JPH07195978A - Vehicle surroundings display unit - Google Patents

Abstract

PURPOSE:To properly and visually recognize your own vehicle and obstacles around the vehicle even if the position of your own vehicle is slipped off from the visual field center of a camera. CONSTITUTION:Obstacles and parking areas which are hidden due to the image deformation of your own vehicle, are properly indicated by synthesizing an image the deformation of which has been corrected (coordinate transformation) in response to the position and direction of your own vehicle, and an image of other areas excluding your own vehicle, by a whole image synthesizing means 109 so as to be displayed wherein the aforesaid position and direction of your own vehicles outputted by a vehicle coordinate transformation means 105, a back ground area extracting means 107 and a vehicle area synthesizing means 108 after your own vehicle has been extracted from an image photographed by a vehicle extracting means 104.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention captures an image of a vehicle and its surroundings in a parking lot, and extracts and synthesizes the contour of the vehicle, its surroundings, background, etc. by image processing to produce an image of the vehicle. The present invention relates to a vehicle surroundings display device which is transmitted to a vehicle and displayed on a monitor.

[0002]

2. Description of the Related Art As a conventional vehicle surrounding display device,
For example, there is a "vehicle safety confirmation system" disclosed in Japanese Patent Laid-Open No. 4-123945. This is to prevent contact accidents in the parking lot by capturing an image of the parking space and its surroundings with a camera installed on the ceiling of the parking lot and transmitting this captured image from the transmitter to the receiver of the vehicle. It is intended.

[0003]

However, in the conventional vehicle surroundings display device as described above, the vehicle shape on the captured image may differ depending on the state of the position of the camera and the vehicle. is there. For example,
As shown in FIG. 29, when the three-dimensional object 2901 is imaged from above, the shape of the three-dimensional object 2901 can be accurately determined when the visual field center 2902 of the camera is almost directly above the three-dimensional object 2901 as shown in (a). Can be captured, (b)
When the three-dimensional object 2901 and the visual field center 2902 of the camera are deviated as shown in (c) or (c), the vehicle appears distorted unlike the image in (a) when viewed from directly above.

As described above, since the host vehicle is a three-dimensional object having a height, the position of the camera and the host vehicle on the two-dimensional image is displaced, which causes distortion of the vehicle shape and the distance from the obstacle. Despite this, there was a problem that the vehicle and obstacles appeared to overlap and it was difficult to grasp the sense of distance.

The present invention has been made in view of the above, and accurately recognizes the own vehicle and obstacles around it even when the position of the own vehicle deviates from the center of the visual field of the camera. The purpose is to be able to.

[0006]

In order to achieve the above object, the vehicle surrounding display device according to claim 1 transmits an image photographed from above by the photographing means to the receiving means of the own vehicle through the transmitting means. Then, in the vehicle surroundings display device for displaying on the monitor, the first image storage means for storing the image taken in the state where the own vehicle does not exist and the second image storage means for storing the image taken in the state where the own vehicle exists. An image storage unit, a vehicle region extraction unit that extracts a vehicle region from a captured image stored in the first image storage unit, information about a vehicle model, and information about an installation state of the imaging unit are stored in advance as parameters. Parameter storage means, vehicle coordinate conversion means for converting the extracted vehicle region into an image viewed from above the vehicle based on each parameter of the parameter storage means, and the vehicle Background area creating means for complementing the extracted vehicle area excluding the vehicle area converted by the coordinate converting means by using the image stored in the first image storing means, the image by the vehicle coordinate converting means, and the A vehicle area synthesizing means for synthesizing an image by the background area creating means, an image by the vehicle area synthesizing means, and an entire surface for synthesizing the image stored in the second image storage means with the vehicle area removed. And an image synthesizing means.

[0007]

A vehicle surroundings display device according to the present invention (claim 1)
Extracts the own vehicle from the captured image, combines the image that has been subjected to image distortion correction (coordinate conversion) according to the position and orientation of the own vehicle, and another area image that does not include the own vehicle. By displaying it, obstacles and parking areas hidden by the distortion of the vehicle are displayed accurately.

[0008]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT An embodiment of a vehicle surrounding display device according to the present invention will be described below with reference to the accompanying drawings. FIG. 1 is a block diagram showing a system configuration of a vehicle surrounding display device according to the present invention. In the figure, 101 is an image input device (hereinafter referred to as a camera) for inputting an image of a region to be monitored using a TV camera or the like. : 102 is an image capturing unit in the claims), 102 is an image memory (A) for storing an image input by the camera 101 (first image storing unit in the claims), and 103 is a captured image A vehicle detection device for detecting the presence / absence of a vehicle in 104, 104 is a vehicle extraction device for extracting a vehicle region from a captured image (which is a vehicle region extraction means in the claims).

Reference numeral 105 denotes a vehicle coordinate conversion device for converting an image of a vehicle region photographed based on an image extracted by the vehicle extraction device 104 and a parameter stored in advance into an image viewed from above (vehicle coordinates in claims). And 106 is an image memory (A) 102 based on inputs from the image memory (A) 102 and the vehicle detection device 103.
Image memory (B) (which is the second image storage means in the claims) for storing the image in FIG. 1 or for continuously storing the stored image, 107 is the vehicle extraction device 104 and the image memory ( B) A background area extracting device (which is a background area creating unit in the claims) that selectively extracts a necessary area based on an input from 106.

Reference numeral 108 denotes a vehicle area synthesizing device (which is a vehicle area synthesizing means in the claims) for synthesizing the vehicle areas after the conversion based on inputs from the vehicle coordinate transformation device 105 and the background region extraction device 107. 109
Is a full-screen synthesizing device for synthesizing and creating a full-screen including the converted vehicle based on inputs from the image memory (A) 102 and the vehicle-region synthesizing device 108 (a full-screen image synthesizing means in claims). ), 110 is an image memory (A)
An output selection device that selects whether to output a composite image or an input image based on inputs from the vehicle 102, the vehicle detection device 103, and the full-screen composition device 109, and 111 transmits the output of the output selection device 110 A device (which is a transmitting means in the claims), 112 is a receiving device (which is a receiving means in the claims) that receives the output of the transmitting device 111, and 113 is a display device that displays a received image of the receiving device 112. Is.

Reference numeral 114 denotes a parameter storage device (a parameter storage means in claims) for storing a vehicle model indicating a predetermined vehicle height, installation of the camera 101, imaging parameters, and the like. Is.

Next, the operation will be described. FIG. 2 is a flow chart showing an example of the processing operation of the vehicle surroundings display device according to the present invention. In the figure, when this processing is started, initialization processing is executed (S201). In this initial setting, for example, a camera 101 mounted downwardly above the parking lot is used to input a captured image of the parking lot into the image memory (A) 102. Further, in the initial setting, an image is captured without a moving object in the imaging range.

Further, the image input to the image memory (A) 102 is output to the image memory (B) 106. FIG. 8 shows a part of the image processed in step S201. Next, after the above initial setting, an image input process for inputting one image is executed (S202). FIG. 9 shows an example of this image input processing, in which a new image is stored in the image memory (A) 10
2 is input at intervals of 33 msec, for example. The size of the image in this case is, for example, horizontal (HN) 5
12 × vertical (VN) 480 pixels.

Next, after the image input in step S202, difference processing is executed (S203). In this difference processing, in the vehicle detection device 103, the image memory (A) 1
The image f (x, y) in 02 and the image g (x, y) in the image memory (B) 106 are Δf (xi, yj) = | f (xi, yj) -g (x,
y) | is calculated. FIG. 10 is an image showing an example of this difference processing, from the image memory (A) 102 to the image memory (B) 1
06 is subtracted, and the pixel value having a value larger than the set value TH1 is set to 1 and the other values are set to 0.

Further, it is determined whether or not there is a vehicle (S204). Here, by the vehicle detection device 103,
The binarized image is scanned, and clustering (described later) is executed for each region in which 1s are consecutive, and a region in which the pixels included in the cluster correspond to a threshold value TH2 set in advance or more is determined to be a vehicle. When it is determined in step S204 that a vehicle exists, the vehicle extraction device 104 extracts the image area in the image memory (A) 102 of the area determined to be the vehicle in the binarized image (S205).

Next, the vehicle coordinate conversion device 105 executes area coordinate conversion processing (described later) as shown in the flowchart of FIG. 3 (S206). FIG. 15 is a vehicle area image before conversion processing by this coordinate conversion, and FIG. 16 is a vehicle area image after coordinate conversion processing. Here, height information is added to the vehicle area image using the vehicle model (see FIG. 6), and coordinate conversion is performed using the height information to convert the vehicle into an image viewed from directly above.

Further, by the background area extraction device 107,
The current vehicle area on the image memory (B) 106 is extracted and set as the background area (S207). FIG. 17 is a screen showing an example of the composition of this vehicle area, and the image memory (B) 1
The brightness value of the pixel at the coordinates corresponding to the vehicle area in 06 is extracted as it is as a background image. The extracted background image becomes an image including an obstacle hidden in the vehicle.

Then, the vehicle area synthesizing device 108 synthesizes the vehicle image converted in step S206 and the background area in step S207 to create a vehicle area (S208). In this synthesizing method, if the vehicle image is present after conversion with all the coordinate values of the background area, the vehicle coordinates are selected and the coordinate values are used as pixel values. If there is no vehicle image, the background area is selected. This processing screen is shown in Figure 1.
8 shows. Here, the converted vehicle area shown in FIG. 16 and the extracted background image shown in FIG. 17 are combined.
At this time, the brightness value of the same coordinates of the background image is applied to the coordinates of which the brightness values have disappeared due to the conversion of the converted vehicle area. This provides an image that shows the relationship between the vehicle and the obstacle.

Next, the vehicle area combined in step S208 is combined with the image in the image memory (A) 102 by the full-screen combining device 109 (S209). This synthesizing method replaces the pixel value with the pixel value of the image memory (A) 102 at all coordinate positions of the vehicle area. At other coordinate positions, the pixel value of the image memory (A) 102 remains unchanged. FIG. 19 shows an example of a screen resulting from this full screen composition.

That is, the vehicle area combined as shown in FIG. 18 is combined with all the images in the image memory (A) 102. Since the combined vehicle area is the same as the vehicle area of the image memory (A) 102, the vehicle area of the image memory (A) 102 is overwritten and the combined vehicle area remains. As a result, the positional relationship between the vehicle and the obstacle can be displayed (part A) without being hidden by the own vehicle.

Thereafter, the output selection device 110 selects the composite image by the full-screen composition device 109 when the vehicle exists, and the image memory (A) 1 when the vehicle does not exist.
The images on 02 are selected, and the selected images are transmitted by the transmission device 111 to the reception device 112 using radio waves (S210).

On the other hand, when it is determined in step S204 that there is no vehicle, it is further determined whether or not the image memory (B) 106 is updated (S21).
1). Here, when there is a cluster in which the pixels included in the cluster are larger than the set threshold value TH3 (TH3 <TH2) in the images clustered in the image memory (B) 106, it is determined that there is a moving object, and the image memory (B )
106 is not updated.

Therefore, in step S211, when it is determined that the image memory (B) 106 is not updated, the process proceeds to step S210, and when it is determined that the image memory (B) 106 is updated, the image memory (B) 106 is updated. ) After updating 106 (S212),
In step S210, the image is transmitted.

Next, the area coordinate conversion processing executed in step S206 will be described with reference to the flow chart shown in FIG. When this process is started, first, as shown in FIG. 4, the vehicle region 402 extracted from the image center 401 is searched for in the extracted image, and the farthest from the center of the coordinates of the vehicle region. Point (farthest point 4
03) is searched (S301). A Hough transform (see FIG. 12) is performed on the farthest point 403 to calculate the line segment that constitutes the point. The search image of the farthest point 403 by the above processing is shown in FIG.
An example of vehicle edge detection is shown in FIG. In this way, the farthest point 403 is detected around the extracted vehicle image.

Further, the vehicle direction is calculated (S30).
2). Here, the center of gravity of the binarized image area clustered with the vehicle is calculated. The position coordinates (xg, yg) of the center of gravity and the main axis are obtained by the following, and the binarized vehicle direction is calculated. FIG. 13 shows an example of calculation of this vehicle direction. The direction of the vehicle is calculated using the vehicle region, the center of gravity is calculated, the principal axis of inertia passing through the center of gravity is calculated, and the direction of the principal axis is the direction of the vehicle. It is the direction.

That is, when the area extracted by the vehicle area extracting device 104 is V (i, j), the following expression 1 is obtained.

[0027]

[Equation 1]

The position coordinates of the center of gravity (xg, yg)
Becomes xg = m ₁₀ / m ₀₀ , yg = m ₁₀ / m ₀₀ ··· (2).

Further, V (i, j) is translated so that the position coordinates (xg, yg) of the center of gravity become the origin of the coordinate system. Assuming that the translated image is V ₂ (i, j), then V ₂ (i, j) = V (i-xg, j-yg) ... (3).

[0030]

[Equation 2]

From the above equation 2, the equation of the principal axis of inertia is i = j · tan θ ₀ ··· (5). Further, the inverse transformation of the parallel movement performed by the above equation (3) is performed to obtain i + xg = (j + yg) · tan θ ₀ ··· (6), which is a straight line indicating the desired vehicle direction.

After the processing of step S302 is executed, the front and rear edges are detected (S303). Here, as shown in FIG. 14, the intersection of the straight line of the vehicle direction and the line segment obtained by the Hough transform is obtained, and the line segment having this intersection is the line segment indicating the front and rear edges of the vehicle. If there are two such line segments, it is determined that they are front and rear edge lines of the vehicle.
Next, as shown in FIG. 14, detection of a side edge is performed in which a line segment having an intersection with the front and rear edge lines is a side edge (S).
304). In this case, a line segment that intersects the edge line is selected from the line segment group detected by the Hough transform. The line segment is the edge of the bonnet or trunk.

Next, the direction of the vehicle is calculated (S30).
5). That is, the length of the line segment is calculated and the front-back direction of the vehicle is estimated. The apparent length of the line segment differs from the actual length due to the camera position and the height of the vehicle. The coordinates of the starting point and the ending point of the line segment are converted using the model information and the following equation (7) to find the actual length. The vehicle model is stored in advance. This model has information as shown in FIG. 6 (described later). The lengths of the trunk and bonnet in this model information are compared with the calculated lengths of the line segments, and the one with the smaller difference is selected to estimate the vehicle orientation.

The above processing will be described in detail below. Figure 5
Is for explaining the real space conversion from the image in the above. In the image f (x, y), (x1, y
When the height from the point of the coordinates of 1) is t2, the image center 4
If 01 is the origin of the coordinates on the ground, this (x1, y1)
Is the following (7) as the coordinates (x1, y1, t1) on the ground.
Is converted by. The focal length of the camera is f, the depression angle is 90 °, the screen size is vertical a × horizontal b, and the installation height is H. _{That, (a / 2x N · (} x N -x1)) / f = x1 / (H-t1) _{Thus, X = (a / 2x N} · f) · (x N -x) · (H-t1) Y = become _{(b / 2y N · f)} · (y N -y) · (H-t1) ·· (7).

FIG. 6 is an explanatory view showing a vehicle model. This model has a constant width W in the y-axis direction,
The three-dimensional coordinates of the twelve feature points are stored in the parameter storage device 114 in advance as a model.

Next, a vehicle model conversion process is executed (S306). Here, the model is subjected to coordinate conversion using the relational expression shown below from the determined direction of the vehicle, edge lines before and after the vehicle, line segment information connected thereto, and installation parameters of the camera 101, and the extracted vehicle is extracted. Height information is added to each pixel of the image. The conversion process and the addition of height information will be described in detail below.

[Conversion of Vehicle Model Information to Real Space] Two points (x1, y) on the edge lines before and after the vehicle obtained from the image.
1, t1), (x2, y2, t1) are selected and converted into the real space using the above equation (7). Note that x is the X coordinate of the image, y is the Y coordinate of the image, and t1 is the height information added to each pixel by the vehicle model conversion. Then, using the coordinates of the converted two points (x1, y1, t1) and (x2, y2, t1), the inclination θ1 of the vehicle with respect to the x coordinate is obtained. That is, θ1 = − (x2−x1) / (y2−y1) · (8)

Further, the inclination θ2 of the vehicle can be similarly obtained from the two points on the side edge line. By averaging these two slopes and using them, it is possible to reduce the influence of an error in edge line detection.

If the intersection of the edge line before and after the vehicle and the side edge line is (x0, y0, t0), the real space coordinates (x0, y0, t0) of this coordinate correspond to the feature point of the model. Among the feature points of the vehicle model shown in 6,
Either 2, 11, or 12 corresponds. Which of these corresponds corresponds to which side of the edge line before and after the vehicle is the feature point and the direction of the vehicle, and this is used as the conversion reference. Then, by determining the reference point, it is possible to coordinate-convert each feature point in the real space using the inclination θ.

Further, when the height information changes between the characteristic points, auxiliary coordinates for image conversion are calculated between the characteristic points. The method of setting the auxiliary coordinates will be described below. FIG. 7 is an explanatory diagram showing a method of setting the auxiliary coordinates. The height Δt, which is determined in advance that the distortion is acceptable, is set. The coordinates in the real space are calculated by the following equations so that the feature points are interpolated for each height Δt. That is, P
1 (x1, y1, t1) and P2 (x2, y2, t2), the auxiliary coordinate (x3, which changes from P1 by nΔt).
y3, t1 + nΔt) can be obtained by x3 = nΔt / (t2-t1) · (x2-x1) y3 = nΔt / (t2-t1) · (y2-x1) ·· (9).

[Conversion of Real Space Coordinates to Image Coordinates] This conversion is an inverse conversion of the above conversion formula (7) and becomes the following formula. _{x = x N - (2x N} f / a (H-t1) · X y = y N - (2y N f / a (H-t1) · Y ·· (10)

[Addition of Height Information] The coordinates of the area surrounded by the feature points and the auxiliary coordinates converted into image coordinates are height information corresponding to the average value of the height information of the feature points and the auxiliary coordinates. Is added. The converted feature points and auxiliary coordinates (x1, y1, t1), (x2, y2, t1), (x
3, y3, t2) and (x4, y4, t2), the height information of the coordinates contained therein is (t1 + t2) / 2 ... (11).

After executing the process of step S306, the image coordinate conversion process is executed (S307), and the process returns. In step S307, the image real space conversion shown in the equation (7) and the following real space → image conversion equation (12) are used for the image to which the height information of the vehicle area image is added. Perform image coordinate transformation. Note that this real space-to-image conversion is to convert an object with a height into an image viewed from directly above. Real space coordinates (X1, Y1,
In t1), t1 = 0 is set, and the above step S30 is performed.
This is the conversion of the conversion of 6. Therefore, from equation (10), x = x _N - a _{(2y N f / aH) ·} Y ·· (12) - (2x N f / aH) · X y = y N.

Next, clustering will be described.
The clustering in this invention is performed in step S203.
Pixels having a pixel value of 1 in the binary image obtained by the difference processing (see FIG. 2) are divided into adjacent groups and labeled (group names). This processing procedure is shown in FIG.
This will be described with reference to the examples shown in FIGS.

[Procedure 1] Pixel value 1 with line number k = 0
Search for images with column numbers from L = 0 to L = 9. In the case of FIG. 20, it is possible to detect when L = 0. The pixel value of the detected pixel is newly set to 1 (label). L
Is incremented by 1 and the same 1 is given as the pixel value when the pixel value of the pixel is 1 (that is, as long as 1 continues). When the pixel value 1 is no longer continuous, the value given to the pixel is incremented by 1. This process is performed up to the maximum of L,
Proceed to the next step 2. Remember the last label given. When this process is performed on the image shown in FIG. 20, the image becomes as shown in FIG.

[Procedure 2] The value of k is increased by 1, and the pixel value is 1
Search for images with L from L = 0 to L = 9. This Figure 2
In the case of 0, it can be detected with L = 1. It is searched whether or not there is a pixel having a pixel value in the k-1th row adjacent to the detected pixel. When there are a plurality of pixels, the pixel value of the pixel having a small L is set as the pixel value of the pixel, and when 1 continues thereafter, this value is newly given as the pixel value. Further, in order to replace the other adjacent labels with the same label, the label to be changed and the label after the change are written in the conversion table. When this process is performed when k = 2, the result is as shown in FIG. 22, and the conversion table shown in FIG. 23 is obtained. This processing is performed up to the maximum of L, and the procedure shifts to step 3.

[Procedure 3] The procedure 2 is repeatedly executed until the maximum value of k. If k is the maximum, go to step 4.

[Procedure 4] The labels of adjacent pixels are made the same label by using the conversion table. When the change target is included after the change, the change target is replaced after the change target is changed. Explaining with the example of the conversion table shown in FIG. 24,
The conversion target 4 is converted to the label 2 after the change, but the label 2
Is also a change target and is converted to label 1. When this is sequentially converted from the top, the label 2 is converted into the label 1, and then the label 4 is converted into the label 2, and the label 2 remains. In order to prevent this, as shown in FIG. 25, the label to be finally converted in the conversion table is replaced after the change.

Further, FIG. 26 shows the label labeled in the example shown in FIG. 20, FIG. 27 shows the conversion table, and FIG. 28 shows the label rearranged using the conversion table. When the image shown in FIG. 26 is produced, labeling of the area where the pixel value 1 is connected is completed, and the number of pixels having the label can be measured using each label.

[0050]

As described above, according to the vehicle surroundings display device (Claim 1) of the present invention, the own vehicle is extracted from the photographed image, and the image is displayed according to the position and direction of the own vehicle. Since the image subjected to the distortion correction (coordinate conversion) and the other region image not including the own vehicle are combined and displayed, it is possible that the position of the own vehicle deviates from the center of the visual field of the camera. Also, it is possible to accurately see the own vehicle and obstacles around it.

[Brief description of drawings]

FIG. 1 is a block diagram showing a system configuration of a vehicle surrounding display device according to the present invention.

FIG. 2 is a flowchart showing an operation process of a vehicle surroundings display device.

FIG. 3 is a flowchart showing a region image conversion process.

FIG. 4 is an explanatory diagram showing an example of an image pickup screen used in the embodiment.

FIG. 5 is an explanatory diagram showing image-to-real space conversion used in the embodiment.

FIG. 6 is an explanatory diagram showing an example of a vehicle model used in the embodiment.

FIG. 7 is an explanatory diagram showing an example of auxiliary coordinates used in the embodiment.

FIG. 8 is an explanatory diagram showing an example of an image in an image memory (B) in an initial setting process.

FIG. 9 is an explanatory diagram showing an example of a screen in an image memory (A) in an image input process.

FIG. 10 is an image example showing difference processing.

FIG. 11 is an image example showing the farthest point search.

FIG. 12 is an image example showing detection of a vehicle edge.

FIG. 13 is an image example showing calculation of a vehicle direction.

FIG. 14 is an image example showing detection of front and rear and side edges.

FIG. 15 is an image example showing before image coordinate conversion.

FIG. 16 is an image example showing an image after coordinate conversion.

FIG. 17 is an image example showing background extraction.

FIG. 18 is an image example showing composition of vehicle regions.

FIG. 19 is an image example showing all-image combination.

FIG. 20 is a matrix showing an example of a binary image related to clustering.

FIG. 21 is a matrix showing an example in which procedure 1 relating to clustering is executed.

FIG. 22 is a matrix showing an example in which procedure 2 relating to clustering is executed.

FIG. 23 is a conversion table example (1) related to clustering.
FIG.

FIG. 24 is a conversion table example (2) related to clustering.
FIG.

FIG. 25 is a conversion table example (3) related to clustering.
FIG.

FIG. 26 is a matrix showing an example of labeling related to clustering.

FIG. 27 is a conversion table example (4) related to clustering.
FIG.

FIG. 28 is a matrix showing an image in which labels are converted by clustering.

FIG. 29 is an explanatory diagram showing distortion of a captured image due to a change in a capturing position of a camera and a three-dimensional object.

[Explanation of symbols]

 Reference Signs List 101 image input device 102 image memory (A) 104 vehicle extraction device 105 vehicle coordinate conversion device 106 image memory (B) 108 vehicle area composition device 109 full screen composition device 111 transmission device 112 reception device 113 display device 114 parameter storage device

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁶ Identification code Internal reference number FI technical display location H04N 7/18 DJ

Claims (1)

[Claims] 1. A vehicle periphery display device for transmitting an image photographed from above by a photographing means to a receiving means of a vehicle via a transmitting means and displaying the image on a monitor, and storing an image photographed in the absence of the vehicle. A first image storage means, a second image storage means for storing an image captured in the state where the vehicle is present, and a vehicle for extracting a vehicle area from the captured image stored in the first image storage means. Area extraction means, parameter storage means for storing in advance information regarding the vehicle model and information regarding the installation state of the photographing means as parameters, and the extracted vehicle area based on each parameter of the parameter storage means. The vehicle coordinate conversion means for converting into a visible image, and the extracted vehicle area excluding the vehicle area converted by the vehicle coordinate conversion means. By the background area creating means for complementing using the image stored in the image storing means, the vehicle area synthesizing means for synthesizing the image by the vehicle coordinate converting means and the image by the background area creating means, and the vehicle area synthesizing means. A vehicle surrounding display device comprising: an image and a full-screen image synthesizing unit for synthesizing the image stored in the second image storage unit and the image excluding the vehicle region.