Calibration pattern and calibration method
The calibration pattern addresses the limitation of conventional patterns by combining large and small indicators, enabling accurate calibration of cameras with varying resolutions and views, enhancing usability and reducing display size for improved convenience.
Patent Information
- Application Number
- PCT/JP2026/018409
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2025-05-16
- Filing Date
- 2026-05-15
- Publication Date
- 2026-11-19
AI Technical Summary
Conventional camera calibration patterns are limited by the specifications of low-resolution cameras, making it impossible to calibrate high-resolution cameras effectively, and there is a need for a common calibration pattern that can accommodate both low- and high-resolution cameras.
A calibration pattern that includes a combination of large- and small-size calibration indicators with a predetermined size ratio, arranged within a unit area to accommodate cameras with different resolutions and fields of view, allowing for accurate calibration by processing images with both low- and high-resolution algorithms.
Enables calibration of multiple cameras with a wider range of resolution and field of view differences, improving accuracy and convenience by allowing the use of a single pattern for both low- and high-resolution cameras, and reducing the size of the display area for easier handling.
Smart Images

Figure JP2026018409_19112026_PF_FP_ABST
Abstract
Description
Calibration pattern and calibration method
[0001] The present disclosure relates to a calibration pattern and a calibration method, and more particularly, to a calibration pattern and a calibration method capable of expanding the combinable range of calibratable cameras and improving convenience.
[0002] Conventionally, as calibration patterns used when obtaining camera calibration parameters for calibrating a camera, patterns such as a chessboard pattern, and a pattern in which circular calibration indices are arranged in a grid, and the like have been used. Further, a calibration pattern in which ArUco markers used for position recognition in AR (Augmented Reality) technology are combined with a chessboard is also known.
[0003] For example, Patent Document 1 discloses a method for calibrating an active sensor and a passive sensor using markers formed on a calibration sheet.
[0004] International Publication No. 2019 / 093136
[0005] By the way, when performing camera calibration using a calibration pattern on which a plurality of calibration indices are arranged, it is necessary to arrange the required number of calibration indices for obtaining camera calibration parameters within the angle of view of the camera at a size that can be recognized by the camera. Furthermore, when calibrating a plurality of cameras with different resolutions and angles of view, the maximum size of the calibration pattern is limited by the specifications of the low-resolution camera.
[0006] Therefore, conventionally, when calibrating a high-resolution camera with a narrow field of view using a calibration pattern of the maximum size specified for a low-resolution camera, it was sometimes impossible to place the necessary number of calibration indicators, which are sized according to the specifications of the low-resolution camera, within the field of view of the high-resolution camera. In this case, it was not possible to calibrate the camera using a common calibration pattern, and the desired camera calibration parameters could not be obtained. Therefore, there was a need to improve convenience by enabling the use of a common calibration pattern for both low-resolution and high-resolution cameras, thereby expanding the combination of cameras that can be calibrated, and consequently reducing the size of the calibration pattern.
[0007] This disclosure is made in light of these circumstances and aims to expand the combination of calibrable cameras and improve convenience.
[0008] A calibration pattern according to one aspect of this disclosure involves arranging one or more markers within a predetermined unit area, arranging at least two types of large-size calibration indicators and small-size calibration indicators of different sizes with a predetermined size ratio or greater within the unit area as calibration indicators used to calculate camera calibration parameters, arranging a predetermined number of the large-size calibration indicators around the markers within the unit area, arranging the small-size calibration indicators to fill the blank areas in the unit area where the markers and large-size calibration indicators are not arranged, or arranging the small-size calibration indicators on top of the large-size calibration indicators, or combining these arrangements to arrange four or more of the small-size calibration indicators within the unit area.
[0009] A calibration method according to one aspect of this disclosure involves arranging one or more markers within a predetermined unit area, arranging at least two types of large-size calibration indicators and small-size calibration indicators of different sizes with a predetermined size ratio or greater within the unit area as calibration indicators used to calculate camera calibration parameters, arranging a predetermined number of the large-size calibration indicators around the markers within the unit area, arranging the small-size calibration indicators to fill the blank areas in the unit area where the markers and large-size calibration indicators are not arranged, or arranging the small-size calibration indicators on top of the large-size calibration indicators, or combining these arrangements to arrange four or more of the small-size calibration indicators within the unit area, and processing an image of the calibration pattern captured by a low-resolution camera according to a low-resolution calibration algorithm, and processing an image of the calibration pattern captured by a high-resolution camera according to a high-resolution calibration algorithm.
[0010] In one aspect of this disclosure, one or more markers are arranged within a predetermined unit area, and within the unit area, at least two types of large-size calibration indicators and small-size calibration indicators of different sizes with a predetermined size ratio or greater are arranged as calibration indicators used to calculate camera calibration parameters, a predetermined number of the large-size calibration indicators are arranged around the markers within the unit area, and the small-size calibration indicators are arranged to fill the blank areas in the unit area where the markers and large-size calibration indicators are not arranged, or the small-size calibration indicators are arranged on top of the large-size calibration indicators, or a combination of these arrangements is used to arrange four or more of the small-size calibration indicators within the unit area.
[0011] This figure shows an example layout of the first embodiment of a calibration pattern to which this technology is applied. This figure shows an example layout of the second embodiment of a calibration pattern to which this technology is applied. This figure illustrates the fourth layout condition of the calibration pattern. This figure illustrates the fourth layout condition of the calibration pattern. This figure illustrates the fourth layout condition of the calibration pattern. This figure illustrates the fourth layout condition of the calibration pattern. This figure illustrates the definition of a calibration pattern. This figure illustrates that it can accommodate the calibration of multiple cameras with a wider difference in resolution. This figure illustrates that it can accommodate the calibration of multiple cameras with a wider difference in field of view. This is a schematic diagram illustrating the effects of using a calibration pattern. This figure illustrates examples of calibration indexes and marker types. This figure shows an example of the first variation of the calibration pattern. This figure shows an example of the second variation of the calibration pattern. This figure shows an example of the third variation of the calibration pattern. This figure shows an example of the fourth variation of the calibration pattern. This figure shows an example of the fifth variation of the calibration pattern. This figure shows an example of the sixth variation of the calibration pattern. This figure shows an example of the seventh variation of the calibration pattern. This figure shows an example of the eighth variation of the calibration pattern. This is a flowchart explaining the calibration process. This figure explains an example of the black paint used in the calibration pattern.
[0012] The following describes in detail a specific embodiment of this technology, with reference to the drawings.
[0013] <First Layout Example of Calibration Pattern> Figure 1 shows a layout example of a first embodiment of a calibration pattern to which this technology is applied.
[0014] As shown in Figure 1, the calibration pattern 11 has a layout in which one marker 22, four large-size calibration indices 23a-1 to 23a-4, and nine small-size calibration indices 23b-1 to 23b-9 are arranged within a display area 21 which is the area for displaying the calibration pattern 11. Hereafter, when it is not necessary to distinguish between large-size calibration indices 23a-1 to 23a-4, they will be referred to as large-size calibration indices 23a, and when it is not necessary to distinguish between small-size calibration indices 23b-1 to 23b-9, they will be referred to as small-size calibration indices 23b. Furthermore, when it is not necessary to distinguish between large-size calibration indices 23a and small-size calibration indices 23b, they will be referred to as calibration indices 23.
[0015] The display area 21 represents the external shape of a board on which the calibration pattern 11 is drawn (printed), for example. For example, when calibrating a camera using the calibration pattern 11, the user holds the board in their hand and takes the necessary number of photos (for example, about 10) to perform the calibration while changing the angle of the calibration pattern 11 relative to a fixed camera. In the example shown in Figure 1, a square shape is used for the display area 21.
[0016] The marker 22 has a uniquely determined pattern that allows it to recognize the up, down, left, and right orientation of the calibration pattern 11. In the example shown in Figure 1, the marker 22 is an ArUco marker with a pattern in which white squares, each 1 / 6 the size of the black square, are arranged inside a black square, and only one is placed in the center of the display area 21.
[0017] The large-size calibration index 23a is a large-sized calibration index used to calculate camera calibration parameters. In the example shown in Figure 1, the large-size calibration index 23a consists of black circles positioned at four locations on the top, bottom, left, and right of the marker 22, and is at least 1.71 times larger than the small-size calibration index 23b.
[0018] The small-size calibration index 23b is a small-sized calibration index used to calculate camera calibration parameters. In the example shown in Figure 1, the small-size calibration index 23b uses a circle that is 1 / 1.71 times or less in size than the large-size calibration index 23a. The small-size calibration indexes 23b-1 to 23b-4 use white circles that are superimposed on the large-size calibration indexes 23a-1 to 23a-4, respectively. The small-size calibration indexes 23b-5 to 23b-9 use black circles that are placed in the blank areas within the display area 21 where the marker 22 and the large-size calibration indexes 23a-1 to 23a-4 are not placed.
[0019] Accordingly, the calibration pattern 11 shown in Figure 1 has a layout in which one large calibration index 23a-1 to one large calibration index 23a-4 are placed above, below, to the left and to the right of the marker 22 located in the central part of the display area 21, one small calibration index 23b-1 to one small calibration index 23b-4 are placed overlapping with the large calibration index 23a-1 to one large calibration index 23a-4, and one small calibration index 23b-5 to one small calibration index 23b-9 are placed in the blank area.
[0020] Such a calibration pattern 11 layout allows for the placement of a larger number of small calibration indicators 23b within the maximum display area 21 corresponding to the low-resolution wide-angle camera field of view 31. For example, within the high-resolution telephoto camera field of view 32, the number of large calibration indicators 23a or small calibration indicators 23b required to determine camera calibration parameters can be placed in sizes corresponding to the specifications of the low-resolution camera.
[0021] This allows calibration to be performed using a common calibration pattern 11 for both low-resolution and high-resolution cameras, thereby improving convenience. In other words, since the calibration pattern 11 can support the calibration of multiple cameras with different resolutions and fields of view, it is possible to perform calibration with a wider range of camera combinations with different resolutions and fields of view using the calibration pattern 11.
[0022] Furthermore, since the calibration pattern 11 has a large number of small calibration indicators 23b arranged within a limited area, it is possible to extract a large number of feature points when calibrating a high-resolution camera, thereby enabling more accurate calibration. In other words, by increasing the number of small calibration indicators 23b arranged within the display area 21, the accuracy of the high-resolution camera calibration can be improved.
[0023] Furthermore, the calibration pattern 11 can improve the accuracy of the calibration of the high-resolution camera by aligning the centroids of the overlapping large-size calibration indicators 23a-1 to 23a-4 and the small-size calibration indicators 23b-1 to 23b-4, thereby averaging their respective detection results.
[0024] Furthermore, by overlapping the small-sized calibration indicators 23b with the large-sized calibration indicators 23a, or by placing the small-sized calibration indicators 23b in a blank area, the calibration pattern 11 can be displayed in a more limited and narrow area, thereby reducing the size of the display area 21. As a result, for example, while conventional calibration patterns were larger than the screens of tablets and personal computers, the calibration pattern 11 can be displayed on the screens of tablets and personal computers due to the reduction in the size of the display area 21. Therefore, the handling of the calibration pattern 11 when calibrating a camera can be made easier, further improving convenience.
[0025] <Second Layout Example of Calibration Pattern> Figure 2 shows a layout example of a second embodiment of a calibration pattern to which this technology is applied. In the calibration pattern 11A shown in Figure 2, the same reference numerals are used for layouts common to the calibration pattern 11 shown in Figure 1, and their detailed explanation is omitted.
[0026] As shown in Figure 2, the calibration pattern 11A has a layout in which two markers 22-1 and marker 22-2, six large-size calibration indicators 23a-1 to 23a-6, and thirty small-size calibration indicators 23b-1 to 23b-30 are arranged within the display area 21A that displays the calibration pattern 11A.
[0027] The display area 21A is a rectangle with an aspect ratio of 1:√2 (the so-called A-size), and in the example shown in Figure 2, the display area 21A is arranged in a vertical orientation.
[0028] Markers 22-1 and 22-2 are arranged vertically in the center of the display area 21A in the left-right direction.
[0029] The large-size calibration indicators 23a-1 to 23a-6 are arranged in a 2x3 grid (width x height).
[0030] For the small-size calibration indicators 23b-1 to 23b-6, white circles are used, which are positioned to overlap with the large-size calibration indicators 23a-1 to 23a-6, respectively. For the small-size calibration indicators 23b-7 to 23b-30, black circles are used, which are positioned to fill the blank areas in the display area 21A where markers 22-1 and marker 22-2 and large-size calibration indicators 23a-1 to 23a-6 are not positioned.
[0031] Calibration patterns 11A with the above layout can be used in common for calibrating both low-resolution and high-resolution cameras, similar to calibration patterns 11 in Figure 1, thereby improving convenience. Furthermore, by using a rectangular calibration pattern 11A as shown in the figure, for example, the usability of the board on which the calibration pattern 11A is drawn can be improved.
[0032] <Layout Conditions for Calibration Pattern> Referring to Figures 3 to 6, the first to fourth layout conditions for the calibration pattern 11 that can achieve the effects described above will be explained. When multiple large-size calibration indicators 23a are arranged around the marker 22, the rectangular area in which the large-size calibration indicators 23a can be arranged at the maximum density (for example, an area corresponding to the size of the square-shaped display area 21 shown in Figure 1) will be referred to below as a unit area.
[0033] The first layout condition is a layout in which one or more markers 22 are placed within a unit area. The second layout condition is a layout in which two or more types of calibration indicators 23 of different sizes (in the example shown in Figure 1, there are two types, large and small, but there may be three types, large, medium and small, or four or more types) are placed within a unit area with a size ratio of 1:1.71 or larger. The third layout condition is a layout in which a specified number of large-size calibration indicators 23a, ranging from four to twelve, are placed around the markers 22 within a unit area. The fourth layout condition is a layout in which four or more small-size calibration indicators 23b are placed within a unit area by placing small-size calibration indicators 23b to fill the blank areas within the unit area (areas where markers 22 and large-size calibration indicators 23a are not placed), or by placing small-size calibration indicators 23b on top of large-size calibration indicators 23a, or by combining these arrangements.
[0034] Regarding the fourth layout condition, a layout may be as shown in the calibration pattern 11a in Figure 3, in which four small-sized calibration indicators 23b-1 to 23b-4 are placed in the blank area within the unit area. Alternatively, a layout may be as shown in the calibration pattern 11b in Figure 4, in which four large-sized calibration indicators 23a-1 to 23a-4 are overlaid with small-sized calibration indicators 23b-1 to 23b-4. Of course, a layout combining both of these is also acceptable.
[0035] Furthermore, when adopting a layout in which small-sized calibration indicators 23b are simply placed in the blank areas within a unit area, as shown in the calibration pattern 11c in Figure 5, a layout in which a larger number of small-sized calibration indicators 23b (in the illustrated example, nine small-sized calibration indicators 23b-1 to 23b-9 in the blank areas within a unit area) may also be used.
[0036] Furthermore, as shown in the calibration pattern 11d in Figure 6, a layout may be used in which small-size calibration indicators 23b-1 to 23b-4 are superimposed on four large-size calibration indicators 23a-1 to 23a-4, and a large number of small-size calibration indicators 23b (in the illustrated example, nine small-size calibration indicators 23b-5 to 23b-13) are placed in the blank areas within the unit area.
[0037] <Definition of Calibration Pattern Layout> The definition of the calibration pattern 11 layout will be explained with reference to Figure 7.
[0038] First, we consider the number of large-size calibration indicators 23a-1 that can be placed around the marker 22 in terms of pixel size. Figure 7 shows a calibration pattern 11 drawn with 2-pixel lines, where diagonally hatched pixels represent the marker 22 and the large-size calibration indicators 23a, and dotted hatched pixels represent the margins necessary to distinguish between the marker 22 and the large-size calibration indicators 23a.
[0039] As shown in Figure 7A, the marker 22 can be represented by 12 x 12 pixels. If a 1-pixel margin is provided around the marker 22, a pixel size of 14 x 14 pixels is required to place the marker 22. Furthermore, for the minimum size of the large-size calibration index 23a, if a 1-pixel margin is provided around the large-size calibration index 23a, a pixel size of 6 x 6 pixels is required to place the large-size calibration index 23a.
[0040] Accordingly, the maximum number of large-size calibration indicators 23a that can be arranged to surround the periphery of the marker 22 is 12, as shown in A of Fig.7. Further, to surround the periphery of the marker 22, at least 4 large-size calibration indicators 23a are required. That is, as in the third layout condition described above, a layout is defined in which 4 to 12 large-size calibration indicators 23a are arranged around the marker 22.
[0041] Next, consideration is given to applying the pixel-size calibration pattern 11 based on coordinates [pix] on an image to the actual-size calibration pattern 11 based on three-dimensional coordinates [mm] in the real world. For example, using the resolution R [pix / mm] determined from the number of pixels / sensor size, the focal length f [mm] of the camera, and three-dimensional coordinates [mm] represented by the x-axis, y-axis, and z-axis, the actual size u of the large-size calibration indicator 23a can be expressed by the following equation (1).
[0042]
[0043] Further, when the minimum size of the large-size calibration indicator 23a is a pixel size of 6×6 pixels as described above, the minimum actual size x of the large-size calibration indicator 23a that can be recognized by a camera with resolution R and focal length f min can be expressed by the following equation (2).
[0044]
[0045] Then, consideration is given to the size of the small-size calibration indicator 23b when the size of the large-size calibration indicator 23a is the minimum actual size x determined by this equation (2). min The size of the small-size calibration indicator 23b in this case is discussed.
[0046] For example, as shown by the black circle in B of Fig.7, the small-size calibration indicator 23b can be arranged in a blank area where the marker 22 and the large-size calibration indicator 23a are not arranged.
[0047] Further, when the small-size calibration indicator 23b is represented by 1×1 pixel, the minimum actual size x of the small-size calibration indicator 23b min ' is the minimum actual size x of the large-size calibration indicator 23amin This becomes 1 / 6 the size. Therefore, the balance between the resolution R and focal length f of a camera capable of recognizing the small-size calibration index 23b of this size can be determined by the following equation (3).
[0048]
[0049] In other words, based on equation (3), the balance between the camera's resolution R and focal length f should be such that f × R is 6 times.
[0050] Furthermore, we will consider the possibility that the difference in f×R between multiple cameras may be less than 6 times.
[0051] For example, when actually calibrating a camera using the calibration pattern 11, the angle of the calibration pattern 11 is changed while taking pictures with a fixed camera, so it is necessary to take into account the robustness of recognition when taking pictures at an angle. For this reason, in practice, as shown in Figure 7C, large-sized calibration indicators 23a of the same size as the marker 22 are sometimes laid out in four places surrounding the marker 22.
[0052] In this layout, small-size calibration indicators 23b can be placed in the blank areas where markers 22 and large-size calibration indicators 23a are not located, as long as they are even slightly smaller than the large-size calibration indicators 23a. However, the size relationship when using small-size calibration indicators 23b of the size shown by the white circles in Figure 7C is negligible (i.e., it is not possible to place a large number of small-size calibration indicators 23b), so it is ignored.
[0053] Therefore, assuming the size relationship when using a small-size calibration index 23b, which is the size shown by the black circle in Figure 7C, the size of the small-size calibration index 23b (black circle) relative to the large-size calibration index 23a is 7 / 12, according to the pixel ratio shown in Figure 7C. From this, according to the following equation (4), the f × R of multiple cameras should be between 12 / 7 and 1.71 times.
[0054]
[0055] Thus, in order to increase the number of small-size calibration indicators 23b that can be placed in a limited area, it is necessary to arrange the small-size calibration indicators 23b to fill the empty areas, or to place the small-size calibration indicators 23b on top of the large-size calibration indicators 23a, as described above.
[0056] <Effects of Calibration Patterns> The effects of calibration patterns 11 will be explained with reference to Figures 8 to 10. In the conventional calibration patterns 12 used in Figures 8 and 9, four large-sized calibration indicators are placed above, below, to the left and right of the marker located in the center, and four medium-sized calibration indicators (calibration indicators that are at least 1 / 1.71 times the size of the large-sized calibration indicators) are placed in the blank areas.
[0057] Figure 8 illustrates an example of calibrating cameras with a constant field of view difference but different resolution differences.
[0058] First, as shown in Resolution Difference Case 1, we consider the case where the calibration pattern 11 of this technology and the calibration pattern 12 of the conventional technology are arranged so that they fit within the camera field of view 32 on the high-resolution telephoto side, and the calibration pattern 11 of this technology and the calibration pattern 12 of the conventional technology are photographed using the camera field of view 31 on the low-resolution wide-angle side and the camera field of view 32 on the high-resolution telephoto side. In this case, regardless of whether the calibration pattern 11 of this technology is photographed using the camera field of view 31 on the low-resolution wide-angle side or the camera field of view 32 on the high-resolution telephoto side, it is possible to photograph the number of large-size calibration indices 23a or small-size calibration indices 23b necessary to determine the camera calibration parameters. Similarly, regardless of whether the calibration pattern 12 of the conventional technology is photographed using the camera field of view 31 on the low-resolution wide-angle side or the camera field of view 32 on the high-resolution telephoto side, it is possible to photograph the number of large-size calibration indices or medium-size calibration indices necessary to determine the camera calibration parameters.
[0059] Therefore, in this case, whether the calibration pattern 11 of this technology or the calibration pattern 12 of the conventional technology is used, it is possible to calibrate both the low-resolution wide-angle camera field of view 31 and the high-resolution telephoto camera field of view 32.
[0060] Next, as shown in Resolution Difference Case 2, we consider the case where the calibration pattern 11 of this technology and the calibration pattern 12 of the conventional technology are arranged so that they fit within the low-resolution wide-angle camera field of view 31, and the calibration pattern 11 of this technology and the calibration pattern 12 of the conventional technology are photographed using both the low-resolution wide-angle camera field of view 31 and the high-resolution telephoto camera field of view 32. In this case, regardless of whether the calibration pattern 11 of this technology is photographed using the low-resolution wide-angle camera field of view 31 or the high-resolution telephoto camera field of view 32, it is possible to photograph the number of large-size calibration indices 23a or small-size calibration indices 23b necessary to determine the camera calibration parameters. On the other hand, when the conventional calibration pattern 12 is photographed with the low-resolution wide-angle camera field of view 31, it is possible to photograph the number of large-size or medium-size calibration indicators necessary to determine the camera calibration parameters. However, when the conventional calibration pattern 12 is photographed with the high-resolution telephoto camera field of view 32, it is not possible to photograph the number of large-size or medium-size calibration indicators necessary to determine the camera calibration parameters.
[0061] Therefore, in this case, by using the calibration pattern 11 of this technology, it is possible to calibrate both the low-resolution wide-angle camera field of view 31 and the high-resolution telephoto camera field of view 32. In contrast, even when using the calibration pattern 12 of the conventional technology, it is possible to calibrate the camera with the low-resolution wide-angle camera field of view 31, but it is not possible to calibrate the camera with the high-resolution telephoto camera field of view 32.
[0062] Thus, by using the calibration pattern 11 of this technology, it is possible to calibrate multiple cameras with a wider range of resolution differences than would be possible with the calibration pattern 12 of the conventional technology.
[0063] Figure 9 illustrates an example of calibrating cameras with constant resolution differences but different field-of-view differences.
[0064] First, we consider the case where the calibration pattern 11 of this technology and the calibration pattern 12 of the conventional technology are photographed using the low-resolution wide-angle camera field of view 31 and the high-resolution telephoto camera field of view 32a, as shown in the field of view difference case 1. In this case, regardless of whether the calibration pattern 11 of this technology is photographed using the low-resolution wide-angle camera field of view 31 or the high-resolution telephoto camera field of view 32a, it is possible to photograph the number of large-size calibration indices 23a or small-size calibration indices 23b necessary to determine the camera calibration parameters. Similarly, regardless of whether the calibration pattern 12 of the conventional technology is photographed using the low-resolution wide-angle camera field of view 31 or the high-resolution telephoto camera field of view 32a, it is possible to photograph the number of large-size calibration indices or medium-size calibration indices necessary to determine the camera calibration parameters.
[0065] Therefore, as shown in the case of angle of view difference 1, when the angle of view difference is small, it is possible to calibrate both the low-resolution wide-angle camera angle of view 31 and the high-resolution telephoto camera angle of view 32a regardless of whether the calibration pattern 11 of this technology or the calibration pattern 12 of the conventional technology is used.
[0066] Next, we consider the case where the calibration pattern 11 of this technology and the calibration pattern 12 of the conventional technology are photographed with the low-resolution wide-angle camera field of view 31 and the high-resolution telephoto camera field of view 32b, as shown in the field of view difference case 2. In this case, when the calibration pattern 11 of this technology is photographed with either the low-resolution wide-angle camera field of view 31 or the high-resolution telephoto camera field of view 32b, it is possible to photograph the number of large-size calibration indices 23a or small-size calibration indices 23b necessary to determine the camera calibration parameters. On the other hand, when the calibration pattern 12 of the conventional technology is photographed with the low-resolution wide-angle camera field of view 31, it is possible to photograph the number of large-size calibration indices or medium-size calibration indices necessary to determine the camera calibration parameters, but when the calibration pattern 12 of the conventional technology is photographed with the high-resolution telephoto camera field of view 32b, it is not possible to photograph the number of large-size calibration indices or medium-size calibration indices necessary to determine the camera calibration parameters.
[0067] Therefore, as shown in the case of angle of view difference 2, when the angle of view difference is moderate, the calibration pattern 11 of this technology can be used to calibrate both the low-resolution wide-angle camera angle of view 31 and the high-resolution telephoto camera angle of view 32b. In contrast, even when using the conventional calibration pattern 12, while it can calibrate the camera with the low-resolution wide-angle camera angle of view 31, it cannot calibrate the camera with the high-resolution telephoto camera angle of view 32b. In this case, since there is margin in the high-resolution telephoto camera angle of view 32b, it is possible to calibrate the camera with the high-resolution telephoto camera angle of view 32b by, for example, increasing the size of the board to enlarge the size of the conventional calibration pattern 12, as shown by the dashed line, and placing a medium-sized calibration index in the enlarged portion. However, with such an enlarged conventional calibration pattern 12, the position where the medium-sized calibration index placed in the enlarged portion of the board is detected is biased within the angle of view of the high-resolution telephoto camera angle of view 32b, raising concerns about a decrease in calibration accuracy.
[0068] Next, we consider the case where the calibration pattern 11 of this technology and the calibration pattern 12 of the conventional technology are photographed with the low-resolution wide-angle camera field of view 31 and the high-resolution telephoto camera field of view 32c, as shown in the field of view difference case 3. In this case, when the calibration pattern 11 of this technology is photographed with either the low-resolution wide-angle camera field of view 31 or the high-resolution telephoto camera field of view 32c, it is possible to photograph the number of large-size calibration indices 23a or small-size calibration indices 23b necessary to determine the camera calibration parameters. On the other hand, when the calibration pattern 12 of the conventional technology is photographed with the low-resolution wide-angle camera field of view 31, it is possible to photograph the number of large-size calibration indices or medium-size calibration indices necessary to determine the camera calibration parameters, but when the calibration pattern 12 of the conventional technology is photographed with the high-resolution telephoto camera field of view 32c, it is not possible to photograph the number of large-size calibration indices or medium-size calibration indices necessary to determine the camera calibration parameters.
[0069] Therefore, as shown in the angle of view difference case 3, when the angle of view difference is large, the calibration pattern 11 of this technology can be used to calibrate both the low-resolution wide-angle camera angle of view 31 and the high-resolution telephoto camera angle of view 32c. In contrast, even when using the conventional calibration pattern 12, while it can calibrate the camera with the low-resolution wide-angle camera angle of view 31, it cannot calibrate the camera with the high-resolution telephoto camera angle of view 32c. In this case, since there is no margin for the high-resolution telephoto camera angle of view 32c, even if the size of the conventional calibration pattern 12 is enlarged as in the angle of view difference case 2 described above, it cannot calibrate the camera with the high-resolution telephoto camera angle of view 32c.
[0070] Thus, by using the calibration pattern 11 of this technology, it is possible to calibrate multiple cameras with a wider field of view difference than would be possible with the calibration pattern 12 of the conventional technology.
[0071] Figure 10 is a schematic diagram illustrating, based on the explanation with reference to Figures 8 and 9, that by using the calibration pattern 11, it is possible to accommodate a wider range of camera combinations with different resolutions and field of view than when using the conventional calibration pattern 12.
[0072] As shown in Figure 10, by using the calibration pattern 11 to which this technology is applied, it is possible to accommodate cameras with a wider resolution difference and a wider field of view difference compared to using the conventional calibration pattern 12. Furthermore, even when using the conventional calibration pattern 12, the range of field of view differences that can be accommodated can be expanded by increasing the size of the conventional calibration pattern 12, as explained with reference to the field of view difference case 2 in Figure 9, but the calibration pattern 11 to which this technology is applied is capable of accommodating cameras with an even wider field of view difference.
[0073] <Types of Calibration Indices and Markers> Figure 11 shows examples of the types of calibration indices 23 and markers 22 used in the calibration pattern 11.
[0074] The calibration index 23 can use basic geometric patterns that allow for the determination of the centroid, such as circles, squares, or equilateral triangles, as shown in Figure 11A. For example, in camera calibration, using the centroid as a feature point instead of corner detection allows for maintaining position detection accuracy better than corner detection, even when the calibration pattern 11 is not directly facing the camera and the calibration index 23 is distorted, or when using a low-resolution camera.
[0075] In addition to the ArUco marker shown in Figure 1, various types of markers can be used for the marker 22, as long as they have a uniquely determined pattern that can recognize the up, down, left, and right orientation of the calibration pattern 11, as shown in Figure 11B.
[0076] <Variations of Calibration Indices> Refer to Figures 12 to 19 to explain the variations of the calibration pattern 11. In the following, the calibration pattern 11 is divided into three sections: upper, middle, and lower, and further divided into three sections: left, center, and right, to explain the layout of the calibration pattern 11 for a 3x3 divided area.
[0077] Figure 12 shows a first variation of the calibration pattern 11.
[0078] As shown in Figure 12, in the calibration pattern 11-1, marker 22 is placed in the central division area of the middle section, large-size calibration index 23a-1 is placed in the central division area of the upper section, large-size calibration index 23a-2 is placed in the right division area of the middle section, large-size calibration index 23a-3 is placed in the central division area of the lower section, and large-size calibration index 23a-4 is placed in the left division area of the middle section.
[0079] Furthermore, in calibration pattern 11-1, the small-size calibration indicators 23b-1 to 23b-4 are superimposed on the large-size calibration indicators 23a-1 to 23a-4, respectively. In calibration pattern 11-1, the small-size calibration indicators 23b-5 to 23b-16 are randomly placed to fill the blank areas in the divided regions of the upper left, upper right, lower right, and lower left sections.
[0080] Thus, the small-size calibration indicators 23b-5 to 23b-16, which are arranged to fill the blank areas, may be arranged randomly rather than regularly.
[0081] Figure 13 shows a second variation of the calibration pattern 11.
[0082] As shown in Figure 13, in calibration pattern 11-2, marker 22 is placed in the central divided area of the middle section, and large-size calibration indices 23a-1 to 23a-8 are placed in the other divided areas. Furthermore, in calibration pattern 11-2, small-size calibration indices 23b-1 and 23b-2 are placed on top of large-size calibration indice 23a-1, small-size calibration indices 23b-3 and 23b-4 are placed on top of large-size calibration indice 23a-2, and so on, with two small-size calibration indices 23b being placed on top of large-size calibration indice 23a.
[0083] In this way, multiple small-size calibration indices 23b (two small-size calibration indices 23b in the example shown in Figure 13) may be placed on top of one large-size calibration indice 23a.
[0084] Figure 14 shows a third variation of the calibration pattern 11.
[0085] As shown in Figure 14, in the calibration pattern 11-3, marker 22 is placed in the central division area of the middle section, large-size calibration index 23a-1 is placed in the central division area of the upper section, large-size calibration index 23a-2 is placed in the right division area of the middle section, large-size calibration index 23a-3 is placed in the central division area of the lower section, and large-size calibration index 23a-4 is placed in the left division area of the middle section.
[0086] In the calibration pattern 11-3, the medium-sized calibration indices 23c-1 to 23c-4 are arranged superimposed on the large-sized calibration indices 23a-1 to 23a-4, respectively, and the small-sized calibration indices 23b-1 to 23b-4 are arranged superimposed on the medium-sized calibration indices 23c-1 to 23c-4, respectively. Furthermore, in calibration pattern 11-3, in the upper left division area, the small size calibration index 23b-5 is superimposed on the medium size calibration index 23c-5; in the upper right division area, the small size calibration index 23b-6 is superimposed on the medium size calibration index 23c-6; in the lower right division area, the small size calibration index 23b-7 is superimposed on the medium size calibration index 23c-7; and in the lower left division area, the small size calibration index 23b-8 is superimposed on the medium size calibration index 23c-8.
[0087] Furthermore, in the calibration pattern 11-3, small-size calibration indicators 23b-9 to 23b-13 are randomly placed to fill the blank areas in the upper left division area, small-size calibration indicators 23b-14 and 23b-15 are randomly placed to fill the blank areas in the upper right division area, small-size calibration indicators 23b-16 to 23b-18 are randomly placed to fill the blank areas in the lower right division area, and small-size calibration indicators 23b-19 to 23b-22 are randomly placed to fill the blank areas in the lower left division area.
[0088] Furthermore, the size ratio of the large-size calibration index 23a to the medium-size calibration index 23c, and the size ratio of the medium-size calibration index 23c to the small-size calibration index 23b, are set to 1:1.71 or higher, in accordance with the second layout condition described above.
[0089] In this way, multiple calibration indices 23 of different sizes (in the example shown in Figure 14, a large calibration indice 23a, a medium calibration indice 23c, and a small calibration indice 23b) may be arranged.
[0090] Figure 15 shows a fourth variation of the calibration pattern 11.
[0091] As shown in Figure 15, in calibration pattern 11-4, marker 22-1 is placed in the left-hand division area of the middle section, and marker 22-2 is placed in the right-hand division area of the middle section. In addition, in calibration pattern 11-4, large-size calibration indices 23a-1 to 23a-3 are placed in the three division areas of the upper section, with small-size calibration indices 23b-1 to 23b-3 superimposed on each. Similarly, in calibration pattern 11-4, large-size calibration indices 23a-4 to 23a-6 are placed in the three division areas of the lower section, with small-size calibration indices 23b-4 to 23b-6 superimposed on each. Furthermore, small-size calibration indices 23b-7 to 23b-10 are placed in a 2x2 arrangement in the central division area of the middle section.
[0092] In this way, multiple markers 22 (in the example shown in Figure 15, markers 22-1 and marker 22-2) may be arranged.
[0093] Figure 16 shows a fifth variation of the calibration pattern 11.
[0094] As shown in Figure 16, in the calibration pattern 11-5, marker 22 is placed in the central division area of the middle section, a large square-shaped calibration index 23d-1 is placed in the central division area of the upper section, a large square-shaped calibration index 23d-2 is placed in the right division area of the middle section, a large square-shaped calibration index 23d-3 is placed in the central division area of the lower section, and a large square-shaped calibration index 23d-4 is placed in the left division area of the middle section.
[0095] Furthermore, in calibration pattern 11-5, the small-sized square calibration indicators 23e-1 to 23e-4 are arranged superimposed on the large-sized square calibration indicators 23d-1 to 23d-4, respectively. In the calibration pattern 11-5, the small square-shaped calibration indicators 23e-5 to 23e-8 are arranged in a 2x2 pattern in the upper left division area, the small square-shaped calibration indicators 23e-9 to 23e-12 are arranged in a 2x2 pattern in the upper right division area, the small square-shaped calibration indicators 23e-13 to 23e-16 are arranged in a 2x2 pattern in the lower right division area, and the small square-shaped calibration indicators 23e-17 to 23e-20 are arranged in a 2x2 pattern in the lower left division area.
[0096] Thus, the calibration index 23 can be a shape other than a circle (in the example shown in Figure 16, it is a square shape).
[0097] Figure 17 shows a sixth variation of the calibration pattern 11.
[0098] As shown in Figure 17, in the calibration pattern 11-6, a different type of marker 22a (in the illustrated example, a pattern with small black circles is provided) is placed in the central divided region of the middle section.
[0099] Furthermore, in calibration pattern 11-6, a large-size calibration index 23a-1 is placed in the central division area of the upper section, a large-size calibration index 23a-2 is placed in the right division area of the middle section, a large-size calibration index 23a-3 is placed in the central division area of the lower section, and a large-size calibration index 23a-4 is placed in the left division area of the middle section. In addition, in calibration pattern 11-6, small-size calibration indexes 23b-1 to 23b-4 are placed on top of each other, with small-size calibration indexes 23b-5 to 23b-8 each placed in the left, right, right, and left division areas of the upper section, respectively.
[0100] Thus, markers other than the ArUco markers described above can be used for marker 22a.
[0101] Figure 18 shows the seventh variation of the calibration pattern 11.
[0102] As shown in Figure 18, in calibration pattern 11-7, a marker 22a of a different type from the ArUco marker described above is placed in the central divided region of the middle section. Furthermore, in calibration pattern 11-7, similar to calibration pattern 11-5 in Figure 16, large square-shaped calibration indicators 23d-1 to 23d-4, and small square-shaped calibration indicators 23e-1 to 23e-20 are placed.
[0103] Thus, it is possible to use a combination of markers 22a other than the ArUco markers described above, and calibration indices 23 with shapes other than circular (in the example shown in Figure 18, a square shape).
[0104] Figure 19 shows the eighth variation of the calibration pattern 11.
[0105] As shown in Figure 19, in calibration pattern 11-8, a marker 22a of a different type from the ArUco marker described above is placed in the central divided region of the middle section. Furthermore, in calibration pattern 11-8, a large circular calibration index 23a-1 is placed in the central divided region of the upper section, a large circular calibration index 23a-2 is placed in the right divided region of the middle section, a large circular calibration index 23a-3 is placed in the central divided region of the lower section, and a large circular calibration index 23a-4 is placed in the left divided region of the middle section.
[0106] In calibration pattern 11-8, the small square-shaped calibration indices 23e-1 to 23e-4 are arranged superimposed on the large circular calibration indices 23a-1 to 23a-4, respectively. Also, in calibration pattern 11-8, similar to calibration pattern 11-5 in Figure 16 described above, the small square-shaped calibration indices 23e-5 to 23e-20 are arranged.
[0107] In this way, calibration indices 23 of different shapes (in the example shown in Figure 19, a large circular calibration indice 23a and a small square calibration indice 23e) can be used in combination.
[0108] The calibration patterns 11 of the first to eighth variations described above can be used, and other variations of the calibration patterns 11 may also be used according to the first to fourth layout conditions described above.
[0109] <Example of Calibration Process> Referring to the flowchart shown in Figure 20, the camera calibration process performed on the computer using the calibration pattern 11 will be explained.
[0110] For example, a low-resolution camera and a high-resolution camera are fixed in place, and the user holds a board on which the calibration pattern 11 is drawn, changing the angle of the calibration pattern 11 while taking multiple images with the low-resolution and high-resolution cameras. As a result, in step S11, the computer acquires the necessary number of images (for example, about 10) of the calibration pattern 11 taken from various angles, which is required for calibration.
[0111] Then, in steps S12 to S17, the image of the calibration pattern 11 captured by the low-resolution camera is processed according to a low-resolution calibration algorithm, and in steps S18 to S21, the image of the calibration pattern 11 captured by the high-resolution camera is processed according to a high-resolution calibration algorithm.
[0112] In step S12, the computer processes one of the images taken with the low-resolution camera acquired in step S11 and detects the markers 22 of the calibration pattern 11 that are depicted in that image.
[0113] In step S13, the computer detects large-size calibration indices 23a-1 to 23a-4 from the image to be processed.
[0114] In step S14, the computer identifies the large-size calibration indicators 23a-1 to 23a-4 detected in step S13.
[0115] In step S15, the computer performs a homography transformation on the image to be processed based on the large-size calibration indices 23a-1 to 23a-4 identified in step S14.
[0116] In step S16, the computer detects and identifies small-size calibration indices 23b-1 to 23b-9 from the image that underwent homography transformation in step S15.
[0117] In step S17, the computer calculates camera calibration parameters for the low-resolution camera based on the small-size calibration indices 23b-1 to 23b-9 identified in step S16.
[0118] Meanwhile, in step S18, the computer processes one of the images captured by the high-resolution camera acquired in step S11 and detects the calibration pattern 11 markers 22 depicted in that image.
[0119] In step S19, the computer performs a homography transformation on the image to be processed based on the four feature points (for example, the four corners of the marker 22) of the calibration pattern 11 detected in step S18.
[0120] In step S20, the computer detects and identifies large-size calibration indices 23a-1 to 23a-4 and small-size calibration indices 23b-1 to 23b-9 from the image that underwent homography transformation in step S19.
[0121] In step S21, the computer calculates camera calibration parameters for the high-resolution camera based on the small-size calibration indices 23b-1 to 23b-9 identified in step S20.
[0122] After the processing in step S17 or step S21, the process proceeds to step S22, in which the computer calculates further camera calibration parameters for the low-resolution camera and the high-resolution camera.
[0123] In step S23, the computer determines whether processing has been performed on the required number of images acquired in step S11. If it determines that processing has not been performed on the required number of images, the process returns to step S12 or step S18. Then, the same process is repeated for the images that have not yet been processed.
[0124] On the other hand, if the computer determines in step S23 that processing has been performed on the required number of images acquired in step S11, the process proceeds to step S24.
[0125] In step S24, the computer calculates the final camera calibration parameters based on the camera calibration parameters obtained from the required number of images. The computer then outputs the final camera calibration parameters to the low-resolution camera and the high-resolution camera, and then terminates the process.
[0126] By performing the calibration process described above, the computer can process images of the calibration pattern 11 captured by a low-resolution camera according to a low-resolution calibration algorithm, and images of the calibration pattern 11 captured by a high-resolution camera according to a high-resolution calibration algorithm, thereby determining the camera calibration parameters for each. In this way, calibration can be performed using a common calibration pattern 11 for both low-resolution and high-resolution cameras, thus improving convenience.
[0127] Furthermore, the processes performed by a computer do not necessarily have to follow the order described in the flowchart. In other words, the processes performed by a computer include processes that are executed in parallel or individually (for example, parallel processing or object-based processing). Also, a program may be processed by a single computer (processor), or it may be processed in a distributed manner by multiple computers. Moreover, it may be transferred to a remote computer and executed there.
[0128] Furthermore, each step described in the flowchart above can be executed by a single device or shared among multiple devices. Additionally, if a single step includes multiple processes, these processes can be executed by a single device or shared among multiple devices. In other words, multiple processes within a single step can be executed as multiple steps. Conversely, processes described as multiple steps can be combined and executed as a single step.
[0129] <An example of black paint used for calibration patterns> Referring to Figure 21, for example, a suitable black paint for printing calibration patterns 11 will be described.
[0130] The calibration pattern 11 to which this technology is applied is expected to be used, for example, when performing calibration in combination with an RGB camera that receives and captures light in the visible light region and an infrared camera that receives and captures light in the near-infrared region. In this case, it is necessary that pattern recognition of the marker 22 and calibration index 23 is possible in both images acquired with a combination of a high-resolution RGB camera and a low-resolution infrared camera, and images acquired with a combination of a high-resolution infrared camera and a low-resolution RGB camera. Therefore, it is desirable that the calibration pattern 11 satisfies conditions such as having a clear contrast between white and black, a low reflectivity of black being preferable, and low reflectivity in both visible and near-infrared light. In other words, it is preferable to print the calibration pattern 11 using black paint that satisfies such conditions.
[0131] Figure 21A shows an example of the reflectivity characteristics of black paint. In Figure 21A, the horizontal axis represents the wavelength of light, and the vertical axis represents the reflectivity of light of each wavelength reflected by the black paint. As shown in Figure 21A, black paint with low reflectivity characteristics that maintain a similar reflectivity below a certain value over a wide range from the visible light region to the near-infrared light region is suitable for use in the calibration pattern 11. Conversely, black paint with reflectivity characteristics that result in higher reflectivity in the near-infrared light region compared to the visible light region is not suitable for use in the calibration pattern 11.
[0132] Figure 21B shows examples of images obtained by photographing a calibration pattern 11 printed with a black paint having suitable reflectivity characteristics using an RGB camera and an infrared camera. Figure 21C shows examples of images obtained by photographing a calibration pattern 11 printed with a black paint having unsuitable reflectivity characteristics using an RGB camera and an infrared camera.
[0133] As shown in Figure 21B, a calibration pattern 11 printed using black paint with suitable reflectivity characteristics can be recognized regardless of whether the image is acquired with an RGB camera or an infrared camera. In contrast, as shown in Figure 21C, a calibration pattern 11 printed using black paint with unsuitable reflectivity characteristics can be recognized in images acquired with an RGB camera, but it is difficult to recognize the pattern in images acquired with an infrared camera. In other words, black paint with unsuitable reflectivity characteristics has high reflectivity in the near-infrared region, so as shown in the figure, the contrast in images acquired with an infrared camera is reduced by the near-infrared light reflected from the black areas.
[0134] Thus, in order to enable pattern recognition of the calibration pattern 11 regardless of whether the image is acquired with an RGB camera or an infrared camera, it is preferable to use a black paint with low reflectivity characteristics over a wide range from the visible light region to the near-infrared light region for the calibration pattern 11.
[0135] Furthermore, the technologies described in this specification can be implemented independently, as long as they do not create a contradiction. Of course, any multiple technologies can also be implemented in combination. For example, some or all of the technologies described in one embodiment can be combined with some or all of the technologies described in another embodiment. In addition, some or all of the above-mentioned technologies can be implemented in combination with other technologies not mentioned above.
[0136] <Examples of Configuration Combinations> The technology can also take the following configurations: (1) A calibration pattern in which four or more small calibration indicators are arranged in the unit area, by arranging one or more markers in a predetermined unit area, at least two types of large-size calibration indicators and small-size calibration indicators of different sizes with a predetermined size ratio or greater as calibration indicators used to calculate camera calibration parameters, arranging a predetermined number of the large-size calibration indicators around the markers in the unit area, and arranging the small-size calibration indicators to fill the blank areas in the unit area where the markers and large-size calibration indicators are not arranged, or by arranging the small-size calibration indicators on top of the large-size calibration indicators, or by combining these arrangements. (2) The calibration pattern according to (1) above, wherein the unit area is a rectangular area in which the large-size calibration indicators can be arranged at the maximum density when multiple large-size calibration indicators are arranged to surround one marker. (3) The calibration pattern according to (1) or (2) above, wherein the size ratio of the large size calibration index to the small size calibration index is 1:1.71 or greater. (4) The calibration pattern according to any of (1) to (3) above, wherein the specified number of large size calibration indexes to be placed is in the range of 4 to 12. (5) The calibration pattern according to any of (1) to (4) above, wherein one marker is placed in the center of the unit area, one large size calibration index is placed above, below, to the left and to the right of the marker, one small size calibration index is placed on top of each large size calibration index, and multiple small size calibration indexes are placed to fill the blank area. (6) The calibration pattern according to any of (1) to (5) above, wherein multiple small size calibration indexes are randomly placed in the blank area. (7) The calibration pattern according to any one of (1) to (6) above, wherein two or more small calibration indicators are arranged in superimposition with respect to one large calibration indicator.(8) A calibration pattern according to any one of (1) to (7) above, in which large, medium, and small calibration indices of different sizes, having a size ratio greater than or equal to the above-mentioned size ratio, are used. (9) A calibration pattern according to any one of (1) to (8) above, in which two markers are arranged within the unit area. (10) A calibration pattern according to any one of (1) to (9) above, in which a circular shape is used as the calibration indice. (11) A calibration pattern according to any one of (1) to (9) above, in which a square shape is used as the calibration indice. (12) A calibration pattern according to any one of (1) to (9) above, in which a combination of a circular shape and a square shape is used as the calibration indice. (13) A calibration pattern according to any one of (1) to (12) above, in which a black paint with reflectivity characteristics that has low reflectivity over a wide range from the visible light region to the near-infrared light region is used. (14) A calibration method comprising: (14) Arranging one or more markers within a predetermined unit area; arranging at least two types of large-size calibration indicators and small-size calibration indicators of different sizes with a predetermined size ratio or greater within the unit area as calibration indicators used to calculate camera calibration parameters; arranging a predetermined number of the large-size calibration indicators around the markers within the unit area; arranging the small-size calibration indicators to fill the blank areas in the unit area where the markers and large-size calibration indicators are not arranged, or arranging the small-size calibration indicators on top of the large-size calibration indicators, or combining these arrangements to arrange four or more small-size calibration indicators within the unit area; processing an image of the calibration pattern captured by a low-resolution camera according to a low-resolution calibration algorithm; and processing an image of the calibration pattern captured by a high-resolution camera according to a high-resolution calibration algorithm.
[0137] It should be noted that this embodiment is not limited to the embodiment described above, and various modifications are possible without departing from the spirit of this disclosure. Furthermore, the effects described herein are merely illustrative and not limiting, and other effects may also exist.
[0138] 11 Calibration pattern, 21 Display area, 22 Marker, 23 Calibration index, 23a Large size calibration index, 23b Small size calibration index, 31 Low resolution wide-angle camera field of view, 32 High resolution telephoto camera field of view
Claims
1. A calibration pattern comprising:
1. Placing one or more markers within a predetermined unit area; 2. Placing at least two types of large-size calibration indicators and small-size calibration indicators of different sizes with a predetermined size ratio or greater within the unit area as calibration indicators used to calculate camera calibration parameters; 3. Placing a predetermined number of the large-size calibration indicators around the markers within the unit area; 4 or more of the small-size calibration indicators within the unit area by arranging the small-size calibration indicators to fill the blank areas in which the markers and large-size calibration indicators are not placed within the unit area, or by arranging the small-size calibration indicators on top of the large-size calibration indicators, or by combining these arrangements.
2. The calibration pattern according to claim 1, wherein the unit area is a rectangular area in which the large-size calibration indicators can be arranged at the maximum density when a plurality of the large-size calibration indicators are arranged to surround one of the markers.
3. The calibration pattern according to claim 1, wherein the size ratio of the large-size calibration index to the small-size calibration index is 1:1.71 or greater.
4. The calibration pattern according to claim 1, wherein the specified number of large-size calibration indicators to be arranged is in the range of 4 to 12.
5. The calibration pattern according to claim 1, wherein one marker is placed in the center of the unit area, one large calibration index is placed above, below, to the left and to the right of the marker, one small calibration index is placed on top of each of the large calibration index, and a plurality of small calibration indexes are placed to fill the blank area.
6. The calibration pattern according to claim 1, wherein a plurality of the small-sized calibration indicators are randomly arranged in the blank area.
7. The calibration pattern according to claim 1, wherein two or more small calibration indicators are arranged in superimposition with respect to one large calibration indicator.
8. The calibration pattern according to claim 1, wherein a plurality of calibration indices of different sizes are used in a size ratio greater than or equal to the aforementioned size ratio.
9. The calibration pattern according to claim 1, wherein two markers are arranged within the unit area.
10. The calibration pattern according to claim 1, wherein a circular shape is used as the calibration index.
11. The calibration pattern according to claim 1, wherein a square shape is used as the calibration index.
12. The calibration pattern according to claim 1, wherein a combination of circular and square shapes is used as the calibration index.
13. The calibration pattern according to claim 1, wherein a black paint having low reflectivity characteristics over a wide range from the visible light region to the near-infrared light region is used.
14. A calibration method comprising: arranging one or more markers within a predetermined unit area; arranging at least two types of large-size calibration indicators and small-size calibration indicators of different sizes with a predetermined size ratio or greater within the unit area as calibration indicators used to calculate camera calibration parameters; arranging a predetermined number of the large-size calibration indicators around the markers within the unit area; arranging the small-size calibration indicators to fill the blank areas in the unit area where the markers and large-size calibration indicators are not arranged, or arranging the small-size calibration indicators on top of the large-size calibration indicators, or combining these arrangements to arrange four or more small-size calibration indicators within the unit area; processing an image of the calibration pattern captured by a low-resolution camera according to a low-resolution calibration algorithm; and processing an image of the calibration pattern captured by a high-resolution camera according to a high-resolution calibration algorithm.
