VR test system and VR test method

Abstract

The present invention assigns a discretionary parallax to left and right images being captured by a stereo camera. Provided is a VR test system of a stereo camera 202, the VR test system comprising a driving environment simulation device 10 for outputting video of a sensor model as mutually independent left and right images, a display device 201 for displaying each of the left and right images, a stereo camera 202 installed in a darkroom together with the display device 201, and a conversion lens 203 disposed between the display device 201 and the stereo camera 202. The stereo camera 202 has left and right cameras for capturing images through the conversion lens 203 independently of left and right, the left image from among the left and right images displayed on the display device 201 being captured by the left camera, and the right image being captured by the right camera. The driving environment simulation device 10 is provided with an image processing unit 101. The image processing unit 101 generates mutually independent left and right images so as to have an equivalent angle of view.

Description

VR test system and VR test method

[0001] The present invention relates to a VR test system and a VR test method.

[0002] Conventionally, evaluation of autonomous vehicles has been carried out in a bench test machine by using a display screen of various driving environments displayed by a simulation device as the subject of photography by a camera of the autonomous vehicle.

[0003] Patent Document 1, an example of the prior art, discloses a vehicle inspection system that inspects a vehicle that performs driving control based on information about the external environment in a predetermined direction detected by two monocular cameras. Patent Document 2, also an example of the prior art, discloses an image processing device in an on-board stereo camera image processing device having a pair of image sensors, in which a parallax image generating unit generates a parallax image based on the difference between the pair of captured images, and performs diagnosis by comparing a diagnostic parallax image with a preset reference parallax image. In Patent Document 2, processing is performed on the measurement side, not on the display side.

[0004] International Publication No. 2020 / 059472 Patent No. 6742423

[0005] Since a stereo camera provides binocular stereoscopic vision, the parallax between the left and right images is important. However, the above-mentioned conventional technology cannot display left and right images with any parallax.

[0006] The present invention has been made in view of the above, and has as its object to provide an arbitrary parallax between left and right images captured by a stereo camera.

[0007] One aspect of the present invention that solves the above-mentioned problems and achieves the object is a stereo camera VR test system that includes a driving environment simulation device that outputs images of a sensor model as left and right images that are independent of each other, a display device that displays each of the left and right images, a stereo camera installed in a darkroom together with the display device, and a conversion lens arranged between the display device and the stereo camera, wherein the stereo camera has left and right cameras that capture the left image and the right image of the left and right images displayed on the display device independently through the conversion lens, and the driving environment simulation device includes an image processing unit that generates the left and right images that are independent of each other so that they have an equivalent angle of view.

[0008] In one aspect of the present invention, a virtual angle of view is used in the sensor model, so that the left and right images have equivalent parallax.

[0009] In one aspect of the present invention, the left and right images are generated so as to have an arbitrary equivalent parallax based on image coordinates on the display device.

[0010] One aspect of the present invention is a stereo camera VR test system comprising: a driving environment simulation device that outputs images of a sensor model as left and right images that are independent of each other; a display device that displays each of the left and right images; a stereo camera installed in a darkroom together with the display device; and a conversion lens arranged between the display device and the stereo camera, wherein the stereo camera has left and right cameras that capture the left image and the right image of the left and right images displayed on the display device independently through the conversion lens; the driving environment simulation device comprises an image processing unit that generates the left and right images that are independent of each other so that they have an equivalent angle of view; and an inverse distortion correction function derivation device that calculates an inverse distortion correction function that corrects distortion of the left and right images caused by the conversion lens.

[0011] One aspect of the present invention is a stereo camera VR test system comprising: a driving environment simulation device that outputs images of a sensor model as left and right images that are independent of each other; a display device that displays each of the left and right images; a stereo camera installed in a darkroom together with the display device; and a conversion lens arranged between the display device and the stereo camera, wherein the stereo camera has left and right cameras that capture the left image and the right image of the left and right images displayed on the display device independently through the conversion lens; the driving environment simulation device comprises an image processing unit that generates the left and right images that are independent of each other so that they have an equivalent angle of view; and an inverse distortion correction function derivation device that calculates an inverse distortion correction function that corrects distortion of the left and right images caused by the conversion lens, and the driving environment simulation device corrects the images of the sensor model using the calculated inverse distortion correction function.

[0012] Alternatively, one aspect of the present invention that solves the above-mentioned problems and achieves the object is a VR test method for a stereo camera that includes simulating a driving environment by outputting images of a sensor model as left and right images that are independent of each other, displaying each of the left and right images, and capturing the left image of the displayed left and right images using a left camera of a stereo camera and a right image of the displayed left and right images through conversion lenses independently for the left and right, wherein in the simulation of the driving environment, the left and right images that are independent of each other are generated so as to have equivalent angles of view.

[0013] One aspect of the present invention is a VR test method for a stereo camera, which includes simulating a driving environment by outputting images of a sensor model as left and right images that are independent of each other, displaying each of the left and right images, capturing the left image of the displayed left and right images using a left camera of a stereo camera and a right image of the displayed left and right images through conversion lenses independently for the left and right, and calculating an inverse distortion correction function that corrects distortion of the left and right images caused by the conversion lenses, wherein in the simulation of the driving environment, the left and right images that are independent of each other are generated so as to have equivalent angles of view.

[0014] One aspect of the present invention is a VR test method for a stereo camera, which includes simulating a driving environment by outputting images of a sensor model as left and right images that are independent of each other; displaying each of the left and right images; capturing the left image of the displayed left and right images using a left camera of a stereo camera and a right image of the displayed left and right images independently through conversion lenses; calculating an inverse distortion correction function that corrects distortion of the left and right images caused by the conversion lenses; and correcting the images of the sensor model using the calculated inverse distortion correction function, wherein in the simulation of the driving environment, the left and right images that are independent of each other are generated so as to have equivalent angles of view.

[0015] According to the present invention, it is possible to provide an arbitrary parallax between left and right images captured by a stereo camera. Furthermore, according to the present invention, it is possible to capture images with reduced distortion caused by a conversion lens.

[0016] FIG. 1 is a diagram showing a VR (virtual reality) test system according to a first embodiment. FIG. 2 is a diagram comparing a grid image captured by a stereo camera without a conversion lens with a grid image captured through a conversion lens. FIG. 3 is a diagram showing left and right images displayed on a display device. FIG. 4 is a diagram showing a VR (virtual reality) test system according to a fourth embodiment. FIG. 5 is a flowchart showing a method for calculating an inverse distortion correction function for correcting distortion caused by a conversion lens, using an inverse distortion correction function derivation device according to the fourth embodiment. FIG. 6 is a diagram showing a VR (virtual reality) test system according to a fifth embodiment. FIG. 7 is a flowchart showing a method for correcting image distortion according to the fifth embodiment.

[0017] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described below with reference to the accompanying drawings. However, the present invention should not be construed as being limited to the following embodiments.

[0018] 1 is a diagram showing a VR (virtual reality) test system 1 in embodiment 1. The VR test system 1 shown in FIG.

[0019] The driving environment simulation device 10 includes simulation software having a scenario, a vehicle model, a driver model, a sensor model, etc., and outputs images of the sensor model as left and right images (left and right images) to a display device 201 of a sensor simulation system 20. Here, the simulator used to generate the left and right images may be simulation software having a sensor model (camera model). The driving environment simulation device 10 also includes an image processing unit 101. As will be described later, the image processing unit 101 performs processing to impart an arbitrary parallax to the left and right images captured by the stereo camera. The left and right images output from the driving environment simulation device 10 to the display device 201 are left and right images to which an arbitrary parallax has been imparted by the image processing unit 101.

[0020] The sensor simulation system 20 includes a darkroom 200 , a display device 201 , a stereo camera 202 , and a conversion lens 203 .

[0021] The darkroom 200 is a box-shaped structure that is provided to prevent surrounding structures and other disturbances from being reflected on the display device 201. Assuming that the stereo camera 202 to be evaluated is mounted on a vehicle, it is necessary to consider the possible size and distance at which the display device 201, which is a display, can be placed. Since such tests are often performed indoors, a compact installation is required. Furthermore, in such tests, reflections due to disturbances should be eliminated. Therefore, the display device 201, the stereo camera 202, and the conversion lens 203 are installed in the darkroom 200.

[0022] The display device 201 displays left and right images with an arbitrary parallax, which are images of the sensor model output from the driving environment simulation device 10. An example of the display device 201 is a 4K display. Here, the left and right images displayed on the display device 201 include left and right images that are independent of each other.

[0023] The stereo camera 202 has a left camera 202L and a right camera 202R, and captures left and right images independently. Here, the left camera 202L captures the left image of the left and right images, and the right camera 202R captures the right image of the left and right images. The stereo camera 202 is controlled from outside the sensor simulation system 20, and sends the captured images to outside the sensor simulation system 20.

[0024] The driving environment simulation device 10 outputs images processed by the image processing unit 101 to the display device 201. The left and right images output from the driving environment simulation device 10 to the display device 201 are given an arbitrary parallax and are independent of each other. The stereo camera 202 captures the left and right images through a conversion lens 203 displayed on the display device 201.

[0025] The conversion lens 203 can be placed and adjusted at any position between the display device 201 and the stereo camera 202, and is placed so that the focal length of the stereo camera 202 matches the distance to the display device 201. Figure 2 is a diagram comparing a grid image captured by the stereo camera without using the conversion lens and a grid image captured through the conversion lens. As can be seen from Figure 2, passing the image through the conversion lens makes the image clearer. This makes it possible to provide any parallax without blurring the image.

[0026] The following describes the arrangement and generation of left and right images on the screen of the display device 201 in order to provide accurate parallax to the stereo camera 202. The range of the left and right angle of view of the display screen of the display device 201 is calculated and specified from the specifications of the stereo camera 202 (baseline length, angle of view, resolution, etc.) and the distance between the stereo camera 202 and the display device 201. The left and right images displayed on the screen of the display device 201 are calculated and generated so as to avoid the intersection range of the left and right angle of view.

[0027] 3 is a diagram showing left and right images displayed on the display device 201. The following describes the coordinate calculations required to display independent left and right images on the display device 201 as shown in FIG. 3. Examples of parameters related to the display device 201 and the stereo camera 202 are shown in Table 1.

[0028]

[0029] Here, as shown in Fig. 3, the origins, y-axis, and z-axis of the display device 201 and stereo camera 202 are aligned, and the stereo camera 202 is positioned directly opposite the display device 201. The parameters used in the coordinate calculation are also shown in Fig. 3. Here, the shooting ranges of the left and right cameras of the stereo camera 202 are calculated from the relationship between the display device 201 and the stereo camera 202. The position, size, and angle of view of the image to be displayed are determined as follows.

[0030]

[0031]

[0032] Here, the parameters in the above formulas (1) and (2) are as follows: H: width of the imaging range; L: distance between the display device 201 and the stereo camera 202; θ: angle of view of the field of view; V: height of the imaging range; P Ｖ : Number of vertical pixels (Vertical pixels) P Ｈ : Number of horizontal pixels

[0033] The width H of the imaging range is calculated from the distance L from the stereo camera 202 to the display device 201 and the angle of view θ of the field of view using the above formula (1). Ｖ / P Ｈ ) is calculated using the above formula (2). Next, the width D of the generated image is calculated. ｗｉｄｔｈ and the height D of the generated image ｈｅｉｇｈｔ is calculated from these values ​​using the following equations (3), (4), and (5).

[0034]

[0035]

[0036]

[0037] Here, the parameters in the above equations (3), (4), and (5) are as follows: D １ : Distance from the origin to the right image D ２ : Distance from the origin to the left image B: Distance between the left and right cameras (baseline length of the stereo camera)

[0038] By taking into consideration the range where the left and right shooting ranges intersect using the above formula (3), the sizes of the left and right images to be generated are calculated using formulas (4) and (5). At this time, as described above, the origins, y-axis, and z-axis of the display device 201 and the stereo camera 202 coincide with each other, so once the sizes of the left and right images are determined, the coordinates and the number of pixels on the display device 201 are determined.

[0039] According to the present embodiment described above, any desired parallax can be given to the left and right images captured by the stereo camera.

[0040] Second Embodiment In the present embodiment, a virtual angle of view θ , which is a virtual FOV (Virtual Fields of View), is calculated based on the results of the first embodiment. ＣＧ is calculated, and the virtual angle of view θ is set to the sensor model (camera model) of the driving environment simulation device 10. ＣＧ In this embodiment, in the driving environment simulation device 10, a virtual angle of view of a sensor model (camera model) to be output to a display device 201, which is a display, is calculated using the results of the first embodiment, and left and right images having equivalent parallax are generated using the virtual angle of view derived from the sensor model (camera model) of the driving environment simulation device 10, and are output to the display device 201, which is a display.

[0041] In order to have the left and right cameras of the stereo camera 202 capture the left and right images on the display device 201 independently, it is necessary to avoid the area where the left and right capture ranges intersect. Therefore, the image to be displayed becomes small as shown in FIG. 3, and the virtual angle of view θ of the sensor model (camera model) of the following equation (6) is ＣＧ is used to generate the image.

[0042]

[0043] where atan is the inverse trigonometric function.

[0044] According to this embodiment, the virtual angle of view θ ＣＧ By using the above, the accuracy of object recognition and distance measurement can be improved compared to that of the first embodiment.

[0045] (Embodiment 3) In this embodiment, a description will be given of a form in which left and right images having any equivalent parallax are generated from image coordinates on the display device 201, which is a monitor, in the sensor model (camera model) of the driving environment simulation device 10. The parallax D to be captured is calculated from the desired capture distance Z by the following equation (7).

[0046]

[0047] Here, the parameters in the above formula (7) are as follows: BF: product of base length (B) and focal length (F); D∞: parallax at infinity;

[0048] In the above formula (7), BF and D∞ are parameters specific to the stereo camera. The disparity D to be captured from the desired capture distance Z is calculated. At this time, the disparity between the left and right images captured by the stereo camera 202 is calculated using SAD (Sum of Absolute Difference) and is calculated based on the upper left y0, z0 of each angle of view. Because the left and right images to be displayed are each at an equal distance from the reference point of the angle of view, the stereo camera 202 can capture images without adjusting the disparity D calculated by the above formula (7).

[0049] By adjusting the equivalent parallax, it becomes possible to adjust (match) the driving environment simulation device 10 and the actual stereo camera 202, thereby improving the accuracy of object recognition and distance measurement compared to the first embodiment.

[0050] (Embodiment 4) In this embodiment, a configuration will be described in which an inverse distortion correction function derivation device is used as a correction means when distortion occurs in an image due to a conversion lens in Embodiments 1 to 3. Note that in Embodiment 4, components common to Embodiments 1 to 3 are denoted by the same reference numerals in the drawings, and descriptions thereof will be omitted.

[0051] 4 is a diagram showing a VR (virtual reality) test system according to embodiment 4. The VR test system 300 shown in FIG. 4 includes a driving environment simulation device 30, a sensor simulation system 20, and an inverse distortion correction function derivation device 304.

[0052] The driving environment simulation device 30 includes simulation software having a scenario, a vehicle model, a driver model, a sensor model, etc., and outputs an image of the sensor model as left and right images (left image and right image) to the display device 201 of the sensor simulation system 20. Furthermore, the driving environment simulation device 30 outputs calibration images (such as a grid image or a checkerboard image) as left and right images (left image and right image) to the display device 201 of the sensor simulation system 20. Furthermore, the driving environment simulation device 30 corrects the calibration images based on the inverse distortion correction function transmitted from the inverse distortion correction function derivation device 304, and outputs the corrected calibration images as left and right images (left image and right image) to the display device 201 of the sensor simulation system 20. The driving environment simulation device 30 also includes an image processing unit 101.

[0053] The sensor simulation system 20 includes a darkroom 200 , a display device 201 , a stereo camera 202 , and a conversion lens 203 .

[0054] While the use of the conversion lens 203 allows the image to be focused on the display device 201, it may cause distortion in the captured image, which may affect parallax. To address this problem, the inverse distortion correction function derivation device 304 calculates an inverse distortion correction function that corrects the image distortion caused by the conversion lens 203.

[0055] Specifically, inverse distortion correction function derivation device 304 determines the distortion function of each grid from an undistorted calibration image (such as a grid image or a checkerboard image) and a distorted calibration image captured by stereo camera 202 through conversion lens 203, and generates an inverse distortion correction function as the inverse function. Furthermore, inverse distortion correction function derivation device 304 again determines the distortion function of each grid from the undistorted calibration image and a corrected calibration image captured by stereo camera 202 through conversion lens 203, which is a calibration image corrected by driving environment simulation device 30 based on the generated inverse distortion correction function, and generates an inverse distortion correction function as the inverse function. These steps are repeated, and when inverse distortion correction function derivation device 304 determines that there is no distortion in the corrected calibration image, the most recently generated inverse distortion correction function becomes the optimized inverse distortion correction function.

[0056] (Method of Calculating Inverse Distortion Correction Function) Next, a method of calculating an inverse distortion correction function for correcting distortion caused by the conversion lens 203 using the inverse distortion correction function derivation device 304 according to this embodiment will be described. Fig. 5 is a flowchart showing a method of calculating an inverse distortion correction function for correcting distortion caused by the conversion lens 203 using the inverse distortion correction function derivation device 304 according to the fourth embodiment.

[0057] First, the image processing unit 101 creates independent left and right images by applying an arbitrary parallax to a calibration image such as a grid image or a checkerboard image, and displays the left and right images on the display device 201 (step S101).

[0058] Next, the stereo camera 202 captures the left and right images displayed on the display device 201 through the conversion lens 203 (step S102).

[0059] Next, the inverse distortion correction function derivation device 304 compares the original calibration image with the calibration image captured by the stereo camera 202 through the conversion lens 203 to determine whether the image is distorted (step S103). Many specific determination methods are possible, but for example, if the original calibration image is a lattice image, it can be determined that the image is free of distortion if all of the intersections of the original lattice image and the intersections of the captured lattice image match. In this case, for example, the original calibration image is stored in the inverse distortion correction function derivation device 304 in advance, or the original calibration image is transmitted from the driving environment simulation device 30.

[0060] If it is determined that the corrected calibration image contains distortion (step S103: No), the inverse distortion correction function derivation device 304 calculates the distortion function of each lattice from the original calibration image and the distorted calibration image captured by the stereo camera 202 through the conversion lens 203, and generates an inverse distortion correction function as the inverse function of the calculated distortion function (step S104). The generated inverse distortion correction function is transmitted to the driving environment simulation device 30.

[0061] Next, the driving environment simulation device 30 corrects the calibration image based on the inverse distortion correction function, and gives an arbitrary parallax to the corrected calibration image to create left and right images that are independent of each other, and displays the left and right images on the display device 201 (step S105).

[0062] Returning to step S102 again, the stereo camera 202 captures the left and right images displayed on the display device 201 through the conversion lens 203, and in step S103, the inverse distortion correction function derivation device 304 compares the original calibration image with the corrected calibration image captured by the stereo camera 202 through the conversion lens 203, and determines whether there is any distortion in the image.

[0063] If it is determined that there is no distortion in the corrected calibration image (step S103: Yes), the most recently generated inverse distortion correction function becomes the optimized inverse distortion correction function, and calculation of the inverse distortion correction function is completed (step S106).

[0064] In this way, by capturing a calibration image, calculating an inverse distortion function from the image obtained, and capturing a corrected image to which the inverse distortion function has been applied through a conversion lens, it is possible to capture images with reduced distortion caused by the conversion lens without having to adjust (correct) the stereo camera body.

[0065] (Embodiment 5) In this embodiment, a display device displays left and right images with an arbitrary parallax, using an image of a sensor model corrected using the inverse distortion correction function calculated in Embodiment 4. Note that in Embodiment 5, components common to Embodiment 4 are denoted by the same reference numerals in the drawings, and descriptions thereof will be omitted.

[0066] 6 is a diagram showing a VR (virtual reality) test system according to embodiment 5. The VR test system 400 shown in FIG. 6 includes a driving environment simulation device 40, a sensor simulation system 20, and an inverse distortion correction function derivation device 304.

[0067] The driving environment simulation device 40 includes simulation software having a scenario, a vehicle model, a driver model, a sensor model, etc., and outputs an image of the sensor model as left and right images (left image and right image) to the display device 201 of the sensor simulation system 20. Furthermore, the driving environment simulation device 40 outputs calibration images (such as a grid image or a checkerboard image) as left and right images (left image and right image) to the display device 201 of the sensor simulation system 20. Furthermore, the driving environment simulation device 40 corrects the calibration images based on the inverse distortion correction function transmitted from the inverse distortion correction function derivation device 304, and outputs the corrected calibration images as left and right images (left image and right image) to the display device 201 of the sensor simulation system 20.

[0068] Furthermore, the driving environment simulation device 40 corrects the image of the sensor model using the inverse distortion correction function calculated by the inverse distortion correction function derivation device 304, and outputs the corrected image as left and right images (left image and right image) to the display device 201 of the sensor simulation system 20. The driving environment simulation device 40 also includes an image processing unit 101.

[0069] The sensor simulation system 20 includes a darkroom 200 , a display device 201 , a stereo camera 202 , and a conversion camera 203 .

[0070] The inverse distortion correction function derivation device 304 calculates an inverse distortion correction function that corrects the image distortion caused by the conversion lens 203 .

[0071] (Image Distortion Correction Method) Next, a method for correcting image distortion using the calculated inverse distortion correction function according to this embodiment will be described. The method for correcting image distortion comprises (1) calculating an inverse distortion correction function, and (2) outputting a corrected image using the calculated inverse distortion correction function. Fig. 7 is a flowchart showing the method for correcting image distortion according to embodiment 5.

[0072] (1) Calculation of the inverse distortion correction function First, the image processing unit 101 applies an arbitrary parallax to a calibration image such as a grid image or a checkerboard image, creates left and right images that are independent of each other, and displays the left and right images on the display device 201 (step S201).

[0073] Next, the stereo camera 202 captures the left and right images displayed on the display device 201 through the conversion lens 203 (step S202).

[0074] Next, the inverse distortion correction function derivation device 304 compares the original calibration image with the calibration image captured by the stereo camera 202 through the conversion lens 203 to determine whether the image is distorted (step S203). Many specific determination methods are possible, but for example, if the original calibration image is a lattice image, it can be determined that the image is free of distortion if all of the intersections of the original lattice image and the intersections of the captured lattice image match. In this case, for example, the original calibration image is stored in the inverse distortion correction function derivation device 304 in advance, or the original calibration image is transmitted from the driving environment simulation device 40.

[0075] If it is determined that the corrected calibration image contains distortion (step S203: No), the inverse distortion correction function derivation device 304 calculates the distortion function of each lattice from the original calibration image and the distorted calibration image captured by the stereo camera 202 through the conversion lens 203, and generates an inverse distortion correction function as the inverse function of the calculated distortion function (step S204). The generated inverse distortion correction function is transmitted to the driving environment simulation device 30.

[0076] Next, the driving environment simulation device 40 corrects the calibration image based on the inverse distortion correction function, and an arbitrary parallax is applied to the corrected calibration image to create left and right images that are independent of each other, and displays the left and right images on the display device 201 (step S205).

[0077] Returning to step S202 again, the stereo camera 202 captures the left and right images displayed on the display device 201 through the conversion lens 203, and in step S203, the inverse distortion correction function derivation device 304 compares the original calibration image with the corrected calibration image captured by the stereo camera 202 through the conversion lens 203, and determines whether there is any distortion in the image.

[0078] If it is determined that there is no distortion in the corrected calibration image (step S203: Yes), the most recently generated inverse distortion correction function becomes the optimized inverse distortion correction function, and calculation of the inverse distortion correction function is completed (step S206). Note that steps S201 to S206 are the same as steps S101 to S106 of the method of calculating the inverse distortion correction function described in the fourth embodiment.

[0079] (2) Output of corrected image using the calculated inverse distortion correction function Next, the driving environment simulation device 40 incorporates the inverse distortion correction function calculated by the inverse distortion correction function derivation device 304 into the device model, and creates an image of the corrected sensor model (step S207).

[0080] Next, the image processing unit 101 provides an arbitrary parallax to the corrected image of the sensor model, creates left and right images that are independent of each other, and displays the left and right images on the display device 201 (step S208).

[0081] Next, the stereo camera 202 captures the left and right images displayed on the display device 201 through the conversion lens 203 (step S209). The images captured by the stereo camera 202 are then sent to the outside of the sensor simulation system 20.

[0082] In this way, while using a conversion lens brings the image into focus on the display device, distortion occurs in the captured image, affecting parallax. However, by capturing a calibration image, calculating an inverse distortion function from the image obtained, and then capturing a corrected image in which the inverse distortion function has been incorporated into the model of the virtual driving simulation device through the conversion lens, it is possible to suppress distortion of the image from the stereo camera viewpoint of the virtual driving simulation device without having to adjust (correct) the stereo camera itself.

[0083] In the first to fifth embodiments, the left and right images are displayed on the same display screen, but the present invention is not limited to this. The display screen on which the left image is displayed and the display screen on which the right image is displayed may be different.

[0084] Although the VR test system has been described in the first to fifth embodiments, the present invention is not limited thereto. That is, the present invention also includes a VR test method for a stereo camera, which includes simulating a driving environment by outputting images of a sensor model as left and right images independent of each other, displaying each of the left and right images, photographing each of the displayed left and right images with left and right cameras of a stereo camera, and performing object recognition and distance measurement for the displayed left and right images, and in the simulation of the driving environment, the left and right images independent of each other are generated so as to have equivalent angles of view.

[0085] Although the fourth and fifth embodiments have been described with reference to the case where the inverse distortion correction function derivation device exists independently, the present invention is not limited to this. The functions of the inverse distortion correction function derivation device may be provided in the driving environment simulation device or the stereo camera.

[0086] The present invention is not limited to the above-described embodiment, but also includes various modifications in which components are added, deleted, or converted from the above-described configuration.

[0087] 1, 300, 400 VR test system 10, 30, 40 Driving environment simulation device 101 Image processing unit 20 Sensor simulation system 200 Darkroom 201 Display device 202 Stereo camera 202L Left camera 202R Right camera 203 Conversion lens 304 Inverse distortion correction function derivation device

Claims

1. A stereo camera VR test system comprising: a driving environment simulation device that outputs images of a sensor model as left and right images that are independent of each other; a display device that displays each of the left and right images; a stereo camera installed in a darkroom together with the display device; and a conversion lens arranged between the display device and the stereo camera, wherein the stereo camera has left and right cameras that capture the left image and the right image of the left and right images displayed on the display device through the conversion lens independently for the left and right; the driving environment simulation device comprises an image processing unit, and the image processing unit generates the left and right images that are independent of each other so as to have an equivalent angle of view.

2. The VR test system of claim 1, wherein a virtual angle of view is used in the sensor model, so that the left and right images have equivalent parallax.

3. The VR test system of claim 1, wherein the left and right images are generated to have any equivalent parallax based on image coordinates on the display device.

4. A stereo camera VR test system comprising: a driving environment simulation device that outputs images of a sensor model as left and right images that are independent of each other; a display device that displays each of the left and right images; a stereo camera installed in a darkroom together with the display device; and a conversion lens arranged between the display device and the stereo camera, wherein the stereo camera has left and right cameras that capture the left image and the right image of the left and right images displayed on the display device through the conversion lens independently for the left and right; the driving environment simulation device comprises an image processing unit, wherein the image processing unit generates the left and right images that are independent of each other so as to have an equivalent angle of view; and further comprises an inverse distortion correction function derivation device that calculates an inverse distortion correction function that corrects distortion of the left and right images caused by the conversion lens.

5. The VR test system of claim 4, wherein a virtual angle of view is used in the sensor model, so that the left and right images have equivalent parallax.

6. The VR test system of claim 4, wherein the left and right images are generated so as to have any equivalent parallax based on image coordinates on the display device.

7. A stereo camera VR test system comprising: a driving environment simulation device that outputs images of a sensor model as left and right images that are independent of each other; a display device that displays each of the left and right images; a stereo camera installed in a darkroom together with the display device; and a conversion lens arranged between the display device and the stereo camera, wherein the stereo camera has left and right cameras that capture the left image and the right image of the left and right images displayed on the display device through the conversion lens independently for the left and right; the driving environment simulation device comprises an image processing unit, wherein the image processing unit generates the left and right images that are independent of each other so as to have an equivalent angle of view, and further comprises an inverse distortion correction function derivation device that calculates an inverse distortion correction function that corrects distortion of the left and right images caused by the conversion lens, and the driving environment simulation device corrects the images of the sensor model using the calculated inverse distortion correction function.

8. The VR test system of claim 7, wherein a virtual angle of view is used in the sensor model, so that the left and right images have equivalent parallax.

9. The VR test system of claim 7, wherein the left and right images are generated so as to have any equivalent parallax based on image coordinates on the display device.

10. A VR test method for a stereo camera, comprising: simulating a driving environment by outputting images of a sensor model as mutually independent left and right images; displaying each of the left and right images; and capturing the left image of the displayed left and right images using a left camera of a stereo camera and a right image of the displayed left and right images through conversion lenses independently for the left and right; wherein, in the simulation of the driving environment, the left and right images that are independent from each other are generated so as to have equivalent angles of view.

11. A VR test method for a stereo camera, comprising: simulating a driving environment by outputting images of a sensor model as left and right images that are independent of each other; displaying each of the left and right images; capturing the left image of the displayed left and right images using the left camera of a stereo camera and the right image using the right camera of the stereo camera through a conversion lens independently for the left and right; and calculating an inverse distortion correction function that corrects distortion of the left and right images caused by the conversion lens, wherein in the simulation of the driving environment, the left and right images that are independent of each other are generated so as to have equivalent angles of view.

12. A VR test method for a stereo camera, comprising: simulating a driving environment by outputting images of a sensor model as left and right images that are independent of each other; displaying each of the left and right images; capturing the left image of the displayed left and right images using the left camera of a stereo camera and the right image using the right camera of the stereo camera through a conversion lens independently for the left and right; calculating an inverse distortion correction function that corrects distortion of the left and right images caused by the conversion lens; and correcting the images of the sensor model using the calculated inverse distortion correction function, wherein in the simulation of the driving environment, the left and right images that are independent of each other are generated so as to have an equivalent angle of view.

Images