JP2020027000A - Correction method for lens marker image, correction device, program, and recording medium - Google Patents

Abstract

To provide a correction method of an image of a lens marker capable of suppressing influence of a foreign matter or crack of a detected part even when the foreign matter adheres or crack of a detected part occurs.SOLUTION: The correction method of an image of a lens marker, comprises: an image acquisition step for acquiring an image of a surface of a lens marker, the lens marker being a marker in which a virtual image moves along at least one movement direction according to inclination to an optical axis; an extraction step for extracting brightness level from one end to the other end in a direction orthogonal to the movement direction, for every existing width along the movement direction; a selection step for calculating a first average value and standard deviation of the brightness level for every existing width, and selecting a range of the standard deviation in which the first average value is a center, as the brightness level region; a correction step for calculating a second average value of the brightness level for a pixel corresponding to the brightness level region for every existing width, and setting the second average value to a corrected brightness level; and a re-construction step for re-constructing a corrected image of the lens marker, on the basis of the corrected brightness level of each existing width.SELECTED DRAWING: Figure 5

Description

  The present invention relates to a lens marker image correction method, a correction device, a program, and a recording medium.

  In fields such as Augmented Reality (hereinafter, also referred to as “AR”) and robotics, so-called visual markers are used to recognize the position and orientation of an object. As the visual recognition marker, for example, there is a VMP marker indicating a variable moiré pattern (VMP). As the VMP marker, for example, a marker in which a lenticular lens is arranged on a striped pattern in which black lines as detected portions are arranged is reported (Patent Document 1).

  As the center axis of the VMP marker is inclined with respect to the optical axis, the virtual image pattern of the black line projected on the lens surface on the front surface moves. Therefore, the inclination of the VMP marker can be determined by performing binarization processing on the image of the VMP marker and detecting the position of the virtual image pattern of the black line in the binarized image. Therefore, by using the VMP marker, for example, it is possible to measure the position and orientation of the object on which the VMP marker is arranged.

JP 2012-145559 A

  However, when the VMP marker is used, foreign substances such as sand and dust may adhere thereto, or the detected portion may be chipped. Therefore, when the above-described binarization processing is performed, in the binarized image, in addition to the virtual image pattern to be detected, a black image due to the adhesion of a foreign substance appears, or the detected portion has A white image due to chipping may appear. For this reason, the position of the virtual image pattern may be erroneously detected, and as a result, an error may occur in the measurement of the position and the posture.

  The attachment of the foreign matter and the chipping of the detected portion may occur during use, for example, or may occur during manufacturing. For this reason, before shipment, it is necessary to check the VMP marker for the attachment of foreign matter and the occurrence of chipping of the detected part. If the attachment of foreign matter or chipping of the detected part is confirmed, shipping is performed. Sometimes you can't.

  Accordingly, an object of the present invention is to provide a method of correcting an image of a lens marker that can suppress the influence of the foreign matter attached to the lens marker such as a VMP marker or the occurrence of chipping of the detected portion even when the detected portion is chipped. I do.

In order to achieve the above object, a method for correcting a lens marker image according to the present invention includes:
Obtaining an image of the lens marker,
Brightness level extraction process,
Brightness level area selection process,
Correcting the brightness level in the brightness level region, and reconstructing a corrected image based on the corrected brightness level,
The lens marker is
On one of the surfaces, a marker in which a virtual image moves along at least one moving direction according to an inclination with respect to the optical axis,
The image obtaining step includes:
Obtaining an image of the surface of the lens marker,
The extracting step includes:
For the image, for each predetermined width along the movement direction, extract the brightness level from one end to the other end in the direction orthogonal to the movement direction,
The selecting step includes:
For each of the predetermined widths,
From the distribution of the number of pixels indicating each lightness level, a first average value and a standard deviation (± σ) of the lightness levels of all pixels are calculated,
Selecting a range of a standard deviation centered on the first average value as a brightness level area;
The correcting step includes:
For each of the predetermined widths,
For a pixel corresponding to the brightness level region, a second average value of the brightness level is calculated, and the second average value is set as a corrected brightness level;
The reconstruction step includes:
A corrected image of the lens marker is reconstructed based on the corrected brightness level of each predetermined width.

The lens marker image correcting device of the present invention includes:
Lens marker image acquisition unit,
Brightness level extractor,
Lightness level area selection section,
Including a brightness level correction unit in the brightness level region, and a reconstructed unit of a corrected image based on the corrected brightness level,
The lens marker is
On one of the surfaces, a marker in which a virtual image moves along at least one moving direction according to an inclination with respect to the optical axis,
The image acquisition unit,
Obtaining an image of the surface of the lens marker,
The extraction unit includes:
For the image, for each predetermined width along the movement direction, extract the brightness level from one end to the other end in the direction orthogonal to the movement direction,
The selection unit includes:
For each of the predetermined widths,
From the distribution of the number of pixels indicating each lightness level, a first average value and a standard deviation (± σ) of the lightness levels of all pixels are calculated,
Selecting a range of a standard deviation centered on the first average value as a brightness level area;
The correction unit,
For each of the predetermined widths,
For a pixel corresponding to the brightness level area, a second average value of the brightness level is calculated, and the second average value is set as a corrected brightness level;
The reconstructing unit,
The corrected image of the lens marker is reconstructed based on the corrected brightness level of each of the predetermined widths.

  A program according to the present invention causes a computer to execute the method for correcting a lens marker image according to the present invention.

  The recording medium of the present invention is a computer-readable recording medium on which the program of the present invention is recorded.

  According to the present invention, for example, with respect to a lens marker such as the VMP marker, even when the foreign matter adheres or the detected portion is chipped, the image of the lens marker is corrected with its influence suppressed. Can be. For this reason, according to the present invention, it is possible to avoid these effects and determine the position and orientation of the target object from the inclination of the lens marker, for example.

FIG. 1A is a top view illustrating an example of a lens marker, and FIG. 1B is a cross-sectional view of the marker as viewed from the II direction in FIG. 1A. FIG. 2A is an image of the lens surface of the lens marker on which the image is projected, and FIG. 2B is a graph showing the lightness level on the detection line D in the image. FIG. 3 is a reference example. FIG. 3A is a graph showing a brightness level averaged in the length direction Y, and FIG. 3B is a corrected image based on the averaged brightness level. . FIG. 4 is a schematic diagram in which a virtual image pattern and a foreign substance image appear in an image of the lens surface of the lens marker. FIG. 5 is a schematic diagram illustrating a result of processing in each step in the correction method according to the first embodiment. FIG. 6 is a schematic diagram illustrating a result of processing in each step in the correction method according to the first embodiment. FIG. 7 is a schematic diagram illustrating a result of processing in each step in the correction method according to the first embodiment. FIG. 8 is a plan view illustrating an example of a lens unit. FIG. 9 is a block diagram illustrating an example of the correction device according to the second embodiment. FIG. 10 is a block diagram illustrating an example of a hardware configuration of the correction device according to the second embodiment.

  The correction method, the correction device, the program, and the recording medium (hereinafter, also collectively referred to as “the present invention”) of the present invention can be used for correcting a lens marker image. The target lens marker may be a marker on one surface of which the virtual image moves along at least one moving direction in accordance with the inclination with respect to the optical axis. Since the virtual image that appears on the surface of the lens marker moves in accordance with the inclination with respect to the optical axis, for example, the position and posture of the lens marker (or the position or posture of the object on which the lens marker is disposed) Etc.) can be analyzed based on the fluctuation of the virtual image appearing on the surface of the lens marker.

  The type of the lens marker is not particularly limited, and examples thereof include a VMP marker. The lens marker has, for example, a lens surface on one surface side, and has a detected portion having a repeated pattern on the other surface side, and the pattern of the detected portion is a virtual image that changes on the lens surface. appear. The lens body in the lens marker may be, for example, a lenticular lens or a lens array. The lenticular lens is, for example, a lens body in which a plurality of cylindrical lenses (hereinafter, also referred to as lens portions) obtained by dividing a cylinder in the axial direction are continuously arranged so that the axial directions are parallel. The lens array is, for example, a lens body in which aspherical or spherical lens surfaces are continuously arranged. Examples of the pattern of the detected portion include a striped pattern and a dot pattern formed by a plurality of lines.

  The lens marker is used as a marker set 4, for example, as shown in FIG. In the marker set 4, the two-dimensional code 3 is usually arranged at the center, the inclination detection marks 2 are arranged at four corners, and the lens marker 1 is arranged between a pair of adjacent inclination detection marks 2. By using the marker set 4, for example, the angle of the marker set 4 is detected as follows. That is, first, the two-dimensional code 3 in the marker set 4 is detected by the camera, and the tilt detection mark 2 is recognized by this detection. Next, when the tilt detection mark 2 is recognized, a distance and an angle from the camera to the marker set 4 are calculated based on the positional relationship of the tilt detection mark 2. Further, an image when the marker set 4 is directly opposed is reconstructed based on the calculation result. Then, by detecting the center position of the virtual image pattern projected on the lens marker 1, the accurate angle of the marker set 4 is calculated. The detection method of the present invention can be used in reconstructing the image of the lens marker 1.

  In the marker set 4, examples of the two-dimensional code 3 include an AR marker and a QR marker. Examples of the AR marker include ARToolKit, Arteaga, Cybercide, ARToolKitPlus, and the like.

  Here, first, a VMP marker will be described as an example of the lens marker, and a structure of the VMP marker and general analysis of a virtual image pattern projected on the VMP marker will be described.

  FIG. 1 shows an example of the VMP marker. FIG. 1A is a plan view of the lens marker 1 as viewed from the viewing side, and FIG. 1B is a cross-sectional view of the lens marker 1 as viewed from the II direction in FIG. 1A. In FIG. 1B, a hatch representing a cross section is omitted for easy viewing. In FIG. 1, for convenience, arrow X is referred to as a width direction, arrow Y is referred to as a length direction, and arrow Z is referred to as a thickness direction.

  As shown in FIG. 1, the lens marker 1 includes a lens body 10. The lens body 10 has a plurality of lens portions 101 that are continuous in a plane direction (specifically, the width direction X in FIG. 1) on one surface, that is, on the upper surface side (viewing side) in FIG. In the lens body 10, a surface on which the lens unit 101 is continuous is also referred to as a lens surface. In addition, the lens body 10 has a plurality of detected portions 11 projected on the lens portion 101 on the other surface side, that is, on the lower surface side in FIG. In FIG. 1, the detected part 11 is a line extending along the length direction Y of the lens body 10, and a plurality of lines form a stripe pattern. The plurality of detected parts 11 are, for example, projected on the upper surface side of the lens body 10 as optically detectable images, and can be detected optically. The shape of the detected portion 11 is not particularly limited, and may be, for example, a shape in which points (dots) are aligned along the length direction Y. For example, the virtual image pattern projected on the surface of the lens unit 101 moves or changes depending on the inclination between the center axis and the optical axis of the lens marker 1. Specifically, in the case of the lens marker 1 shown in FIG. 1, a linear virtual image pattern extending in the length direction Y moves in the width direction X. Hereinafter, the width direction X is also referred to as a movement direction X.

  Next, FIG. 2 shows an image of the lens marker 1 in which a virtual image is projected on the lens surface. FIG. 2A is an image 20 obtained by photographing the lens surface of the lens marker 1 from the viewing side, and FIG. 2B is a schematic diagram illustrating the brightness of the detection line D in the image 20 of FIG. It is.

  As shown in an image 20 of FIG. 2A, a black virtual image pattern V on which the plurality of detection units 11 are projected appears on the lens surface of the lens marker 1. Then, usually, for example, the brightness of each pixel of the image 20 is extracted, a binarization process is performed based on a predetermined threshold, and a binary image represented by two gradations (black and white) is created. With the center line in the direction X as the detection line D, the position of the lens marker and the like are determined from the position of the virtual image pattern V on the detection line D. However, for example, when foreign matter such as dust adheres to the lens marker 1, images F1, F2, and F3 due to the foreign matter appear in addition to the virtual image pattern V as shown in FIG. For this reason, when the lightness at the detection line D is extracted from the image 20 of FIG. 2A, as shown in FIG. 2B, as in the case of the virtual image pattern V, the low lightness dependent on the foreign matter F1, F2, F3 is obtained. Is extracted. For this reason, even if the binarized image is created as described above, since a black portion is generated in the detection line D other than the virtual image pattern V, erroneous detection occurs, and accurate analysis based on the original virtual image pattern V is performed. It will be difficult. The problem of erroneous detection is the same, for example, when the detected portion 11 of the lens marker 1 is chipped, as described later, regardless of the adhesion of foreign matter.

  In such a case, as a method of avoiding the influence of the foreign matter, for example, a simple averaging process of lightness can be considered. A simple averaging process will be described with reference to FIG. FIG. 2B described above is a schematic diagram showing the brightness of only the detection line D of the lens marker 1. On the other hand, in the simple averaging process, the brightness of each pixel in the length direction Y orthogonal to the detection line D is extracted instead of the brightness of a point on the detection line D, and the average brightness in the length direction Y is calculated. I do. FIG. 3A shows this average lightness. However, as shown in FIG. 3A, even if the average brightness in the length direction Y is calculated, a low brightness that depends on the foreign substances F1, F2, and F3 remains. For this reason, even if the image is reconstructed based on the average brightness in FIG. 3A, a black image is formed at the positions of the foreign substances F1, F2, and F3 as shown in the reconstructed image 21 in FIG. 3B. Will remain. Therefore, even if a binarized image is further generated using the reconstructed image 21, it is apparent that a problem of erroneous detection still occurs.

  On the other hand, according to the present invention, as described above, in the extracting step, for each of the predetermined widths along the moving direction, the brightness of the image from one end to the other end in the direction orthogonal to the moving direction is determined. Extracting a level and calculating a first average value and a standard deviation (± σ) of the brightness levels of all pixels from the distribution of the number of pixels indicating each brightness level for each of the predetermined widths in the selection step. A range of a standard deviation centered on the first average value is selected as a brightness level area, and in the correction step, a second average value of the brightness level is obtained for each pixel corresponding to the brightness level area for each of the predetermined widths. Calculating, setting the second average value to the corrected lightness level, and reconstructing the corrected image of the lens marker based on the corrected lightness level of each of the predetermined widths in the reconstructing step. It is characterized by that. Thus, according to the present invention, for example, it is possible to cancel the influence of the adhesion of the foreign matter or the chipping of the detection target portion in the image, and thus, it is possible to solve the problem such as erroneous detection.

  Next, an embodiment of the present invention will be described with reference to the drawings. The present invention is not limited or limited in any way by the following embodiments. In the drawings, the same portions are denoted by the same reference numerals. In the drawings, for convenience of explanation, the structure of each part may be simplified in some cases, and the dimensional ratio of each part is not limited to the conditions in the drawing.

[Embodiment 1]
In the present embodiment, the correction method of the present invention will be specifically described.

  First, the correction method of the present invention will be described with reference to FIGS. 4 and 5 by taking an example in which a foreign substance is attached to a lens marker. FIG. 4 is a schematic diagram of the image 30 of the lens marker, in which the virtual image pattern V and the foreign matter image F appear.

  In the image obtaining step, an image 30 is obtained. The image 30 may be, for example, a raw image obtained by capturing the lens marker or a processed image obtained by performing a filter process such as a Gaussian filter.

  Next, in the extraction step, a brightness level from one end to the other end in the direction (length direction Y) orthogonal to the movement direction X is extracted for the image 30 for each predetermined width along the movement direction X. Specifically, as shown in FIG. 4, the area is divided along the moving direction X by a predetermined width W (W1 to Wn), and the brightness level from one end Y0 to Y1 in the orthogonal direction Y is determined for each area. Is extracted.

  For example, in the region W1, as shown in FIG. 4, since no black image appears, the brightness level of each pixel from the one end Y0 to the other end Y1 in the orthogonal direction Y is almost constant and high. Brightness level). Further, since the region Wv includes the virtual image pattern V as shown in FIG. 4, the brightness level of each pixel from one end Y0 to the other end Y1 in the orthogonal direction Y is substantially constant low brightness level (black brightness level). ).

  On the other hand, since the region Wf includes the foreign matter image F as shown in FIG. 4, the brightness level of each pixel from one end Y0 to the other end Y1 in the orthogonal direction Y is not constant. Here, the region Wf will be further described with reference to FIGS. As shown in the image 30 of FIG. 5A, the region Wf surrounded by the dotted line includes the foreign matter image F. Then, in this area Wf, when the brightness level of each pixel from one end Y0 to the other end Y1 in the orthogonal direction Y is extracted, a graph shown in FIG. 5B is obtained. In FIG. 5B, the X axis is a position from one end Y0 to the other end Y1 in the orthogonal direction Y, and can be said to be a pixel position from one end Y0 to the other end Y1. As shown in FIG. 5B, as the area Wf moves from one end Y0 to the other end Y1 in the orthogonal direction Y, the brightness level sharply decreases at the position of the foreign matter image F.

  Next, the selecting step calculates a first average value and a standard deviation (± σ) of the lightness levels of all the pixels from the distribution of the number of pixels indicating each lightness level for each of the predetermined widths, and A range of the standard deviation centered on the average value is selected as a lightness level area, and the correcting step calculates a second lightness level average value for pixels corresponding to the lightness level area for each of the predetermined widths. , The second average value is set as a corrected brightness level.

  Here, first, the region Wf will be further described. The relationship between the lightness level and the position of the area Wf shown in FIG. 5B can be represented, for example, by a histogram of FIG. 5C as a distribution of the number of pixels indicating each lightness level. In FIG. 5C, the X-axis is the brightness level, and the Y-axis is the number of pixels indicating each brightness level. For this distribution, first, the average value (first average value A1) and the standard deviation (± σ) of the brightness levels of all pixels are calculated, and the range of the standard deviation centered on the first average value A1 is calculated as the brightness. Select as the level area (Bf). Then, for the pixel corresponding to the brightness level area (Bf), a second average value A2 of the brightness level is calculated, and the second average value A2 is set as the corrected brightness level. As shown in FIG. 5C, in the region Wf, the peak F is observed even at a low lightness level indicating black due to the foreign matter image F. However, by performing the above-described processing, the region Wf is canceled. A corrected brightness level of Wf is obtained. In other words, the corrected average brightness level of the area Wf is not the first average value A1 affected by the foreign matter image but the second average value A2 canceling the influence of the foreign matter image.

  On the other hand, as described above, in the region W1, the region Wv, and the like, as described above, the brightness level of each pixel from one end Y0 to the other end Y1 in the orthogonal direction Y indicates a substantially constant brightness level. Therefore, in the selecting step, the first average value and the standard deviation (± σ) are calculated, the brightness level area is selected, and the calculated second average value is set as the corrected brightness level in the correcting step. Since there is no image to be canceled which causes erroneous detection, as a result, the first average value and the corrected lightness level (second average value) are substantially the same.

  Then, in the reconstructing step, the corrected image of the lens marker is reconstructed based on the corrected brightness level of each of the predetermined widths. As described above, for each of the predetermined widths, a corrected lightness level in which the image causing the erroneous detection is canceled has been obtained. Therefore, if the corrected image is reconstructed from these corrected lightness levels, the foreign matter image can be obtained. A corrected image that does not appear can be obtained.

  When the image 20 shown in FIG. 6A, which is the same as FIG. 2A, is processed by this method, the corrected lightness level of the image 20 is expressed as shown in FIG. 6B. Then, when the corrected image is reconstructed from the corrected brightness level of the image 20, the foreign matter images F1, F2, and F3 are canceled in the corrected image 22, and only the virtual image pattern V appears in the corrected image 22, as shown in FIG. State. For this reason, even if a binarized image is created from the corrected image 22 and the virtual image pattern V is analyzed, erroneous detection caused by the foreign matter F1, F2, F3 can be avoided.

  Next, an example in which the detection portion of the lens marker is chipped will be described with reference to FIG. FIG. 7A is a schematic diagram of the image 40 of the lens marker. In the virtual image pattern V, a chip F appears.

  In this case, the processing is performed in the same manner as described above. As shown in FIG. 7A, when the brightness level of each pixel from one end Y0 to the other end Y1 in the orthogonal direction Y is extracted in the area Wf where the chip is generated, as shown in the graph of FIG. become. In FIG. 7B, the X axis is a position from one end Y0 to the other end Y1 in the orthogonal direction Y, and can be said to be a pixel position from one end Y0 to the other end Y1. As shown in FIG. 7B, as the area Wf moves from one end Y0 to the other end Y1 in the orthogonal direction Y, the brightness level sharply increases at the position of the notch F.

  The relationship between the lightness level and the position of the area Wf shown in FIG. 7B can be represented as a distribution of the number of pixels indicating each lightness level, for example, by a histogram in FIG. 7C. In FIG. 7C, the X-axis is the brightness level, and the Y-axis is the number of pixels indicating each brightness level. For this distribution, first, the average value (first average value A1) and the standard deviation (± σ) of the brightness levels of all pixels are calculated, and the range of the standard deviation centered on the first average value A1 is calculated as the brightness. Select as the level area (Bf). Then, for a pixel corresponding to the brightness level area (Bf), a second average value A2 of brightness levels is calculated, and the second average value is set as a corrected brightness level. As shown in FIG. 7C, in the region Wf, the peak F is seen even at a high lightness level indicating white due to the chipping F. However, the region Wf in which the peak F is canceled by performing the above-described processing. Will be obtained. In other words, the corrected average brightness level of the area Wf is not the first average value A1 affected by the missing F, but the second average value A2 canceling the effect of the missing F.

  Therefore, in the reconstructing step, if the corrected image of the lens marker is reconstructed based on the corrected brightness level of each of the predetermined widths, it is possible to obtain a corrected image in which a chip does not appear.

  In the extraction step, the predetermined width is not limited at all, and can be set, for example, by the number of pixels. The lower limit of the number of pixels is, for example, 1 pixel, and the upper limit is, for example, 3 pixels, 5 pixels, 10 pixels, and 50 pixels. Note that the upper limit of the number of pixels is not particularly limited, and can be appropriately determined depending on, for example, the resolution of a device such as a camera used for detection. As the number of pixels having the predetermined width is set to be relatively small, the influence of foreign matter and chipping can be more effectively eliminated.

  In the extraction step, the gray level of the lightness level is not particularly limited. For example, when the calculation is performed with 8 bits, the gray level is 256 gray levels from 0 to 255.

  In the selection step, since the width of the brightness level region can be adjusted, for example, the standard deviation may be further multiplied by a coefficient.

  According to the present invention, in the predetermined width, the number of pixels in the Y direction where the brightness is extremely changed in the distance from one end Y0 to the other end Y1 in the orthogonal direction Y is, for example, 1/5 or less, preferably 1 or less. In the case of / 10 or less, the above-described foreign matter and chipping can be effectively canceled.

  The present invention is preferably applied as an image correction when, for example, a lens marker is placed on an object and its posture and position are analyzed. According to the present invention, even if foreign matter is attached or chipped on the lens marker, the influence of the foreign matter can be avoided and analysis can be performed. Lens markers that have been generated and could not be used because of insufficient accuracy can also be used.

[Embodiment 2]
In the present embodiment, the correction device of the present invention will be specifically described. According to the correction device of the present embodiment, for example, the correction method of the first embodiment can be executed. In the present embodiment, the description of Embodiment 1 can be referred to unless otherwise specified.

  FIG. 9 is a block diagram illustrating an example of the correction device according to the present embodiment. The correction device 50 includes a lens marker image acquisition unit 51, a brightness level extraction unit 52, a brightness level area selection unit 53, a brightness level correction unit 54, and a corrected image reconstruction unit 55. The correction device 50 further includes, for example, a storage unit. The correction device 50 may be, for example, a single device including the respective units, or a device to which the respective units can be connected via, for example, a communication network. The correction device 50 is also referred to as, for example, a correction system.

  The lens marker image obtaining unit 51 obtains an image of the surface (lens surface) of the lens marker. According to the image acquiring unit 51, the image acquiring step in the first embodiment can be executed. The correction device 50 itself may, for example, capture an image of the lens marker itself. In this case, the image acquisition unit 51 is, for example, a camera such as a CCD. Further, the correction device 50 may acquire, for example, an image of the lens marker captured by an external device. In this case, the image acquisition unit 51 can acquire the image from the external device, for example, by connecting to the external device. The correction device 50 and the external device may be connectable, for example, via a communication network. The communication may be, for example, wired or wireless. The external device includes, for example, the camera.

  The brightness level extraction unit 52 extracts, for each of the images, a brightness level from one end to the other end in a direction orthogonal to the moving direction for each predetermined width along the moving direction. According to the extraction unit 52, the extraction step in the first embodiment can be executed.

  The selection unit 53 calculates a first average value and a standard deviation (± σ) of the brightness levels of all pixels from the distribution of the number of pixels indicating each brightness level for each of the predetermined widths, and calculates the first average value. The standard deviation range around the center is selected as the brightness level area. According to the selecting unit 53, the selecting step in the first embodiment can be executed.

  The correction unit 54 calculates a second average value of the brightness level for the pixels corresponding to the brightness level region for each of the predetermined widths, and sets the second average value as the corrected brightness level. According to the correction unit 54, the correction process in the first embodiment can be executed.

  The reconstructing unit 55 reconstructs a corrected image of the lens marker based on the corrected brightness level of each of the predetermined widths. According to the restructuring unit 55, the restructuring step in the first embodiment can be executed.

  The storage unit stores, for example, information acquired by each unit, imaging conditions of the image, conditions of the predetermined width, and the like.

  FIG. 10 illustrates a block diagram of a hardware configuration of the correction device 50. The correction device 50 includes, for example, a CPU (central processing unit) 102, a memory 103, a bus 104, an input device 105, a display 106, a communication device 107, a storage device 108, and the like. The components of the correction device 50 are mutually connected via the bus 104 by, for example, respective interfaces (I / F).

  The CPU 102 performs overall control of the correction device 50. In the correction device 50, for example, the program 109 (including the program of the present invention) is executed by the CPU 102, and various kinds of information are read and written.

  The bus 104 connects between functional units such as the CPU 102 and the memory 103, for example. The bus 104 can be connected to, for example, an external device. The correction device 50 can be connected to a communication network by a communication device 107 connected to the bus 104, and can be connected to the external device via the communication network. The communication line network is not particularly limited, and a known communication line network can be used. Specific examples include an Internet line, a telephone line, a LAN (Local Area Network), and WiFi (Wireless Fidelity). The method of connecting to the external device is not particularly limited, and a known method can be used, and specific examples thereof include USB, GiGE, CoaPress, and CameraLink.

  The memory 103 includes, for example, a main memory, and the main memory is also referred to as a main storage device. When the CPU 102 performs processing, the memory 103 reads various operation programs 109 such as the program of the present invention stored in an auxiliary storage device described later, and the CPU 102 receives data from the memory 103 and Then, the program 109 is executed. The main memory is, for example, a RAM (random access memory). The memory 103 further includes, for example, a ROM (read only memory).

  The storage device 108 is also referred to as a so-called auxiliary storage device for the main memory (main storage device), for example. The storage device 108 includes, for example, a storage medium and a drive that reads from and writes to the storage medium. The storage medium is not particularly limited. For example, the storage medium may be a built-in type or an external type, and may be HD (hard disk), FD (floppy (registered trademark) disk), CD-ROM, CD-R, CD-RW, MO, Examples include a DVD, a flash memory, and a memory card, and the drive is not particularly limited. As the storage device 108, for example, a hard disk drive (HDD) in which a storage medium and a drive are integrated can be exemplified. The storage device 108 is, for example, the storage unit, and may store information obtained by each unit as described above.

  The correction device 50 may further include an input device 105, a display 106, and the like. The input device 105 is, for example, a touch panel, a keyboard, or the like. The display 106 includes, for example, an LED display, a liquid crystal display, and the like.

  The result obtained by the correction device 50 may be output to, for example, the display 106 or may be output to an external device via the communication network.

  In addition, the present invention may be, for example, a determination device that determines a position of a virtual image pattern of a lens marker. In this case, the determining device of the present invention determines the position of the virtual image pattern from the correcting device of the present invention and a binarizing processing unit that binarizes the obtained corrected image. And a position determining unit. According to the correction device of the present invention, for example, as described above, a corrected image in which foreign matter and chipping have been canceled is obtained. Therefore, except for using the corrected image, a conventional and a binarization processing unit and a position determination unit , The position of the virtual image pattern can be determined.

[Embodiment 3]
The program of the present embodiment is a program that can execute the correction method of the above embodiment on a computer. Alternatively, the program of the present embodiment may be recorded on, for example, a computer-readable recording medium. The recording medium is not particularly limited, and includes, for example, the above-described storage medium.

  Although the present invention has been described with reference to the exemplary embodiments, the present invention is not limited to the above exemplary embodiments. Various changes that can be understood by those skilled in the art can be made to the configuration and details of the present invention within the scope of the present invention.

  As described above, according to the present invention, for example, even when a plurality of virtual images (for example, black lines) occur over a plurality of lens portions in the VMP marker, it is possible to accurately determine the center of the virtual image pattern. it can.

1 Lens marker 20, 30, 40 Image 21, 22 Corrected image 2 Tilt detection mark 3 Two-dimensional code 4 Marker set 50 Correction device 102 CPU
103 memory 104 bus 105 input device 106 display 107 communication device 108 storage device 109 program

Claims (6)

Obtaining an image of the lens marker,
Brightness level extraction process,
Brightness level area selection process,
Including a brightness level correction step in the pre-brightness level area, and a reconstructed step of a corrected image based on the corrected brightness level
The lens marker is
On one of the surfaces, a marker in which a virtual image moves along at least one movement direction according to an inclination with respect to the optical axis,
The image obtaining step includes:
Obtaining an image of the surface of the lens marker,
The extracting step includes:
For the image, for each predetermined width along the movement direction, extract the brightness level from one end to the other end in the direction orthogonal to the movement direction,
The selecting step includes:
For each of the predetermined widths,
From the distribution of the number of pixels indicating each lightness level, a first average value and a standard deviation (± σ) of the lightness levels of all pixels are calculated,
Selecting a range of a standard deviation centered on the first average value as a brightness level area;
The correcting step includes:
For each of the predetermined widths,
For a pixel corresponding to the brightness level area, a second average value of the brightness level is calculated, and the second average value is set as a corrected brightness level;
The reconstruction step includes:
A method for correcting a lens marker image, comprising reconstructing a corrected image of the lens marker based on the corrected brightness level of each of the predetermined widths. The correction method according to claim 1, wherein in the extracting step, the predetermined width is set to a width of 1 to 5 pixels. Lens marker image acquisition unit,
Brightness level extractor,
Lightness level area selection section,
Including a brightness level correction unit in the brightness level region, and a reconstructed unit of a corrected image based on the corrected brightness level,
The lens marker is
On one of the surfaces, a marker in which a virtual image moves along at least one moving direction according to an inclination with respect to the optical axis,
The image acquisition unit,
Obtaining an image of the surface of the lens marker,
The extraction unit includes:
For the image, for each predetermined width along the movement direction, extract the brightness level from one end to the other end in the direction orthogonal to the movement direction,
The selection unit includes:
For each of the predetermined widths,
From the distribution of the number of pixels indicating each lightness level, a first average value and a standard deviation (± σ) of the lightness levels of all pixels are calculated,
Selecting a range of a standard deviation centered on the first average value as a brightness level area;
The correction unit,
For each of the predetermined widths,
For a pixel corresponding to the brightness level region, a second average value of the brightness level is calculated, and the second average value is set as a corrected brightness level;
The reconstructing unit,
An apparatus for correcting a lens marker image, wherein a corrected image of the lens marker is reconstructed based on the corrected brightness level of each predetermined width. The correction device according to claim 3, wherein the extraction unit sets the predetermined width to a width of 1 to 5 pixels. A program for causing a computer to execute the lens marker image correction method according to claim 1 or 2. A computer-readable recording medium on which the program according to claim 5 is recorded.