional map that is easy for the driver to grasp and has a small amount of information. The map updating unit 60 updates the three-dimensional map using the image from which the moving object has been removed by the moving object removing unit 54. Thereby, it can prevent that a mobile body is stored in a three-dimensional map. The map update unit 60 updates the three-dimensional map using the image of the surrounding subject detected by the surrounding subject selection unit 58. Thereby, it is possible to avoid an unnecessary image such as an image on the road or an image of the road itself from being updated to the three-dimensional map. Details of the operation for updating the three-dimensional map will be described later.

  In addition, the map update part 60 of this example selects the image of the area | region different from the prestored solid map by comparing the prestored solid map with the imaged image. In this case, the map update unit 60 calculates the difference between the pixel value at each coordinate of the three-dimensional map stored in advance in the three-dimensional map DB 64 and the pixel value at each coordinate in the image received from the surrounding subject selection unit 58. When the absolute value of the difference is larger than a predetermined value, the coordinates are selected as a region of the 3D map to be updated. The transmission unit 67 transmits, to the base station 400, image information indicating a region of the three-dimensional map to be updated using radio. Thereby, the amount of data transmitted to the server 200 can be reduced. In this example, the image information indicating the region of the three-dimensional map to be updated is each coordinate indicating the region and a pixel value to be updated at each coordinate.

  The server 200 includes a map update unit 61 and a latest 3D map DB (database) 68. The map update unit 61 updates the latest 3D map stored in the latest 3D map DB 68 using the image information indicating the area of the 3D map to be updated received from the plurality of vehicles 10 via the network 300. . Thus, the latest 3D map DB 68 stores the latest 3D map.

  Here, when a route search is performed by the car navigation system 62, the latest map acquisition unit 66 of the map update device 100 acquires the latest three-dimensional map on the route from the server 200 via the network 300 and the base station 400, and Store in the navigation system 62. As described above, since the latest map acquisition unit 66 acquires the latest 3D map from the server 200 and stores it in the car navigation system 62, the map update apparatus 100 of the present invention acquires the latest image of each location, The map can be updated to the latest contents. And even if it is the place where it passes for the first time, the newest three-dimensional map can be displayed to the driver. Moreover, since the map update apparatus 100 of this invention acquires only a required three-dimensional map from the server 200, the communication time with the server 200, a communication charge, and update time can be reduced.

  4 and 5 are flowcharts showing an example of the operation of the map update device 100. FIG. The car navigation system 62 receives GPS information from the GPS satellite 500. Thereby, the car navigation system 62 acquires the current position and the traveling direction of the vehicle 10 (S100). Note that the car navigation system 62 of this example identifies the traveling direction of the vehicle 10 from the history of the position of the vehicle 10.

  The 3D camera 42 captures a forward image from the vehicle 10 (S105). The 3D camera 42 of this example includes two cameras inside, and obtains the distance to the subject with reference to the vehicle 10 using the shift of the position of the feature point of the subject imaged by each camera. The 3D camera 42 calculates the direction of the subject relative to the traveling direction of the vehicle 10 using the position of the subject in the captured image. Thereby, the 3D camera 42 detects the relative position of the subject as viewed from the vehicle 10.

  When the image picked up by the 3D camera 42 has a high-frequency component whose size is larger than the reference value compared to the size of the low-frequency component (S110: YES), the high-frequency removing unit 44 Is removed (S115). If the image captured by the 3D camera 42 does not have a high-frequency component whose size is larger than the reference value compared to the size of the low-frequency component (S110: NO), the flowchart goes to step S120. move on.

  The vehicle speed detector 48 detects the moving speed of the vehicle 10, and the steering angle detector 50 detects the moving turning angle of the vehicle 10 (S120). The turning speed calculation unit 51 calculates the turning speed of the vehicle 10 based on the moving speed detected by the vehicle speed detection unit 48 and the steering angle detected by the steering angle detection unit 50 (S125). The turning speed calculation unit 51 of this example calculates the turning speed of the vehicle 10 by multiplying the moving speed detected by the vehicle speed detection unit 48 by the magnitude of the change in the steering angle detected by the steering angle detection unit 50.

  The optical flow calculation unit 46 calculates the optical flow of the stationary subject using the vehicle speed detected by the vehicle speed detection unit 48 and the turning speed calculated by the turning speed calculation unit 51 (S130). The optical flow calculation unit 46 of this example calculates a three-dimensional optical flow.

  The moving body detection unit 52 calculates a relative movement vector of the subject with respect to the vehicle 10 based on the position of the subject calculated by the 3D camera 42. In this case, the moving body detection unit 52 calculates the movement vector of the feature point by obtaining the difference in position between different frames of the same feature point. When the optical flow of the stationary subject calculated by the optical flow calculation unit 46 has a movement vector different from the movement vector calculated by the moving object detection unit 52 (S135: YES), the moving object detection unit 52 The subject is detected as a moving body (S140). Here, the moving body detection unit 52 calculates a difference between the optical flow calculated by the optical flow calculation unit 46 and the calculated movement vector, and if the predetermined size is exceeded, the subject having the movement vector is calculated. Is determined to be a moving object. Then, the moving body removing unit 54 removes the moving body detected by the moving body detecting unit 52 (S145).

If the optical flow of the stationary subject calculated by the optical flow calculation unit 46 does not have a movement vector different from the movement vector calculated by the moving object detection unit 52 (S135: NO), this flowchart proceeds to step S150. move on.

  The surrounding subject detection unit 56 acquires information indicating the width of the currently traveling road and the position on the road based on the position information acquired from the car navigation system 62. Thereby, the surrounding subject detection unit 56 determines the road and the area on the road in the image captured by the 3D camera 42 (S150), and excludes the road and the subject on the road (S155). Here, the surrounding subject may be a building or a sign existing around the road on which the vehicle 10 travels. The surrounding subject selection unit 58 selects only the surrounding subject detected by the surrounding subject detection unit 56 (S160).

  The map update unit 60 updates the three-dimensional map of the car navigation system 62 using the image captured by the 3D camera 42 (S165). The map update unit 60 of this example displays an image from which the high frequency has been removed by the high frequency removing unit 44, an image from which the moving body has been removed by the moving body removing unit 54, and an image of the surrounding subject detected by the surrounding subject selecting unit 58. Use to update the 3D map. Alternatively, the map update unit 60 may update the three-dimensional map using any one of the image from which the high frequency is removed, the image from which the moving body is removed, or the image of the surrounding subject.

  The map update unit 60 of this example calculates the difference between the pixel value at each coordinate of the stereoscopic map stored in advance in the stereoscopic map DB 64 and the pixel value at each coordinate in the image received from the surrounding subject selection unit 58. When the absolute value of the difference is larger than a predetermined value, the coordinates are selected as a region of the 3D map to be updated. Then, the transmission part 67 transmits the image information which shows the area | region of the solid map which should be updated to the server 200 via the base station 400 using a radio | wireless (S170). Then, the latest 3D map DB 68 of the server 200 updates the latest 3D map stored in the latest 3D map DB 68 using the image received from each vehicle 10.

  Here, when a route search is performed by the car navigation system 62 (S175: YES), the latest map acquisition unit 66 acquires the latest 3D map on the route from the server 200 (S180), and the acquired latest 3D map. Is stored in the car navigation system. When the route search is not performed by the car navigation system 62 (S175: NO), the flowchart proceeds to step S100, and the operations from step S100 to step S175 described above are repeated. Above, this flowchart is complete | finished.

  FIG. 6 is a diagram for explaining display of a three-dimensional map. The car navigation system 62 stores, in the 3D map DB 64, intersections and places where roads turn as nodes, and stores each road as a link connecting the nodes. In the example of this figure, links 804 to 815 are connected to the nodes 800 and 802. Here, the car navigation system 62 stores position information in association with nodes and stores road information in association with links. For example, the car navigation system 62 stores position information indicated by latitude and longitude in association with a node, and stores information such as a legal speed of a road corresponding to the link as road information in association with the link. By providing such information to the driver, the car navigation system 62 supports the driving of the driver.

  This figure shows a state where the vehicle 10 traveling on the road 606 (see FIG. 2) is viewed from directly above. For convenience of explanation, nodes and links corresponding to intersections and roads are shown in an overlapping manner. The three-dimensional map DB 64 stores three-dimensional image information for generating a three-dimensional map in association with each link, and the car navigation system 62 is stored based on the current position of the vehicle 10 at each link. A three-dimensional map is generated using the image information. The three-dimensional map DB 64 of this example stores, as three-dimensional image information, three-dimensional coordinates having a node as the origin of coordinates and pixel values corresponding to the respective three-dimensional coordinates. Then, the car navigation system 62 displays a three-dimensional map to be displayed to the driver based on the position of the vehicle 10 on the link while generating it. The car navigation system 62 of this example displays an image representing the road and the surface of a building as viewed from the driver's viewpoint as a three-dimensional map to be displayed to the driver.

  For example, the position 820 on the surface of the building 616 is defined by three-dimensional coordinates (X, Y, Z) with the node 800 as the origin, and the three-dimensional map DB 64 stores the pixel value A at the position 820. Similarly, the three-dimensional map DB 64 stores three-dimensional coordinates and image values in association with roads and all positions around the road. When the vehicle 10 is located on the link 808 indicated by the arrow 804 and is moving in the direction indicated by the arrow 816, that is, from the node 800 to the node 802, the car navigation system 62 is visible on the road viewed from the driver of the vehicle 10. A three-dimensional map of each building is generated and displayed. In the example of this figure, when the area to be displayed to the driver is the area 850 indicated by the dotted line, the car navigation system 62 displays a three-dimensional map including the buildings 614, 616, 608, 610, and 612. . When the traveling direction is opposite to the arrow 816, the car navigation system 62 converts the node 802 into three-dimensional coordinates with the origin as the origin and displays a three-dimensional map.

  FIG. 7 is a flowchart showing an example of detailed operation in step S165. When a new building 614 has been built, the 3D camera 42 of the map update device 100 captures an image including the building 614. When the current position is acquired from the GPS satellite 500 (see FIG. 3), the car navigation system 62 converts the current position information of the vehicle 10 received from the GPS satellite 500 into coordinates on the link. The car navigation system 62 of the present example converts the current position of the vehicle 10 into coordinates with the node connected in the direction opposite to the traveling direction of the vehicle 10 as the origin. In the example of FIG. 6, the car navigation system 62 specifies a position on the link 808 with the node 800 as the origin (S200).

  Next, the map update unit 60 uses each of the three-dimensional image information stored in advance, that is, the old three-dimensional image information, to generate a three-dimensional map to be displayed to the driver as the current position of the vehicle 10. Point coordinates and pixel values are calculated (S205). Then, at each coordinate, the pixel value of the three-dimensional image captured by the 3D camera 42 is compared with the pixel value of the old three-dimensional image, and when the difference is larger than a predetermined value, the pixel value of the old three-dimensional image Is replaced with a pixel value of an image newly captured by the 3D camera 42 (S210). Thereby, the map update part 60 updates a three-dimensional map.

  Note that the optical flow calculation unit 46 of this example calculates the three-dimensional optical flow of each subject, but it may calculate the two-dimensional optical flow more simply. In this case, the moving body detection unit 52 calculates the optical flow generated in the entire image by calculating the movement vector of each subject from the continuously captured images, and has a movement vector different from the calculated optical flow. The subject may be detected as a moving body.

  Moreover, although the map update apparatus 100 of this example acquired the image using the 3D camera 42, the map update apparatus 100 in another example may acquire an image using a 2D camera. In this case, the map update unit 60 updates the three-dimensional map by specifying image information stored in a three-dimensional format in advance, which is necessary for generating a three-dimensional map corresponding to the image captured by the 2D camera. To do.

  Alternatively, as another example, the map update device 100 may store in advance a two-dimensional image to be displayed to the driver according to the position of the vehicle. In this case, the map update device 100 performs pattern matching processing between the image stored in advance and the image acquired by the 2D camera, and detects a subject that is not included in the image stored in advance, whereby the image of the map update device 100 is detected. Update.

  As is apparent from the above description, according to the present embodiment, the map update unit 60 updates the 3D map of the car navigation system 62 using the image captured by the 3D camera 42. Even when a simple building is built, the latest stereoscopic image can be provided to the driver of the vehicle. The high frequency removal unit 44 extracts a region having only a high frequency component from the image captured by the 3D camera 42 and removes the high frequency component in this region, so that the high frequency component in the boundary region of the subject is retained. Information such as dirt contained in the subject can be effectively reduced. In addition, the moving body detection unit 52 detects the moving body included in the captured image by comparing the positions of the subjects included in the images continuously captured by the 3D camera 42, and thus moves to the three-dimensional map. The body can be prevented from being stored.

  As mentioned above, although this invention was demonstrated using embodiment, the technical scope of this invention is not limited to the range as described in the said embodiment. It will be apparent to those skilled in the art that various modifications or improvements can be added to the above-described embodiment. It is apparent from the scope of the claims that the embodiments added with such changes or improvements are also included in the technical scope of the present invention.

It is a figure which shows the structure of the map update system 900 of this invention. It is a figure explaining the three-dimensional map which the map update apparatus has. (A) shows the solid map previously stored in the map update apparatus 100. FIG. (B) shows the image which the camera which the map update apparatus 100 has imaged. (C) shows the updated three-dimensional map stored in the map update device 100. It is a figure which shows an example of a detailed structure of the map update system. 5 is a flowchart showing an example of the operation of the map update device 100. 5 is a flowchart showing an example of the operation of the map update device 100. It is a figure explaining the display of a three-dimensional map. It is a flowchart which shows an example of the detailed operation | movement in step S165.

Explanation of symbols

10a to n ... vehicle, 42 ... 3D camera, 44 ... high frequency removing unit, 46 ... optical flow calculating unit, 48 ... vehicle speed detecting unit, 50 ... steering angle detecting unit, 51 ... turning speed calculation unit, 52 ... moving body detection unit, 54 ... moving body removal unit, 56 ... surrounding subject detection unit, 58 ... surrounding subject selection unit, 60 ... map update , 61... Map update unit, 62... Car navigation system, 64... 3D map DB, 66... Latest map acquisition unit, 67. , 100a to n ... map update device, 200 ... server, 300 ... network, 400a to m ... base station, 500 ... GPS satellite, 600 ... pedestrian, 602, 604 ..Vehicle, 606 ... road, 608, 610, 6 2, 614, 616 ... Building, 620 ... Intersection, 700, 702, 704 ... Window, 706 ... Dirt, 708 ... Entrance, 800, 802 ... Node, 804, 816 ..Arrow, 806, 808, 810, 812, 814, 815 ... link, 820 ... position, 850 ... area, 900 ... map update system

Claims (8)

A camera mounted on the vehicle,
A car navigation system having a three-dimensional map and acquiring the position and traveling direction of the vehicle;
In the image captured by the camera, a high-frequency removing unit that extracts a region where the magnitude of the high-frequency component is greater than the reference value by the magnitude of the low-frequency component,
A map updating system comprising: a map updating unit that updates a three-dimensional map of the car navigation system using an image from which a high frequency component has been removed by the high frequency removing unit .   A surrounding subject for detecting a surrounding subject, which is a subject around the road, in the image captured by the camera, using information indicating the width of the running road and the position on the road acquired from the car navigation system Detection unit
Further comprising
  The map update system according to claim 1, wherein the map update unit updates a three-dimensional map using an image of a surrounding subject detected by the surrounding subject detection unit.   A camera mounted on the vehicle,
  A car navigation system having a three-dimensional map and acquiring the position and traveling direction of the vehicle;
  A surrounding subject for detecting a surrounding subject, which is a subject around the road, in the image captured by the camera, using information indicating the width of the running road and the position on the road acquired from the car navigation system A detection unit;
  A map updating unit that updates a three-dimensional map of the car navigation system using an image of the surrounding subject detected by the surrounding subject detection unit;
Map update system with A moving body detection unit that detects a moving body included in the captured image by comparing positions of subjects included in the images continuously captured by the camera;
A moving body removing unit that removes the moving body detected by the moving body detecting unit;
The map update system according to any one of claims 1 to 3, wherein the map update unit updates a three-dimensional map using an image from which a moving object has been removed by the moving object removal unit. A vehicle speed detector for detecting the moving speed of the vehicle;
An optical flow calculation unit that calculates an optical flow of a stationary subject using the moving speed detected by the vehicle speed detection unit is further provided. The map update system according to claim 4 , wherein a subject having a different movement vector is detected as a moving body. A server storing the latest 3D map;
The map updating unit, using the image received from a plurality of vehicles, the map updating system according to any one of claims 1 to 5 for updating the latest three-dimensional map in which the stored in the server.   The map update system according to claim 6, further comprising a latest map acquisition unit that acquires the latest three-dimensional map from the server and stores it in the car navigation system.   The map update system according to claim 7, wherein when a route search is performed by the car navigation system, the latest map acquisition unit acquires the latest three-dimensional map on the route from the server and stores it in the car navigation system.

Images