JP2009025078A5 - - Google Patents

Description

Measurement method of luminance unevenness of uniform light source for shading correction

       The present invention is a lens applied to a camera used for image processing and the like, and is basically based on the condition that the image capturing has a detection angle like a reduction system or an enlargement system, and the spherical aberration of the lens. Shading correction applied to correct shading of an image processing apparatus applied to improve the accuracy of image measurement, etc., from distortion (hereinafter referred to as shading) in which the brightness level of the brightness caused by the curve is inversely curved. Data and a measurement method for creating a shading correction device.

       To create shading correction data in a conventional image processing apparatus, correction values are extracted by a uniform light source measured at a directly facing position or a method for guaranteeing uniformity for an integrating sphere.

As a method other than the uniform light source method described in 0002, a structure for which a patent application has already been filed in Patent Document 1 or a method of applying shading correction using an application method and applying it to image processing has been described.
Japanese Patent Application No. 2006-243533

       Luminance distortion caused by the spherical aberration of the lens called shading based on the case where the characteristics of the image capturing lens to be applied to image processing are other than parallel beam capturing such as a reduction system and an enlargement system, that is, the lens It is applied to correct the situation where the vicinity of the core becomes the maximum brightness and becomes darker as it reaches the periphery. This shading correction is the same as the lens setting under the conditions applied to actual image processing, and the same distance and angle of view are important points, and it was assumed that the distance and angle of view were the same. To deal with this condition, another factor of detection angle cannot be ignored.

In the measurement of uniform light source applied to shading correction, referencing the uniform light source is also one of the causes because it is affected by passing through the lens. It is important to measure the characteristics.

       The object of the present invention is that it is basically impossible for all pixels to face each other when detecting the luminance of a uniform light source, and the detection angle by a lens that changes for each corresponding pixel other than the center of the lens can be ignored. Can not. In addition, since there is a difference between individual lenses, it is assumed that an apparatus that is basically applied to measurement is applied.

       When the uniform light source to be applied is a system with ideal uniformity that can confirm that the uniform light source, which is caused by the characteristics of the lens, changes the level change of the inverted barrel shape concentrically In the shading correction process, which is a pre-process, it is not always necessary to target all pixels. This is not only due to the problem of processing time, but also due to characteristics such as fluctuations in the luminous flux of the light source. Due to the characteristics of the filter of the applied light source, it may not be possible to expect sufficient surface accuracy, so vertical and horizontal lines and diagonal lines are used as the line segment based on the pixel that is the core of the lens of the basic measurement angle of view. Alternatively, a line segment drawn at 45 degrees between vertical and horizontal lines, and processing is performed by comprehensively processing the peripheral pixels of these line segments to obtain line segment data, and a position where a plurality of these line segments are adapted Of pixels corresponding to Application of the method for deploying write data to the surface data as a base is also possible, even to meet the exclude purposes unnecessary pixel components from the measurement target object data that do not apply to the processing as a measurement result.

       In the measurement of unevenness of a uniform light source applied to shading correction, the measurement target line segment described in 0007 is used and divided into four quadrant angles with the vertical and horizontal lines passing through the center of the lens as boundaries. Using the data of these line segments in the quadrant, the mura value is measured for each quadrant corresponding to the pixel data corresponding to the lens core, and finally the data in the four quadrants is used to determine the total mura value. What to do.

       Regarding the detection of unevenness of the uniform light source, together with the data detected from the pixels corresponding to the line segments divided into the four quadrants described in 0007, the line is similarly rotated by rotating 180 degrees around the midpoint of the quadrant for each quadrant. Comparing the data detected from the pixels corresponding to the minute, either the data obtained by rotating or the quadrant data extracted in 0008 is aligned as the reverse position and the comparison process is performed to make the murarator in the quadrant. In general, the data of all quadrants is used to make the unevenness of the uniform light source.

       The final purpose of shading correction is to detect the shading characteristics of the lens. Furthermore, this characteristic can be used to correct the uniformity of the uniform light source. Eliminates the uniform light source from the physical conditions of the correction process, and can move the same position two-dimensionally in front of the lens. In addition, it has a mechanism that can face the lens or pixel, and can be applied to all pixels or 0007. It is equipped with a structure that automatically performs fine adjustment so that the brightness level of the target pixel for alignment or direction alignment corresponding to the line segment to be maximized, and the target pixel data is extracted and transmitted through the lens. The luminance attenuation rate with respect to the pixel position is extracted, and this data is detected data for all pixels or a line designated by 0007. Detection pixel data and performs a comparison operation, for the application of uniform light source to the correction value and then after shading correction of unevenness of the uniform light source according to.

       It is easy to create and adjust a uniform light source that can be applied to shading correction processing to support techniques that enable high-precision measurement of image processing, including image measurement. Since maintenance of processing accuracy can be established, it has become possible to perform high-speed image processing at low cost.

       The measurement method of the present invention will be described with reference to examples.

       A situation in which the present invention is applied as a method for extracting a tendency of shading characteristics will be described with reference to the drawings.

       FIG. 1 shows a transition of a scanning line of an image of a basic measurement angle of view corresponding to the center of a lens near the center when a uniform light source is referred to, or a transition of a luminance level on the horizontal axis. It is a horizontal line and shows uniformity, and this line is after shading correction. On the other hand, there is a line that shows the maximum luminance near the center, but it can be judged that this line basically shows a tendency of upside-down type, and the influence of passing through the lens is prominent. Furthermore, the basic measurement angle of view shown in FIG. 2 is divided into four equal parts with the lens core as a boundary, and divided into four quadrants, and a feature is extracted for each. 2-1, 2-2, 2-3, and 2-4 indicate the split angle of view divided into four quadrants, and 2-5 indicates a split angle of view equivalent to the area divided into four quadrants. The center is designated as the position of the lens core. Since 2-5 may be the part with the least change, this part is regarded as uniform and compared with the four quadrants. Shading correction is performed in advance with reference to 2-5. A correction value is extracted. Data of each pixel in each quadrant is extracted by applying this correction value to four quadrants. FIG. 3 shows the state of the small angle of view when performing processing that includes peripheral pixels, and FIG. 4 shows the relationship between the position of the lens core of FIG. 3 and the small angle of view. . This usage includes kernel setting in shading correction and setting of a frame of peripheral pixels for a line segment. In FIG. 5, since the measurement values for all pixels are complicated, the condition in the case where the luminance level is recognized to decrease so that the uniform light source draws a concentric circle around the center of the lens is shown. And, as a representative measurement point, I will explain with the content only for pixels related to the diagonal line drawn with the basic measurement angle of view, but this object was drawn in the vertical and horizontal line segments drawn from the lens core and this vertical and horizontal It is also possible to target a line segment drawn at 45 degrees from the line segment, and there are also conditions for comprehensive processing of peripheral pixels for the line segment, as described in FIGS. 3 and 4. . A relative ratio is obtained by how 2-1, 2-2, 2-3, and 2-4 are detected with respect to 2-5 in a situation set to be uniform by performing shading correction. 2-1 is shown as a representative of the small angle of view in FIG. Reference numeral 5-1 denotes a lens core, which is a central part of the basic measurement field angle, and 5-2 is a peripheral part of the basic measurement field angle. FIG. 6 is a graph for explaining the situation of the luminance level of the pixels detected in 5-1 to 5-2. FIG. 5 shows the situation from the central part of 5-1 which is the core of the lens at the position of FIG. 5 to the peripheral part of 5-2, and the situation where the luminance level decreases to the peripheral part of 5-2. By extracting the brightness levels in all four quadrants of this situation, it is possible to grasp the situation from the lens core to the four corners of the basic measurement angle of view.

       The present invention will be described with reference to a method of creating a correction coefficient for all pixels for extracting a light source correction coefficient and correcting light source unevenness on data.

       Data for each quadrant is extracted using the method of processing divided into four quadrants indicated by 0007, and data is similarly extracted by the method described in FIG. FIG. 7 shows a sub-view angle of 2-1 as an example, but the same processing is performed for the sub-view angles of the other three quadrants. The oblique angle of view is the target, and the same measurement is performed by rotating 180 degrees around the midpoint of the angle of view as shown in 7-1. In this case, the pixel corresponding to the core of the lens is a position corresponding to the peripheral portion, and the same condition of the lens is used and the conditions are reversed. FIG. 8 is a graph of luminance levels represented by 2-1 extracted by the method of FIG. 8-1 is a luminance graph at the basic measurement angle of view, and 8-2 shows a case where the luminance graph extracted in FIG. 7 has high uniformity. 8-3 assumes a case where the actually extracted luminance graph is in such a situation. The deviation in this case is to compare and contrast 8-1 and 8-3, but 8-3 is reversed to the left or right, or 8-1 is reversed to calculate the correlation ratio. We will seek change. Similarly, the overall state can be grasped by performing processing in other quadrants. This shows the relationship between the lens described in FIG. 9 and the imaging target such as the uniform light source. In the imaging, the relationship between the lens 9-1 and the imaging target shows that the uniformity is measured when the uniform light source measures directly. For those evaluated to be high, it can be assumed that the center of the lens shows the maximum brightness level with the light beam facing 9-2, but the brightness level at the periphery of the angle of view as in 9-3. This indicates that the luminance level decreases from 9-2 to 9-3 as it reaches the periphery. This method is also applied to brightness data from vertical and horizontal line segments that pass through the core of the lens, which is another line segment, a 45-degree line segment to this vertical and horizontal line segment, or pixels that encompass the surrounding pixels of the line segment. Thus, since the concentric circles are drawn, it is possible to develop the correction values corresponding to all the pixels of the basic measurement angle of view.

       A technique of directly detecting the luminance level due to the spherical aberration of the lens, which is the gist of the present invention, and applying it to adaptation as a shading correction value or correction for a uniform light source will be described.

       A two-dimensional drive mechanism capable of mechanical scanning equivalent to the surface of the uniform light source is removed at the same position, and the head is swung. The target mechanism is provided with a point light source (hereinafter referred to as a standard light source) whose luminance value is stable for a long time. The light emission position of the standard light source is set so as to be the same position as the uniform light source, and these mechanisms are provided with a driving device. At this time, the setting of the lens is set to a distance and an angle of view applied to actual image processing, and is fixed in a direction facing the uniform light source. For purposes other than this usage, it is possible to detect the attenuation rate based on the capture angle by directing the lens toward the standard light source as shown in FIG. FIG. 11 illustrates, as a representative example, that data collection corresponding to pixels on a diagonal line is performed, but the line segment described in 0007 is another target. With respect to this line segment, a position and an adaptive angle with respect to the pixel are obtained in advance, and the position with respect to the pixel moves to a position where it can be developed in the same manner as the surface of the uniform light source, using a two-dimensional drive mechanism, Similarly, a light source is irradiated at a detection angle corresponding to a pixel using a mechanism for swinging. In FIG. 12, although drawn as the target pixel of 12-1, the address of the basic measurement angle of view where the data detected as the actual target pixel is arranged is shown. On the other hand, although the detection by the 13-1 kernel in FIG. 13 has the same correspondence, the pixel toward the center of one pixel is drawn for reference, and the surrounding 3 pixels × 3 pixels are shown as a representative. Has a structure that automatically drives the swing position so that the target pixel has the maximum luminance level, and has a function to hold the detected luminance level at the address of the basic measurement angle of view. It is. The 14-1 standard light source position A illustrated in FIG. 14 is in a position shifted from the surface of the uniform light source in the illustrated state, but the actual position is required to be the same as the surface of the uniform light source. However, there is also a method in which the attenuation rate is obtained from the value of the 14-2 standard light source position B and the distance at the same brightness level by separating the distance at the directly facing position as illustrated in 14-1. Luminance data extracted by this method can be applied to the correction value of the uniform light source by comparing with the pixel value of the same address of the uniform light source. It is also possible to develop the plane data by creating correlation data. This data can also be applied as a shading correction value.

       Conventionally, there was no guideline for correcting the lens with reference to the uniform light source, but since a method for indicating a state where the uniform light source is uniform is shown, it is possible to display the accuracy of the uniform light source. became. It became important for the creation of correction values when high accuracy such as image measurement is required, because the data for the condition setting can be compared and dealt with by creating data that matches each condition.

It shows the signal at the basic measurement angle of view, and displays before and after the shading correction processing of the center line of the horizontal line (horizontal line) of the angle of view. The figure explaining the subdivision picture with respect to a basic measurement angle of view. The figure which shows the promotion part in the basic measurement image of a small angle of view. The figure which shows the way of thinking of the center part of the promotion part of a small angle of view. The figure which shows the state of the diagonal line drawn with respect to the representative small drawing of a basic measurement view angle. The figure which displayed the signal generated in FIG. The figure which shows the condition which rotates the midpoint of a small angle of view by 180 degree | times. The figure which shows the signal on the diagonal of a small split angle of view and the measurement small split angle of view after rotation. The figure which shows the condition of the detection angle of a camera lens and a uniform light source. The figure which turned the camera and the lens to the corner | angular part which is a measurement point. The figure which drawn the diagonal line at the basic measurement angle of view. The figure which shows the position of an object pixel in the measurement point of a corner | angular part. The figure which shows the position of the measurement kernel of a corner | angular part. The figure which shows the relative relationship which changes and changes a measuring object into a standard light source from a uniform light source.

2-1 A frame showing the split angle of view.
2-2 A frame showing the small angle of view.
2-3 A frame showing the split angle of view.
2-4 A frame showing the split angle of view.
2-5 A frame indicating a small angle of view from which a shading correction coefficient is to be extracted.
3-1 shows the split angle of view encompassing the assisted part, which is a part of 3-1.
The promotion part which is 3-5 spread is shown centering on the center part.
4-1 The promotion part which is a part of the spread is shown.
5-1 This corresponds to the center of the basic measurement angle of view.
5-2 The peripheral part corresponding to the corner of the basic measurement angle of view is shown.
6-1 The graph of the luminance level detected from 5-1 to 5-2 in FIG. 5 is shown.
7-1 A diagram showing that the basic measurement angle of view is rotated by 180 degrees at the center of the small angle of view.
8-1 A graph of the luminance level generated in FIG.
8-2 A diagram showing a case where a uniform light source is uniform in a graph of a luminance level when measurement is performed by rotating a small angle of view by 180 degrees.
8-3 A graph showing an actual luminance level when measurement is performed by rotating the small angle of view by 180 degrees.
9-1 A diagram showing the position and direction of a lens for measurement.
9-2 is a diagram showing that a uniform light source is directly facing.
9-3 A diagram showing a light beam directly facing a uniform light source in the peripheral portion.
FIG. 9-4 is a diagram showing a light beam traveling from the periphery to the lens.
10-1 A view of directing the lens toward the periphery.
12-1 The figure which shows the corner | angular part used as a detection target, and the position of a pixel.
13-1 The figure which shows the corner | angular part used as a detection target, and the position of a measurement kernel.
14-1 The lens of the camera is facing the uniform light source.
14-2 A diagram in which a standard light source is arranged at a directly facing position instead of a uniform light source.
14-3 A diagram in which a standard light source is arranged in the peripheral portion instead of a uniform light source, and a direction is directed to the center of the pixel to be arranged or the kernel.

Claims (3)

In an image processing apparatus in which a lens having characteristics of a reduction system or an enlargement system is applied to a camera or the like that is an image capturing device, the brightness level of brightness caused by the spherical aberration of the lens is a distortion (hereinafter referred to as a curved shape). As the measurement data of the uniform light source applied to the correction of the shading), the basic situation in the entire captured image corresponding to the basic measurement angle of view is the pixel corresponding to the center, or the lens core (hereinafter, Both are referred to as the lens core.) The situation where the luminance level decreases in a concentric manner with the pixel corresponding to the center point as the midpoint is a theoretically correct uniform light source, and it corresponds to the luminance level of the pixel at the lens core. The correlation with the luminance level of each pixel is calculated as a shading correction value, and its inverse is applied for correction. Is also a measurement of unevenness inclusive of deviation due to the spherical aberration of the lens with respect to the unevenness of the uniform light source. However, since processing of measurement values for too many pixels becomes complicated in the case of targeting all pixels, and there is a possibility that unstable elements for processing often occur, there is a necessity to avoid this, As a precondition, when it is confirmed that the level of luminance in the structure of the uniform light source is concentrically reversely curved with the center of the lens as the midpoint, the vertical and horizontal lines passing through the center of the lens are used as the boundary. Divided into two quadrant angles and placed in 45 degree lines corresponding to the vertical and horizontal lines and diagonal lines of the basic measurement angle of view or half of the vertical and horizontal lines drawn from the center of the lens in each quadrant Applying a method for processing the data of the pixel to be processed and taking into account the average value of neighboring pixels for these line segments, and further using the data based on the plurality of conditions for the lens Phase to the core of Uniform light source luminance unevenness measurement method having a function of an overall non-uniformity value by utilizing the final four quadrant data subjected to uneven metering per each quadrant corresponding to the data of a pixel to be The measurement is performed by dividing the image into four quadrants as described in claim 1 and extracting the luminance unevenness in all pixels or various line segments, and the quadrant center for each quadrant is rotated for the extracted data. The same measurement is performed in a state rotated 180 degrees as the center, and the purpose is to extract data by using the same lens and inverting the core and peripheral or corners of the lens. Thus, the brightness unevenness measurement method for a uniform light source having a function of aligning the peripheral portion and the lens core values and extracting the uniform light source capture unevenness rate for the two types of extracted data. As a measurement method for detecting the transmission unevenness value of only the lens, which is the basis of the shading correction value, or by correcting the shading correction value data described in claim 1 for this data, the correction value for the uniformity of the uniform light source is obtained. It is also used as an extraction function, and its structure maintains the relationship between the distance between the lens and uniform light source, which is the setting condition for performing shading correction processing, and the angle of view, and the same light source with the uniform light source removed. A point light source (hereinafter referred to as a standard light source) that is stable for a long time is mounted on the mechanism with a mechanism that can change the position and irradiation angle mechanically at the position, and detection is required in advance. The position and irradiation angle corresponding to the pixel to be specified are specified, and the standard light source is moved to that position and irradiation angle. It has a function of moving both mechanisms so as to maximize the value and holding the value, and extracts the luminance data of the measurement position of all pixels or the line segments described in claim 1 and claim 2 and shading. A method for measuring the brightness of a uniform light source, which has a function for the purpose of extracting unevenness values of the uniform light source as a correction value.