JP5544595B2 - Map image processing apparatus, map image processing method, and computer program - Google Patents

Description

  The present invention relates to a map image processing device, a map image processing method, and a computer program.

  Information on facilities and roads on the surface of the earth is displayed on the display of electronic devices, including technology that provides route guidance to vehicles and pedestrians using computerized 2D map data that can be used electronically. Things are widely done. At present, the development of three-dimensional map data that reproduces the situation of the ground surface in a three-dimensional coordinate space is also in progress (for example, Patent Document 1).

JP 2005-346650 A JP 2004-126633 A JP 2003-111128 A

  In order to create such 3D map data realistically using a technique such as texture mapping, there is a problem that an enormous amount of work is required, and a technology for efficiently creating 3D map data is required. ing.

  An object of this invention is to provide the technique which produces efficient three-dimensional map data.

  SUMMARY An advantage of some aspects of the invention is to solve at least a part of the problems described above, and the invention can be implemented as the following forms or application examples.

[Application Example 1] A map image processing apparatus, which stores map data;
A coordinate conversion unit that perspectively projects the map data under a predetermined perspective projection condition and converts the map data into a two-dimensional image;
A captured image acquisition unit that acquires a captured image of the real world in the traveling direction of the vehicle;
A road surface sign recognition unit for recognizing a road sign actual image included in the captured image;
The road sign actual image and the road sign projection image corresponding to the road sign actual image included in the two-dimensional image converted using the perspective projection condition are compared, and the road sign actual image and the road sign are compared. A perspective projection condition correcting unit that corrects the perspective projection condition so that the projection image matches;
A map image processing apparatus comprising:

  According to the map image processing apparatus of the application example 1, the perspective projection conditions of the two-dimensional image obtained by perspective projection of the map data are set so that the photographed image obtained by photographing the real world and the two-dimensional image overlap with each other based on the road surface sign. doing. The road surface signs in the traveling direction of the vehicle can be recognized efficiently because the timing at which photographing can be performed without being obstructed by an obstacle occurs. Since the position of the road surface marker is stably fixed, it is possible to accurately superimpose a two-dimensional image obtained by perspective projection of map data on the captured image.

[Application Example 2] The map image processing apparatus according to Application Example 1,
By associating the feature image included in the captured image with the feature included in the two-dimensional image that is perspective-projected under the modified perspective projection condition, the image of the feature included in the captured image is obtained. A map image processing apparatus comprising: a texture application unit that texture maps the map data.

[Application Example 3] The map image processing apparatus according to Application Example 1,
The road surface sign recognition unit analyzes the captured image and sets a plurality of feature points of the road surface sign actual image on the captured image,
The perspective projection condition correcting unit corrects the perspective projection condition so that a plurality of feature points of the road sign projected image included in the two-dimensional image and a plurality of feature points of the road sign actual image are matched. A map image processing apparatus.

  The present invention can be realized in various forms, for example, in the form of a map image processing method, a computer program for creating map data, a recording medium on which the computer program is recorded, and the like. be able to.

It is explanatory drawing which shows schematic structure of the vehicle 100 which comprises some map image processing systems as an Example. It is explanatory drawing which shows schematic structure of the map image processing apparatus which comprises some map image processing systems as an Example. It is explanatory drawing explaining the outline | summary of the three-dimensional map data stored in the three-dimensional map data storage part D13. It is a figure explaining perspective projection. It is a figure which shows the image 500 which converted the image data D131 which is three-dimensional data into the two-dimensional image on a certain perspective projection condition. It is a figure which shows the frame image Fn which actually image | photographed the place which the two-dimensional image shown in FIG. 5 shows from the substantially same viewpoint. It is a flowchart showing the process step of map image processing. It is a 1st figure explaining the image process performed as a pre-processing in the case of a road surface sign recognition process. It is a 2nd figure explaining the image process performed as a pre-processing in the case of a road surface sign recognition process. It is a 3rd figure explaining the image process performed as a pre-process in the case of a road surface sign recognition process. It is a 4th figure explaining the image processing performed as a pre-processing in the case of a road surface marking recognition process. It is a 5th figure explaining the image processing performed as a pre-processing in the case of a road surface marking recognition process. It is a 1st figure shown about a shape recognition process. It is a 2nd figure shown about a shape recognition process. It is explanatory drawing explaining a temporary camera position estimation process. It is a figure explaining the setting of the initial value of perspective projection conditions, and the feature point of a road marking. It is a figure explaining a perspective projection condition correction process. FIG. 6 is a diagram illustrating an image obtained by optimizing the two-dimensional image 500 illustrated in FIG. 5 and adjusting the image to overlap the frame image Fn illustrated in FIG. 6. It is a figure explaining a texture application process.

  Hereinafter, embodiments of the present invention will be described based on examples with reference to the drawings.

A. Example:
・ Configuration of map image processing system:
FIG. 1 is an explanatory diagram showing a schematic configuration of a vehicle 100 constituting a part of a map image processing system as an embodiment. As illustrated, the vehicle 100 includes an imaging device 110, a GPS receiver 120, and a calculator 130.

  The imaging device 110 is a camera device that can generate digital or analog image data at a predetermined time interval, such as a digital video camera or a digital still camera capable of continuous shooting. The imaging device 110 is operable in accordance with the control of the computer 130 and is communicably connected to the computer 130 so that the generated image data can be transmitted to the computer 130. For example, the imaging device 110 is installed so as to photograph the front of the vehicle 100 through the front window of the vehicle 100.

  The GPS receiver 120 is a device that receives radio waves transmitted from artificial satellites that constitute a GPS (Global Positioning System / Global Positioning System). The GPS receiver 120 converts radio waves received from the artificial satellite into digital data (GPS information) and transmits the digital data to the computer 130.

  The computer 130 is a known personal computer such as a notebook computer. The computer 130 includes a central processing unit (CPU) (not shown), an internal storage device such as a RAM and a ROM, and an external storage device such as a hard disk. Based on the GPS information acquired from the GPS receiver 120, the computer 130 generates current position information (latitude and longitude) representing the current position of the vehicle 100 at regular intervals, and always recognizes the current position of the vehicle 100. Yes. The vehicle 100 may further include an odometer or a gyro sensor to improve the recognition accuracy of the current position by the computer 130. The computer 130 uses the current position when the image capturing apparatus 110 captures the image as the image capturing position, and stores the image data received from the image capturing apparatus 110 in the external storage device in association with the image capturing position of the captured image represented by the image data. The shooting position is the current position of the vehicle 100 at the shooting time.

  By taking an image with the imaging device 110 while traveling the route on which the vehicle 100 is to create 3D map data, the computer 130 can store the captured image of the route on which the 3D map data is to be created and the current position as data. it can.

  FIG. 2 is an explanatory diagram illustrating a schematic configuration of a map image processing apparatus that constitutes a part of the map image processing system according to the embodiment. The map image processing apparatus 200 is a well-known computer, and includes a CPU 201, an input / output unit 202, an internal storage device 203 such as a ROM or a RAM, and an external storage device 204 such as a hard disk. The input / output unit 202 is an interface for exchanging data with an external device, and is connected to the computer 130 described above. As a result, the map image processing apparatus 200 can acquire the image data of the photographed image stored in the computer 130 and the photographing position information indicating the photographing position of the photographed image.

  A map image processing program is stored in the internal storage device 203, and the functions of the functional blocks illustrated in the internal storage device 203 in FIG. 2 are realized by the CPU 201 executing the program. The realized functional blocks are a captured image acquisition unit M11, a road surface sign recognition unit M12, a three-dimensional map data acquisition unit M13, a coordinate conversion unit M14, a temporary camera position estimation unit M15, and a perspective projection condition correction unit M16. And a texture application unit M17.

  The captured image acquisition unit M11 acquires the image data captured by the above-described imaging device 110 and the imaging position information associated with the image data via the input / output unit 202. In this embodiment, the imaging device 110 is a video camera, and a moving image composed of a plurality of frame images and shooting position information associated with each frame image are acquired. As will be described later, the road surface sign recognition unit M12 analyzes one frame image to recognize a predetermined road surface sign, and sets a plurality of feature points of the recognized road surface sign on the frame image. The 3D map data acquisition unit M13 acquires 3D map data described later from the 3D map data storage unit. The coordinate conversion unit M14 perspectively projects the three-dimensional map data under a predetermined perspective projection condition, and performs coordinate conversion into a two-dimensional image. As a result, a two-dimensional image obtained by viewing the three-dimensional map data from a predetermined viewpoint can be displayed on a display or the like. The temporary camera position estimation unit M15 estimates the shooting direction and shooting position of the frame image from the shooting position information associated with the frame image and the shooting position information associated with the frame image group before and after the frame image. To do. The temporary camera position estimation unit M15 temporarily determines the perspective projection conditions for the frame image and the three-dimensional map data from the estimation result of the shooting direction and the shooting position of the frame image. The perspective projection condition correcting unit M16 performs an optimization process, which will be described later, with the provisional projection condition determined as an initial value, and finally the two-dimensional image obtained by perspective projection of the three-dimensional map data and the frame image are accurately obtained. Modify to overlap perspective projection conditions. The texture application unit M17 associates an image of a feature such as a building included in the frame image with a feature such as a building included in the two-dimensional image that is perspective-projected under the corrected perspective projection condition. Texture mapping of an image of a feature such as a building to be included in 3D map data.

  The external storage device 204 includes a model code data storage unit D11, an analysis data storage unit D12, and a 3D map data storage unit D13. The model code data storage unit D11 stores a model chain code (described later) in which the shape of the road surface sign is converted into data. The analysis data storage unit D12 stores analysis data such as a road surface sign recognition result by image analysis. The 3D map data storage unit D13 stores 3D map data represented in a 3D coordinate space described later.

  FIG. 3 is an explanatory diagram for explaining the outline of the 3D map data stored in the 3D map data storage unit D13. The three-dimensional map data includes image data D131 and route data D132. The image data D131 is three-dimensional data representing the surface condition in the three-dimensional coordinate space (X, Y, Z). As the image data D131, a surface image that represents the surface of the ground two-dimensionally is drawn in more detail, and includes accurate data such as a two-dimensional shape of a feature such as a road surface sign RH or a building BL. On the other hand, the image data D131 has relatively little data in the height direction. For example, a feature such as a tall building BL has only information about the height h of the building BL.

  The route data D132 is data relating to a traffic route existing on the surface of the ground surface of the map image represented by the map image data. The route data D132 is data representing a network indicating the arrangement of traffic routes on the ground surface. The network data includes node data representing element points ND (referred to as nodes) in the traffic route and link data representing line segments LK (referred to as links) connecting the nodes. The node ND represents, for example, an intersection, a branch point, an end point, a start point, a boarding / exiting position such as a station, and the like. The link LK represents a traffic route such as a roadway, a track, and a pedestrian road.

  FIG. 4 is a diagram illustrating perspective projection. The three-dimensional map data represented in the three-dimensional coordinate space (X, Y, Z) is converted into a two-dimensional coordinate space (U, V) for two-dimensional display on a display or the like. The projection center EP is placed at a predetermined position in the three-dimensional space, and the projection plane TM is placed in front of it. A projection line is drawn from each point of the three-dimensional figure A toward the projection center. Then, the two-dimensional figure a is formed as a collection of intersections of the projection plane and the projection line. Such a projection method is called perspective projection. An image photographed by a camera is generally a perspective projection. The projection plane TM has a predetermined rectangular size, and only an image projected in this range is drawn as a two-dimensional image. The size of the projection plane TM is determined by the angle of view GK from the projection center and the distance d from the projection center EP to the projection plane, assuming that the aspect ratio is constant. FIG. 4 shows an example in which the projection center EP is placed at the origin of the three-dimensional coordinate space (X, Y, Z) and the center of the projection plane (projection plane center) CP is on the Z axis.

  The conditions (perspective projection conditions) that uniquely define a two-dimensional image obtained by projecting and projecting predetermined three-dimensional data are the coordinates (Ex, Ey, Ez) of the projection center EP and the coordinates (Cx, Xy, Cz) and an up vector UV (Ux, Uy, Uz) representing a rotation component about the angle of view GK and the line-of-sight direction. If these nine variables are determined, a two-dimensional image obtained by projecting and projecting predetermined three-dimensional data can be uniquely determined.

  FIG. 5 is a diagram showing an image 500 obtained by converting image data D131, which is three-dimensional data, into a two-dimensional image under certain perspective projection conditions. The image 500 may include a building 501 such as a building, a road 505, road markings 502 to 504 drawn on the road 505, and the like.

  FIG. 6 is a diagram showing a frame image Fn obtained by actually capturing the place indicated by the two-dimensional image shown in FIG. 5 from substantially the same viewpoint. The image 500 shown in FIG. 5 and the frame image Fn do not completely coincide with each other in viewpoint and angle of view, and do not completely overlap.

-Operation of the map image processing apparatus 200:
Next, map image processing for texture mapping the frame image Fn to the image data D131 will be described. FIG. 7 is a flowchart showing processing steps of map image processing. First, the CPU 201 reads the frame image Fn into the internal storage device 203 together with the shooting position of the frame image Fn and the shooting position of the frame image group before the frame image Fn (step S100).

  The CPU 201 executes road surface sign recognition processing for the read frame image Fn (step S110). The road surface sign recognition process is a process for recognizing a road surface sign reflected in the frame image Fn by analyzing the frame image Fn. In the present embodiment, the road surface marker is recognized as an arrow-shaped traveling direction indicating straight ahead, left turn, right turn or the like as the road surface marker. 8 to 12 are diagrams for describing image processing performed as pre-processing in the road surface sign recognition processing. FIG. 8 shows a frame image Fn (original image) before image processing. First, the CPU 201 binarizes the frame image Fn into white and black. FIG. 9 shows the frame image Fn after the binarization process. Since the road surface marker related to the arrow-shaped traveling direction is generally white, it is included in the white region in the binarized frame image Fn. The CPU 201 performs noise removal processing (FIG. 10) and contour extraction processing (FIG. 11) of the frame image Fn after binarization processing. The noise removal process is performed using a known noise removal filter, and the contour line extraction process is performed by a known edge extraction process. The CPU 201 further thins the contour line extracted by the contour line extraction process. All the thinned lines are closed loop (FIG. 12).

  The CPU 201 performs shape recognition processing for each thinned loop. 13 and 14 are diagrams showing the shape recognition process. By applying the 8-way approximation matrix shown in FIG. 13 to each closed loop, each closed loop is converted into a chain code corresponding to its shape. The model code data storage unit D11 stores model data chain code data in which the shape of the road surface marker to be searched is represented by a chain code.

  The CPU 201 calculates the cost by DP matching for each closed-loop chain code (extracted chain code) and the model data chain code (FIG. 14). DP matching is a typical expansion / contraction matching method, and it is determined that both data are approximated as the cost value between the two data is small, and that both data are not approximated as the cost value is large. When the cost value obtained by DP matching between the extracted chain code and the model data chain code is smaller than a predetermined threshold, the CPU 201 indicates that the closed loop corresponding to the extracted chain code is a road surface sign represented by the model data chain code. Judge that there is.

  When the road surface sign recognition process ends, the CPU 201 determines whether or not a desired road surface sign exists on the frame image Fn (step S120). Here, if it is determined that there is no road marking in the frame image Fn (step S120: NO), the CPU 201 returns to step S100 and executes the above-described processing for another frame image Fn. If it is determined that there is a road marking in the frame image Fn (step S120: YES), the CPU 201 executes a temporary camera position estimation process (step S130).

  FIG. 15 is an explanatory diagram illustrating the temporary camera position estimation process. Here, it is assumed that the road surface sign M0 recognized in the frame image Fn is a road surface sign in which a straight arrow and a right turn arrow are combined as shown in FIG. In the image data D131 of the three-dimensional map, the CPU 201 searches for a road surface sign having the same shape as the road surface mark M0 with a range within a predetermined distance from the shooting position P0 of the frame image Fn as a search range TA. As a result, the road surface sign M1, the road surface sign M3, and the road surface sign M5 are searched as candidates. Further, the CPU 201 estimates the traveling direction R of the vehicle 100 at the shooting position P0 based on the shooting positions P1 to P6 of the frame image group before the frame image Fn. As a result, the CPU 201 determines that the road marking M0 recognized in the frame image Fn is the road marking M5 in the three-dimensional map image data D131. The CPU 201 tentatively determines that a position DP that is a predetermined distance from the road surface sign M5 is a position where the frame image Fn is captured.

  When the temporary camera position estimation process ends, the CPU 201 determines whether or not the camera position has been provisionally determined (step S140). If it is determined that provisional determination cannot be made (step S140: NO), the CPU 201 returns to step S100 and executes the above-described processing for another frame image Fn. If it is determined that provisional determination has been made (step S140: YES), the CPU 201 executes perspective projection condition correction processing (step S150).

  In the perspective projection condition correction processing, the recognized road surface sign M0 in the frame image Fn and the two-dimensional image TM of the road surface sign M5 corresponding to the road surface sign M0 in the three-dimensional map image data D131 are accurately overlapped. This is a process for correcting the perspective projection condition of the projection surface TM.

  As described above, the perspective projection conditions include the coordinates (Ex, Ey, Ez) of the projection center EP, the coordinates CP (Cx, Xy, Cz) of the projection plane center, and the up vector UV (Ux, Uy, Uz) 9. One value. FIG. 16 is a diagram for explaining the initial values of the perspective projection conditions and the setting of the feature points of the road marking. In this embodiment, there is a restriction that the distance d between the projection center EP and the projection plane center CP is a constant value, and a restriction that the straight line connecting the projection center EP and the projection plane center CP and the up vector UV are orthogonal. Yes. The initial value of the projection center EP is set to the temporarily determined camera position in the previous step. The initial value of the projection plane center CP is determined using the direction vector CV associated with the road marking M5. The initial value of the up vector UV is set to an appropriate value in consideration of the angle of view of the imaging device 110. That is, the initial value of the projection plane center CP is set to a position advanced by a distance d from the projection center EP in the direction of the direction vector CV. Further, the feature points PR (PRx, PRy, PRz) of the road surface sign M5 are set at, for example, each vertex of the road surface sign M5 as shown in FIG. Alternatively, the feature point PR of the road surface sign M5 may be set at each vertex of a rectangle circumscribing the road surface sign M5 as shown in FIG.

  Although illustration is omitted, for the road surface sign M0 recognized in the frame image Fn, the characteristic points FP (Uc, Vc) are set at positions corresponding to the characteristic points PR set on the road surface sign M5.

FIG. 17 is a diagram illustrating the perspective projection condition correction process. First, the CPU 201 converts the feature point PR (PRx, PRy, PRz) of the road surface marker M5 into coordinates VP (Ur, Vr) on a two-dimensional image that is perspective-projected under the set perspective projection conditions. Next, the CPU 201 calculates a sum S for each feature point of the distance between the feature point VP (Ur, Vr) on the two-dimensional image and the feature point FP (Uc, Vc) of the corresponding frame image Fn. The calculation formula is represented by the following formula (1).

  In Equation 1, i is an identifier for identifying each feature point, and N is the number of feature points. The perspective projection condition correction process results in an optimization problem for obtaining a perspective projection condition that minimizes the sum S obtained by Equation 1. Various optimization algorithms for obtaining a perspective projection condition that minimizes the sum S while gradually changing the perspective projection condition can be used for the optimization process. In this embodiment, the well-known simulated door annealing method ( (Annealing method) will be used.

  When the perspective projection condition is finally corrected, the frame image Fn and the two-dimensional image obtained by perspective projection of the image data D131 of the three-dimensional map are accurately overlapped in the entire image. FIG. 18 is a diagram showing an image that has been optimized so that it overlaps the frame image Fn shown in FIG. 6 by optimizing the two-dimensional image 500 shown in FIG. As shown in FIGS. 18 and 6, it can be seen that the two-dimensional image obtained by perspective projection of the image data D131 of the three-dimensional map and the frame image Fn accurately overlap by the perspective projection condition setting process.

  When the perspective projection condition setting process ends, the CPU 201 executes a texture application process (step S160). The texture application process associates a realistic image of a feature such as a building BL included in the frame image Fn with a corresponding feature included in the two-dimensional image that is perspective-projected under the perspective projection condition corrected in step S150. Thus, the real image of the feature such as the wall surface of the building BL included in the frame image Fn is mapped to the image data D131 of the three-dimensional map. FIG. 19 is a diagram illustrating the texture application process. The pixel value (RGB value, HSV value, etc.) of the pixel TP on the frame image Fn is associated with the coordinate MP on the two-dimensional image 500 corresponding to the pixel TP. Since the frame image Fn and the two-dimensional image 500 are associated with each other so as to be correctly overlapped by the perspective projection condition setting process, the coordinate MP on the two-dimensional image 500 that associates the pixel TP on the frame image Fn is obtained. Easy. The pixel value of the pixel TP on the frame image Fn is finally associated (mapped) with the coordinates XP (a, b, c) on the three-dimensional map inversely transformed under the perspective projection conditions corrected in step S150. The

  According to the embodiment described above, the perspective projection conditions of the two-dimensional image obtained by perspective projection of the three-dimensional map data are set so that the photographed image obtained by photographing the real world and the two-dimensional image are overlapped with reference to the road surface sign. ing. The road surface signs in the traveling direction of the vehicle can be recognized efficiently because the timing at which photographing can be performed without being obstructed by an obstacle occurs. Since the position of the road surface marker is stably fixed, it is possible to accurately superimpose a two-dimensional image obtained by perspective projection of the three-dimensional map data on the captured image. Further, after obtaining the perspective projection conditions so that the road marking on the two-dimensional image obtained by perspective projection of the image data D131 of the three-dimensional map overlaps with the corresponding road marking on the captured image, the contents of the captured image are obtained. Is mapped to the image data D131 of the three-dimensional map, so that the texture mapping can be automatically performed with high accuracy on the image data D131 of the three-dimensional map. Furthermore, since the perspective projection conditions are obtained using road markings that currently have accurate and substantial data, it is possible to enrich the image data D131 of a three-dimensional map near many intersections.

B. Modifications-First modification:
In the above-described embodiment, the two-dimensional image that has been perspective-projected and the photographed image are superimposed using the road surface sign relating to the traveling direction, but other road sign may be used. The other road surface sign is preferably one in which accurate position data is prepared in the image data D131 of the three-dimensional map, and in general, a feature drawn two-dimensionally on the ground surface is preferable. For example, a sign related to a maximum speed, a sign related to a vehicle traffic section, a sign indicating that there is a pedestrian crossing or a bicycle crossing band, and the like can be used.

・ Second modification:
In the above embodiment, a part of the configuration realized by hardware may be replaced by software, and conversely, a part of the configuration realized by software may be replaced by hardware.

Third modified example In the above embodiment, it is used for texture mapping, but for a vehicle equipped with a camera device, a two-dimensional image in which a perspective projection condition is set is displayed on a display near the driver's seat and used for navigation. May be. In the present invention, since the position of the vehicle and the projection center of the two-dimensional image can be made substantially coincident, for example, navigation for grasping in which lane the vehicle is located can be performed.

Fourth Modification In the above embodiment, the three-dimensional map data is perspectively projected and converted into a two-dimensional image. However, the two-dimensional map data is perspectively projected and two-dimensional as an overhead view from a specific angle. You may convert into an image. When texture mapping is performed on the feature image of the photographed image using this two-dimensional image, a contour line is extracted based on the feature image of the photographed image, and inversely transformed under the perspective projection conditions to obtain a three-dimensional wire. It is also possible to generate 3D map data by generating a frame.

  As mentioned above, although the Example and modification of this invention were demonstrated, this invention is not limited to these Example and modification at all, and implementation in a various aspect is possible within the range which does not deviate from the summary. It is.

    DESCRIPTION OF SYMBOLS 100 ... Vehicle, 110 ... Imaging device, 120 ... GPS receiver, 130 ... Calculator, 200 ... Map image processing device, 201 ... CPU, 202 ... Input / output unit, 203 ... Internal storage device, 204 ... External storage device

Claims (5)

A map image processing device, a storage unit for storing map data;
A coordinate conversion unit that perspectively projects the map data under a predetermined perspective projection condition and converts the map data into a two-dimensional image;
A captured image acquisition unit that acquires a captured image of the real world in the traveling direction of the vehicle;
A road surface sign recognizing unit for recognizing a road surface sign photographed image relating to an arrow-shaped traveling direction included in the captured image;
The road sign actual image and the road sign projection image corresponding to the road sign actual image included in the two-dimensional image converted using the perspective projection condition are compared, and the road sign actual image and the road sign are compared. A perspective projection condition correcting unit that corrects the perspective projection condition so that the projection image matches;
A map image processing apparatus comprising: The map image processing apparatus according to claim 1,
The road surface sign recognition unit analyzes the captured image and sets a plurality of feature points of the road surface sign actual image on the captured image,
The perspective projection condition correcting unit corrects the perspective projection condition so that a plurality of feature points of the road sign projected image included in the two-dimensional image and a plurality of feature points of the road sign actual image are matched. A map image processing apparatus. The map image processing apparatus according to claim 1,
By associating the feature image included in the captured image with the feature included in the two-dimensional image that is perspective-projected under the modified perspective projection condition, the image of the feature included in the captured image is obtained. A map image processing apparatus comprising: a texture application unit that texture maps the map data. A map generation method using an information processing apparatus having a storage unit for storing map data,
The map data is perspective-projected under predetermined perspective projection conditions and converted into a two-dimensional image,
Acquire a captured image of the real world in the direction of travel of the vehicle,
Recognizing the road surface photographed image regarding the direction of travel of the arrow included in the captured image,
The road sign actual image and the road sign projection image corresponding to the road sign actual image included in the two-dimensional image converted using the perspective projection condition are compared, and the road sign actual image and the road sign are compared. A map image processing method for correcting the perspective projection condition so that the projection image matches. A computer program for performing map image processing on an information processing apparatus having a storage unit for storing map data,
A coordinate conversion function for perspectively projecting the map data under a predetermined perspective projection condition and converting the map data into a two-dimensional image;
A captured image acquisition function for acquiring a captured image of the real world in the traveling direction of the vehicle;
A road surface sign recognition function for recognizing a road surface actual image related to the arrow-shaped traveling direction included in the captured image;
The road sign actual image and the road sign projection image corresponding to the road sign actual image included in the two-dimensional image converted using the perspective projection condition are compared, and the road sign actual image and the road sign are compared. A perspective projection condition correction function for correcting the perspective projection condition so that the projection image matches;
A computer program for causing a computer of the information processing apparatus to realize the above.

Images