JP4700080B2 - Car navigation system and method - Google Patents

Description

  The present invention relates to a car navigation system and method, and more particularly, to a car navigation system and method for displaying a point of interest (POI) on an image collected in real time using a camera mounted on a vehicle.

  The present invention is derived from research conducted as part of the IT growth dynamic innovation technology development project of the Korea Information and Communication Department and the Institute of Information and Communication Technology [Problem Management Number: 2005-S-114-03, Problem] Name: Technology Development for Construction and Management of Tangible Content for Telematics Services].

  In a car navigation system based on a map that navigates the progress of a vehicle in a two-dimensional map form, a point of interest (POI) is displayed on a map screen in the form of an icon or text. The POI refers to an object such as a building, a park, or a facility. The POI information includes data including coordinates in the form of points or planes on a map coordinate system and text information.

  Further, in a video car navigation system that navigates the running of a vehicle using video collected in real time using a camera mounted on the vehicle, the POI is displayed over the front video of the vehicle. However, in order to output POI to video, an accurate camera position and posture are required. In a car navigation system based on a two-dimensional map, there is no major problem in executing functions such as route guidance and map output as long as the approximate position of the vehicle can be confirmed. The presence of a slight error in posture greatly affects the car navigation function. For example, when traveling on one lane on a wide road and traveling on four lanes, it is difficult to distinguish between ordinary GPS receivers and functions of a two-dimensional car navigation system, and it is not necessary to distinguish them. There is a big difference in video output of information.

  Conventionally, in order to obtain accurate camera information, a technique for correcting camera position and orientation information using an inertial sensor such as a gyroscope is known. Conventionally, a technique for detecting an azimuth angle and an inclination angle of a camera using an inertial sensor such as a gyroscope and an inclinometer is also known.

  Another thing to consider in a video car navigation system that is not considered in a two-dimensional map-based car navigation system is that if the building is far away from the roadside entrance or if the area of the building is very large, There is a possibility that a difference occurs between the actual position of the video and the video output position of the POI. For example, when only the center point of the building is stored as POI coordinates, the entrance for entering the building is actually on the roadside and enters the field of view, but the building itself may not be output as a POI on the screen. Can occur.

  Since the conventional video car navigation system uses an expensive inertial sensor in order to obtain accurate camera information, there is a problem that the manufacturing unit cost increases. Further, since the conventional video car navigation system provides only a single coordinate related to the POI, a large building may be included in the field of view, but the building may not be output as a POI on the video. There was a problem that could not.

  Accordingly, an object of the present invention is to provide a video car navigation system and method capable of acquiring inexpensive and accurate camera information.

  In addition, the present invention does not require an expensive inertial sensor, and uses only the GPS receiver and the camera, and uses the result recognized from the video acquired from the camera, so that the POI information can be superimposed on the real-time video and output. It is an object to provide a car navigation system and method.

  Further, the present invention differs from the conventional POI data construction method by using additional point data associated with POI coordinates (points or faces), so that even when the building does not appear on the screen, it can be An object of the present invention is to provide a video car navigation system and method capable of outputting a position of an actually required entrance / exit on a roadside on a screen to help a driver search for a road.

  In order to achieve the above object, the video car navigation system of the present invention is a video car navigation system for displaying a target point on a screen, and data for acquiring the screen using a camera mounted on a vehicle. An input unit; a video recognition unit that recognizes a reference object from the screen; an electronic map unit that provides information on the target point and the position of the reference object on the map; and a position of the reference object on the map; A calculation unit that determines a target point position on the screen of the target point based on the reference object, using a relative positional relationship between the position of the target point on the map; And an output unit for displaying the target point at the target point position.

  In order to achieve the above object, a video car navigation method of the present invention is a video car navigation method for displaying a target point on a screen, and is mounted on a vehicle by (a) a data input unit. Obtaining the screen using a camera; (b) recognizing a reference object from the screen by the video recognition unit; and (c) a position of the reference object on the map provided by the electronic map unit; Determining a target point position on the screen of the target point based on the reference object by a calculation unit using a relative positional relationship between the position on the map of the target point; (D) The step of displaying the target point at the target point position on the screen by the output unit.

  Here, the reference object may be a signal lamp.

  According to this configuration of the present invention, the target point can be specified only by using the position of the target point relative to the reference object on the map without using an existing expensive inertial sensor.

  In the video car navigation system and method of the present invention, the calculation unit includes a screen reference region setting unit that divides the screen based on the reference object into a plurality of screen reference regions, and the screen reference region setting The unit determines the position of the target spot by causing the target spot to correspond to one screen reference area of the plurality of screen reference areas. Here, the correspondence relationship of the target point with respect to the screen reference region is determined by the positional relationship among the vehicle, the reference object, and the target point.

  Further, in the video car navigation system and method of the present invention, the output unit calculates a two-dimensional image coordinate from a position of the target point on the map, and adds the screen reference region associated with the target point position to the screen reference region. If the two-dimensional image coordinates are not included, the target point is displayed at the target point position instead of displaying the target point at the two-dimensional image coordinates.

  According to this configuration of the present invention, when there is an error in the two-dimensional image coordinates of the target point calculated from the three-dimensional map coordinates, the target point position specified in the reference region setting process is used instead of the two-dimensional image coordinates. Since the target point can be displayed, the target point can be displayed at a more accurate position on the screen.

  In the video car navigation system and method of the present invention, the video recognition unit includes a lane recognition unit that recognizes a lane ahead of the vehicle from the screen, and the calculation unit is recognized by the lane recognition unit. An azimuth angle calculating unit that calculates a vehicle azimuth angle of the vehicle from information on the lane can be included.

  Here, the output unit provides from the vehicle azimuth calculated by the calculation unit and the GPS receiver as a vehicle azimuth used to calculate a two-dimensional image coordinate from the position of the target point on the map. Any one of the vehicle azimuths thus selected can be selected.

  According to this configuration of the present invention, the vehicle azimuth from the GPS receiver is used in a vehicle speed section (eg, 20 km / h or more) with high reliability with respect to GPS data, otherwise the recognized lane Therefore, the vehicle azimuth angle used for calculating the two-dimensional image coordinates from the target point position on the map can be specified more accurately.

  In the video car navigation system and method of the present invention, the video recognition unit includes a traveling roadway recognition unit that recognizes a traveling roadway of the vehicle from information on the lane recognized from the screen, and the calculation unit includes the traveling The position calculating part which calculates the position of the said vehicle from the information regarding the said traveling roadway recognized by the roadway recognition part can be included.

  Here, the data input unit receives position information of the vehicle from a GPS receiver, and the position calculation unit recognizes the position information of the vehicle received from the GPS receiver by the traveling roadway recognition unit. The position of the vehicle is calculated by performing correction using the information related to the traveling roadway.

  According to this configuration of the present invention, it is possible to eliminate a distance error in a direction perpendicular to the road center line with respect to the position of the vehicle specified by the road center line based on information from the GPS receiver.

  Further, in the video car navigation system and method of the present invention, the target point is associated with a plurality of target point coordinates on a map, and the calculation unit determines the relative positional relationship in order to determine the relative positional relationship. Any one of the target point coordinates can be specified as the position of the target point on the map.

  According to this configuration of the present invention, the position of the target point is displayed on the screen as needed by a plurality of coordinates including coordinates representing the target point and coordinates representing the entrance / exit of the target point linked to the coordinate. Since it can represent above, a target point can be pinpointed more correctly.

  According to the present invention, it is possible to specify the target point only by the relative position of the target point on the map with respect to the reference object without requiring an existing expensive inertial sensor.

  Further, according to the present invention, when there is an error in the two-dimensional image coordinates of the target point calculated from the three-dimensional map coordinates, the target point position specified in the reference region setting process is used instead of the two-dimensional image coordinates. Since the target point can be displayed, the target point can be displayed at a more accurate position on the screen.

  In addition, according to the present invention, in a vehicle speed section (for example, 20 km / h) with high reliability with respect to GPS data, the vehicle azimuth from the GPS receiver is used, and if not, the vehicle calculated from the recognized lane Since the azimuth angle can be used, the vehicle azimuth angle used for calculating the two-dimensional image coordinates from the target point position on the map can be specified more accurately.

  Further, according to the present invention, it is possible to eliminate a distance error in a direction perpendicular to the road center line with respect to the position of the vehicle specified by the road center line based on information from the GPS receiver.

  Further, according to the present invention, the position of the target point is displayed on the screen as necessary by using a plurality of coordinates including coordinates representing the target point and coordinates representing the entrance and exit of the target point. Therefore, the target point can be specified more accurately.

Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings.
FIG. 1 is a block diagram showing a video car navigation system 100 according to an embodiment of the present invention, and FIG. 2 is a block diagram showing details of FIG. The video car navigation system 100 includes a data input unit 10, an electronic map unit 20, a video recognition unit 30, a calculation unit 40, and an output unit 50.

  The data input unit 10 acquires a current position of the vehicle and an image ahead of the vehicle using a GPS receiver and a camera (not shown). The electronic map unit 20 includes POI data including POI position coordinate values and text describing the POI (for example, a building name such as “ABC building”), signal lamp data including signal lamp position coordinate values, and an electronic road map. The POI data and the signal light data associated with the current position of the vehicle are retrieved and map-matched with the electronic road map. The image recognizing unit 30 recognizes the lane and the signal light from the road image acquired using the camera, and calculates the current traveling road position of the vehicle. The calculation unit 40 calculates the vehicle azimuth angle from the GPS signal, and calculates the vehicle azimuth angle and the current vehicle position from the image coordinates of the lane and the signal light received from the video recognition unit 30 and the current traveling road position of the vehicle. To do. The output unit 50 converts the POI coordinates into image coordinates using the vehicle position and the vehicle azimuth angle, and outputs them by superimposing them on the video.

  In this specification, an area having a width where a vehicle travels on a road is referred to as a “roadway”, and a line (including a center line and a road boundary line) dividing the “roadway” is referred to as a “lane”. The “traveling lane position” refers to the roadway number assigned in order from the center line of the road. For example, in the case of a four-lane road on one side, if the vehicle is traveling on the second lane from the center, the traveling lane position = 2 “Vehicle azimuth” refers to an angle formed by the traveling direction of the vehicle (same as the direction in which the optical axis of the camera faces) with respect to the N axis of the map coordinate system.

  In this embodiment, a signal light is used as a reference object for POI video display. However, any object other than the signal light can be used as long as the object has high discrimination power against other objects and high arrangement density. May be used as a reference object.

  FIG. 2 shows a detailed configuration diagram of the video car navigation system 100 of FIG. The data input unit 10 includes a camera 11 for acquiring a real-time video ahead of the vehicle and a GPS receiver 12 for acquiring GPS data including a vehicle position, a vehicle azimuth angle, and a vehicle speed. The direction of the camera is assumed to be parallel to the traveling direction of the vehicle, and in this specification, the traveling direction of the vehicle and the direction of the camera are used to represent the same meaning. Further, the distance between the camera 11 and the GPS receiver 12 is arranged so as to be negligible.

  The electronic map unit 20 includes an electronic road map unit 21, a map matching processing unit 22, a POI data unit 23, a POI search unit 24, a signal lamp data unit 25, and a signal lamp search unit 26. The electronic road map unit 21 has road coordinates.

  The map matching processing unit 22 receives GPS coordinates from the GPS receiver 12 and executes a map matching process. In the map matching process, for example, the current GPS coordinates are compared with the road coordinates of the electronic road map unit 21, and the position on the road closest to the current GPS coordinates is calculated. However, the present invention is not limited to this. . That is, the map matching processing unit 22 maps the vehicle position to the road center line as the position on the road closest to the GPS coordinate value in order to minimize an error existing in the GPS coordinate value. Therefore, the GPS error in the direction perpendicular to the road center line is removed. The coordinate values obtained through the map matching process as described above are transmitted to the calculation unit 40.

  The POI data section 23 has respective POI coordinate values and text attribute values related to various objects (buildings, parks, facilities, etc.). In the case of a building, the POI coordinates are configured by linking an object coordinate value representatively representing the position of the building itself with an additional coordinate value representing the position of the entrance / exit on the roadside of the building. The configuration of such data will be described in more detail with reference to FIG.

  FIG. 3B is an object coordinate value table including coordinate values of the object itself (that is, the city hall itself). The object coordinate values are distinguished from each other by a text attribute value such as “city hall”. FIG. 3C is an additional coordinate value table including coordinate values additionally representing the entrance / exit position of the road side of the building. The additional coordinate values are distinguished from each other by text attribute values such as “main gate” and “east gate”. The object coordinate value table and the additional coordinate value table are linked by “name”. In the case of “city hall”, the additional coordinate values of “main gate” and “east gate” are linked to the object coordinate values of the city hall itself.

  The POI search unit 24 searches for POI information in the POI data unit 23 according to conditions such as text attribute values and coordinate values. In FIG. 3A, when the vehicle travels in the direction A with the city hall as the destination, the POI search unit 24 searches for the additional coordinate value of “main gate”, not the object coordinate value of the city hall itself. At this time, even if an object coordinate value of the city hall itself or an additional coordinate value of “East Gate” is used, such a coordinate value does not appear on the screen. Similarly, when the vehicle travels in the direction B with the city hall as the destination, the POI search unit 24 searches for additional coordinate values of “East Gate”.

  The signal light data unit 25 has signal light information including the coordinate value of the signal light on the road, and the signal light search unit 26 searches the signal light data unit 25 for necessary signal light information in the calculation unit 40 and calculates the calculation unit. 40.

  The image recognition unit 30 includes a lane recognition unit 31, a traveling roadway recognition unit 32, and a signal light recognition unit 33.

  The lane recognition unit 31 recognizes a lane from an image acquired using a camera. For example, the lane recognition unit 31 performs image processing on a video in front of the vehicle and extracts lane data on the left and right sides of the vehicle, and then compares the lane data with prestored roadway recognition pattern data. By this, the lane can be recognized. The lane recognition unit 31 transmits the value recognized as such in the form of a straight line indicated by parameters on the image coordinates to the traveling roadway recognition unit 32 and the calculation unit 40.

  The traveling roadway recognition unit 32 calculates the current traveling roadway position using the lane recognized from the video acquired using the camera, and transmits the result to the calculation unit 40. For example, when a lane is drawn on the road so that each roadway can be distinguished from other roadways, for example, the center line is straight, the right lane of the lane is broken, the right side of the lane When the lane is classified as a one-dot chain line, etc., recognition pattern data related to the roadways to be distinguished from each other is stored in advance, and the lane recognition pattern that matches the lane recognition pattern recognized by the lane recognition unit 31 is stored. By searching from the recognized pattern data, it is possible to calculate the current traveling road position using a known method for determining which traveling road the vehicle is currently traveling.

  The process of the signal light recognition unit 33 recognizing the signal light is similar to the process of the lane recognition unit 31 recognizing the lane. That is, the signal lamp recognition unit 33 processes the vehicle front image acquired using the camera, and uses video data that matches the signal lamp recognition pattern data stored in advance as the signal lamp from the video data obtained by such processing. Recognize and transmit the recognized image coordinate value of the signal lamp to the calculation unit 40. The image coordinate value is a coordinate value in an image displayed on the screen of the image car navigation system.

  The computing unit 40 calculates the vehicle position and the vehicle azimuth from the coordinate values received from the map matching processing unit 22 and the video recognition result received from the video recognition unit 30.

  The position calculation unit 41 corrects the current vehicle position recognized by the map matching data received from the map matching processing unit 22 based on the traveling road position information received from the traveling road recognition unit 31. As described above, since the map matching processing unit 22 performs map matching on the vehicle position to the road center line, correction of the vehicle position is necessary in order to grasp the exact position of the vehicle. The position calculation unit 41 recognizes the traveling road position of the vehicle based on the traveling road position received from the traveling road recognition unit 32. Therefore, the position calculation unit 41 uses the traveling roadway position and the roadway width to calculate the distance at which the vehicle is away from the road centerline in a direction perpendicular to the road centerline, thereby matching the road centerline with the vehicle. Can be corrected to the actual vehicle position. At this time, if there is precise data representing the width of each roadway, the precise data is used for the roadway width; otherwise, a value that is standardized by region (for example, in the case of an urban road, 3.5 m) is used.

  The azimuth calculation unit 42 calculates the vehicle azimuth using the lane information received from the lane recognition unit 32. In many cases, since the lane is parallel to the road center line, the angle that the vehicle makes with the road can be considered to be the same as the angle that the vehicle makes with the lane. However, when the vehicle is actually traveling, the lane and the vehicle are not parallel and may have an angle. In this case, the angle between the road and the N axis of the map and the lane of the road The vehicle azimuth is calculated by adding the angle between

  The method of calculating the angle that the vehicle makes with the lane by the azimuth calculating unit 42 is, for example, as follows. A virtual normal is drawn from the left and right lanes of the vehicle on the screen recognized by the lane recognition unit 32 toward the center of both lanes, and a virtual horizontal axis passing through a point on the screen where the two normals intersect is drawn. If the first angle formed by the normal line from the left lane and the horizontal axis is equal to the second angle formed by the normal line from the right lane and the horizontal axis, the angle formed by the vehicle with the lane is 0 degrees. You can say that. In this way, the angle that the vehicle makes with the lane can be calculated from the difference between the first angle and the second angle. As the difference between the first angle and the second angle increases, the angle that the vehicle makes with the lane increases proportionally. The vehicle azimuth angle calculated in this way is selectively used by the output unit 50 together with the vehicle azimuth angle provided by the GPS receiver 12.

  The screen reference area setting unit 43 divides the screen of the video car navigation system into a plurality of reference areas using the image coordinate value of the signal lamp received from the signal lamp recognition unit 33 as a reference coordinate value, and the POI is displayed on the screen. In order to approximate the position, the POI position information (coordinate value) received from the POI search unit 24 is compared with the signal lamp position information (coordinate value) received from the signal lamp search unit 26, and the POI is displayed on the screen. Correspond to the screen reference area divided above. That is, each POI is appropriately displayed on the screen according to the relative position of the POI with respect to the position of the signal lamp.

  Next, with reference to FIG. 4, a method in which the screen reference area setting unit 43 associates the POI with the reference area divided on the screen will be described. On the map of FIG. 4A, a map area is represented by a broken line (HL) connecting the coordinates of the signal light and the vehicle position, and a broken line (VL) perpendicular to the current traveling road and passing through the signal light coordinates. It is divided into four areas. A, B, C, D, and E on the map indicate POI.

  FIG. 4B shows a screen of the video car navigation system. First, the screen reference area setting unit 43 displays the signal lamp on the screen using the image coordinate value of the signal lamp received from the signal lamp recognition unit 33, and the four screen reference areas S1, S2, S3, And divide the screen into S4.

  Next, the screen reference area setting unit 43 displays the POIs of A, C, and E displayed on the left side of the signal light on the map when viewed from the vehicle, and the screen reference area (S1, Corresponding to S3), the POIs of B and D are made to correspond to the screen reference area (S2, S4) on the right side of the signal lamp on the screen. Further, the screen reference area setting unit 43 associates the POIs of A and B that are closer to the signal lamp when viewed from the vehicle with the screen reference areas (S1, S2) above the signal lamp on the screen, and C, The POIs of D and E are made to correspond to the screen reference area (S3, S4) below the signal lamp on the screen. As a result, each POI of A, B, C, D, and E corresponds to one screen reference area among the four screen reference areas on the screen. In the example of FIG. 4, the POI of E is arranged in the screen reference area of S3, and the POI of B is arranged in the screen reference area of S2.

  On the other hand, since the screen reference area setting unit 43 can recognize the coordinate value of the signal lamp and the coordinate value of each POI, the screen reference area setting unit 43 uses the relative coordinate value of each POI based on the coordinate value of the signal lamp. A somewhat accurate POI display position can be determined on the top. Thereby, the screen reference area setting unit 43 can determine the display position of the POI on the screen from the relative coordinate value of each POI with respect to the coordinate value of the signal light without dividing the screen reference area. However, the reason why the reference area is divided in this embodiment is that the screen reference area setting unit 43 corresponds to the POI image coordinates calculated by projecting the 3D map coordinates onto the 2D image coordinates in the output unit 50 described later. This is because if it does not correspond to the reference region, it can be confirmed that the image coordinates are not accurately calculated.

  The output unit 50 receives the input of the vehicle position, the vehicle azimuth angle, the camera parameters known in advance, the POI data, and the screen reference area, and calculates the 3D map coordinates of the output target POI by projecting to the 2D image coordinates. Outputs the POI image coordinates. At this time, the output unit 50 is one of the vehicle azimuth provided from the GPS receiver 12 and the vehicle azimuth calculated by the azimuth calculation unit 42 according to the vehicle speed provided from the GPS receiver 12. The image coordinates of the POI are calculated using one vehicle azimuth.

  Specifically, for a vehicle speed of 20 km / h or more, which is known to have reliable GPS information, the vehicle azimuth provided by the GPS receiver 12 is used, and the vehicle speed is less than 20 km / h. The vehicle azimuth calculated by the azimuth calculation unit 42 is used. Further, when the lane cannot be correctly recognized due to congestion in front of the vehicle, the vehicle azimuth provided from the GPS receiver 12 is used regardless of the vehicle speed.

  Further, the output unit 50 checks whether or not the in-screen area of the calculated POI image coordinates matches the screen reference area associated by the screen reference area setting unit 43. The image coordinates on the screen reference area associated by the screen reference area setting unit 43 that are not the POI image coordinates are output to the screen as the POI image coordinates.

  In addition, the output unit 50 outputs a text attribute value of POI data (for example, a building name) in the form of a three-dimensional nameboard with a tag on the left side of the object (such as a building) on the left side of the screen. Name) is output to the video, and for the object located on the right side of the screen, the text attribute value of the POI data is output to the video in the form of a three-dimensional nameboard with a tag on the right side of the object. Further, the name board output to the video is configured to be transparent to some extent so as not to hide other information on the screen.

  FIG. 5 shows a flowchart of the operation of the video car navigation system according to the present embodiment. When the system is started (S61), the camera 11 of the data input unit 10 acquires an image ahead of the vehicle at a constant frame rate, and the GPS receiver 12 detects the current position coordinate value and the vehicle azimuth at regular time intervals. The vehicle speed is detected (S62).

  The data input unit 10 transmits the video to the video recognition unit 30 and the output unit 50, and transmits the position coordinate value to the map matching processing unit 22 of the electronic map unit 20. The map matching processing unit 22 executes map matching that approximates the position coordinate value to the closest point (that is, the center line) on the road in the electronic road map unit 21 (S63), and the map coordinated position coordinate value Is transmitted to the arithmetic unit 40.

  The lane recognition unit 32 of the video recognition unit 30 recognizes the lane from the received video and transmits the result to the traveling roadway recognition unit 31. The traveling roadway recognition unit 31 uses the lane recognition result from the received video. The travel roadway position is calculated (S64). The signal lamp recognition unit 33 of the video recognition unit 30 recognizes the signal lamp from the received video (S65). The lane is in the form of image coordinates, the result of signal light recognition is in the form of image coordinates of the signal light arranged in the video, and the traveling road position is calculated in the form of numbers (that is, “2” in the case of two lanes). Is transmitted to the unit 40.

  The calculation unit 40 calculates the vehicle azimuth from the image coordinates of the lane, corrects the vehicle position from the traveling road position, and calculates the accurate vehicle position by the method described above (S66).

  Next, the POI search unit 24 searches the output target POI from the POI data (S68). The search method is as described above based on FIG. The POI search unit 24 transmits the searched POI data (coordinate values and text attribute values) to the screen reference area setting unit 43 of the calculation unit 40. The screen reference area setting unit 43 compares the coordinate value in the POI data with the vehicle position, and associates the POI coordinate output area with one screen reference area in the screen reference area.

  Finally, the output unit 50 receives the input of the vehicle position, the vehicle azimuth angle, the previously known camera parameters, the POI data, and the screen reference area, and projects the 3D map coordinates of the output target POI onto the 2D image coordinates. The image coordinates of the POI calculated in this way are output. At this time, the output unit 50 determines the vehicle azimuth and azimuth calculation unit 42 provided from the GPS receiver 12 according to the vehicle speed provided from the GPS receiver 12 (for example, based on 20 km / h). The image coordinates of the POI are calculated using any one of the vehicle azimuths calculated by the above. Further, the output unit 50 checks whether the calculated in-screen area of the image coordinates of the POI matches the screen reference area associated by the screen reference area setting unit 43, and if not, the calculated POI The image coordinates on the screen reference area associated by the screen reference area setting unit 43 that are not image coordinates are output as the image coordinates of the POI (S69).

  At this time, the output unit 50, for an object (such as a building) located on the left side of the screen, in the form of a three-dimensional nameboard with a tag on the left side of the object (for example, a building) For the object located on the right side of the screen, the text attribute value of the POI data is output to the video in the form of a three-dimensional nameboard with a tag on the right side of the object. Further, the name board output to the video is configured to be transparent to some extent so as not to hide other information on the screen.

  Such a process is repeatedly executed in real time for each frame, and ends when the system receives an end command or video input is interrupted (S70).

  The present invention can also be embodied by a computer-readable recording medium on which a program for executing the above method is recorded. The above-described embodiments of the present invention are specified by specific configurations and drawings, but the scope of the present invention is not limited by such specific embodiments. Therefore, the present invention should be understood as including various modifications and equivalents thereof without departing from the essence of the present invention.

1 is a configuration diagram illustrating a video car navigation system according to an embodiment of the present invention. It is a block diagram which shows the detail of the image | video car navigation system of FIG. It is a conceptual diagram for demonstrating a structure and search of POI data. It is a conceptual diagram for demonstrating screen reference area | region setting. It is a flowchart which shows operation | movement of the video car navigation system which concerns on embodiment of this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 Data input part 20 Electronic map part 30 Image | video recognition part 40 Calculation part 50 Output part 11 Camera 12 GPS receiver 21 Electronic road map part 22 Map matching process part 23 POI data part 24 POI search part 25 Signal lamp data part 26 Signal lamp search part 31 Lane recognition unit 32 Traveling road recognition unit 33 Signal lamp recognition unit 41 Position calculation unit 42 Azimuth angle calculation unit 43 Screen reference area setting unit

Claims (16)

A video car navigation system for displaying a target point on a screen,
A data input unit that acquires the screen using a camera mounted on the vehicle;
A video recognition unit for recognizing a reference object from the screen;
An electronic map unit for providing information on a map of the target point and the reference object;
A target point on the screen of the target point based on the reference object using a relative positional relationship between the position of the reference object on the map and the position of the target point on the map A calculation unit for determining the position;
An image car navigation system comprising: an output unit that displays the target point at the target point position on the screen. The calculation unit includes a screen reference region setting unit that divides the screen into a plurality of screen reference regions based on the reference object,
The screen reference area setting unit determines the position of the target spot by associating the target spot with one screen reference area of the plurality of screen reference areas. Video car navigation system. The output unit calculates a two-dimensional image coordinate from a position of the target point on the map, and when the two-dimensional image coordinate is not included in a screen reference region associated with the target point, The video car navigation system according to claim 2, wherein the target point is displayed at the target point position instead of displaying the target point in two-dimensional image coordinates. Before Symbol image recognition unit may include a recognizing lane recognition unit in front of the lane of the vehicle from the screen,
2. The video car navigation system according to claim 1, wherein the calculation unit includes an azimuth angle calculation unit that calculates a vehicle azimuth angle of the vehicle from information on the lane recognized by the lane recognition unit. The video recognition unit includes a traveling roadway recognition unit that recognizes a traveling roadway of the vehicle from information on a lane recognized from the screen,
The video car navigation system according to claim 1, wherein the calculation unit includes a position calculation unit that calculates a position of the vehicle from information related to the traveling road recognized by the traveling road recognition unit. The data input unit receives position information of the vehicle from a GPS receiver,
The position calculation unit calculates the position of the vehicle by correcting the position information of the vehicle received from the GPS receiver using information on the traveling road recognized by the traveling road recognition unit. The video car navigation system according to claim 5. The target point is associated with a plurality of target point coordinates on the map,
The calculation unit specifies any one of the plurality of target point coordinates as a position on the map of the target point in order to determine the relative positional relationship. The video car navigation system according to claim 1. The said calculating part specifies the target point coordinate nearest to the current position of the said vehicle among the said several target point coordinates to the position on the said map of the said target point. Video car navigation system. The video car navigation system according to claim 7, wherein the plurality of target point coordinates include coordinates of an entrance / exit of the target point. The video car navigation system according to claim 1, wherein the reference object is a signal light. A video car navigation method for displaying a target point on a screen,
(A) a step of acquiring the screen by a data input unit using a camera mounted on a vehicle;
(B) a step of recognizing a reference object from the screen by the video recognition unit;
(C) Using the relative positional relationship between the position of the reference object on the map provided by the electronic map unit and the position of the target point on the map, the calculation unit uses the relative position relationship as a reference. Determining a target point position on the screen of the target point
And (d) displaying the target point at the target point position on the screen by the output unit. The step (c)
Dividing the screen into a plurality of screen reference areas based on the reference object;
Determining the position of the target point by associating the target point with one screen reference area of the plurality of screen reference areas,
The step (d)
Calculating two-dimensional image coordinates from a position of the target point on a map;
When the two-dimensional image coordinates are not included in the screen reference area associated with the target point, instead of displaying the target point in the two-dimensional image coordinates, the target point is displayed at the target point position. The video car navigation method according to claim 11, further comprising a step of displaying. The step (b)
Recognizing a lane ahead of the vehicle from the screen,
The step (c)
Calculating a vehicle azimuth of the vehicle from information about the recognized lane,
The step (d)
Among the vehicle azimuths calculated in the step (c) and the vehicle azimuths provided from the GPS receiver as the vehicle azimuths used for calculating the two-dimensional image coordinates from the position of the target point on the map The video car navigation method according to claim 11, further comprising: selecting any one of the vehicle azimuth angles based on a current speed of the vehicle. The step (b)
Recognizing the traveling road of the vehicle from information on the lane recognized from the screen,
The step (c)
The video car navigation method according to claim 11, further comprising: calculating a position of the vehicle from information on the recognized traveling roadway. The target point is associated with a plurality of target point coordinates on the map,
The step (c)
In order to determine the relative positional relationship, the method includes the step of specifying any one of the plurality of target point coordinates as a position on the map of the target point. The video car navigation method according to claim 11. The step (b)
The video car navigation method according to claim 11, further comprising: recognizing a signal lamp as the reference object from the screen.

Images