JP6806587B2 - MTF measuring device and its program - Google Patents

Description

The present invention relates to an MTF measuring device for measuring an MTF (Modulation Transfer Function) indicating the spatial frequency characteristics of an imaging system and a program thereof.

The greatest feature of high-resolution television is high spatial resolution, and the resolution characteristics of the camera are important. As a resolution measurement of the current high-definition camera, a method using a test chart in which square waves having a plurality of spatial frequencies are spatially arranged is known. This test chart is generally used as a test chart (ITE high-definition inmegacycle chart: see Fig. 9) provided by The Institute of Image Information and Television Engineers (ITE). Be done.
As shown in FIG. 9, the inmegacycle chart is a rectangle in which black and white vertical stripes corresponding to video frequencies of 1 MHz to 36 MHz are arranged, and 800 TVL / ph (27.7 MHz) vertical stripes are arranged in the center of the chart and four corners. It has a wave pattern.

In the method using the immegacycle chart, it is common to read the CTF (Contrast Transfer Function) representing the spatial frequency characteristic of the modulation degree of the square wave from the waveform monitor.
The method using this inmegacycle chart has the convenience of being able to visually read the CTF from the waveform monitor, but there is ambiguity that the amplitude fluctuates due to the influence of the sampling phase. Further, in general, since the resolution characteristics are different between the center and the periphery of the lens to be measured, the current situation is that only the rectangular wave response at the center of the chart corresponding to 800 TVL / ph is measured.

Further, in the method using the inmegacycle chart, it is necessary to accurately frame the imaging angle of view to the chart size in order to obtain the desired spatial frequency characteristics. However, for example, a 4K / 8K camera often uses a wide-angle lens, which requires a large-sized inmegacycle chart, which is unrealistic.

Therefore, a slanted-edge method (patent document 1 to 3, non-patent documents 1 to 3) has been proposed in which the spatial frequency characteristic (MTF) is measured using a chart including an edge image instead of the inmegacycle chart. See Patent Document 1). The Slanted-edge method is a method in which the chart size is relatively small and framing is not required. A slightly inclined edge image on the chart is imaged, and the MTF is calculated from the spread of the edge.

In this slanted-edge method, first, a rectangular region of interest (ROI [Region Of Interest]; see FIG. 10) including an edge is selected from a chart image obtained by capturing a chart. Then, the Slanted-edge method detects an edge from the ROI by an algorithm based on ISO12233 (see Non-Patent Document 1) or the like.

Then, in the Slanted-edge method, each pixel of the ROI is projected along the inclination of the edge on the projection axis (x-axis) divided at equal intervals by the sub-pixels. Then, in the Slanted-edge method, the average value of the pixel values of a plurality of pixels projected on each division is obtained, and the oversampled edge spread function (edge profile) is obtained (in the case of ISO12233, it is oversampled four times). .. Further, in the Slanted-edge method, the edge spread function is differentiated to calculate the line spread function, and the line spread function is Fourier transformed to obtain the absolute value to obtain the MTF.

Normally, in the Slanted-edge method, as shown in FIG. 11A, a region (a region that includes an edge that is a vertically long rectangle and slightly inclined in the long side direction is included in the center of the chart image obtained by capturing the chart CH. ROI) is set, and each pixel of ROI is projected on the projection axis (x-axis). Further, in the Slanted-edge method, as shown in FIG. 11B, when the region (ROI) is set so as to include an edge that is a horizontally long rectangle and slightly inclined in the long side direction around the center, the ROI is set. Rotate it 90 degrees and project the pixels on the same projection axis (x-axis) as in FIG. 11A.

Further, in the slanted-edge method disclosed in Patent Document 2, as shown in FIG. 12, MTF is measured by a chart (arrow wheel chart) CH in which the inclination of the edge exists in multiple directions (12 directions in FIG. 12). In this case, since the rotation angles of each ROI (R _{1 to} R ₁₂ ) are known, the pixels are projected on the same projection axis (x axis) by rotating the ROI by that angle.
In this way, conventionally, the short side of the rectangular ROI is rotated so as to be parallel to the projection axis (x axis), and the MTF is measured.

Japanese Unexamined Patent Publication No. 2010-197201 JP-A-2010-237177 JP-A-2015-94701

ISO 12233: 2014, “Photographing --Electronic still picture imaging --Resolution and spatial frequency responses”

In the conventional method of setting a rectangular ROI and measuring the MTF, the position of the ROI may be limited when another pattern exists in the vicinity of the ROI.
For example, as shown in FIG. 13, when measuring the MTF around the lens with a chart (arrow chart) CH including edges in multiple directions, the ROI can be set correctly as shown in FIG. 13 (a). However, when measuring the MTF at the center of the chart, there is a problem that the MTF cannot be measured correctly because other edges are included in the ROI as shown in FIG. 13 (b).

Therefore, an object of the present invention is to provide an MTF measuring device and a program thereof capable of measuring MTF by setting a region including an edge in an arbitrary shape in the slanted-edge method.

In order to solve the above problems, the MTF measuring device according to the present invention is an MTF measuring device that measures MTF representing the spatial frequency characteristics of an imaging system by using charts having different contrasts at boundaries, and includes ROI setting means and ROI setting means. A configuration including a region extraction means, an edge extraction means, an edge inclination detection means, an image rotation means, a function approximation means, a projection information generation means, a projection pixel value averaging means, and a frequency characteristic calculation means. did.

In such a configuration, the MTF measuring device sets a region including a boundary in the chart image captured by the imaging system by the ROI setting means. At this time, the ROI setting means sets the area as a polygon, a circle, or an ellipse having an arbitrary shape without limiting the area to a rectangle.
Then, the MTF measuring device extracts the image of the region set by the ROI setting means as the ROI image from the chart image by the region extraction means.
Then, the MTF measuring device extracts the edge of the ROI image from the pixel distribution in the ROI image by, for example, the Sobel method, by the edge extraction means, and generates an edge extraction image.

Further, the MTF measuring device detects the inclination angle of the edge with respect to one predetermined coordinate axis of the orthogonal coordinates from the edge extracted image by the edge inclination detecting means by Hough transform or the like.
Then, the MTF measuring device rotates the ROI image by the tilt angle with respect to one coordinate axis in the direction of eliminating the tilt of the edge by the image rotation means. As a result, regardless of the shape of the ROI, the direction of the edge of the rotated ROI image becomes a direction orthogonal to the other coordinate axis of the Cartesian coordinates.
Further, the MTF measuring device approximates the rotated edge with a function such as a quadratic function by the function approximation means.
Then, the MTF measuring device corresponds the pixel position of the rotated ROI image with the position of the sub-pixel unit on the other coordinate axis of the Cartesian coordinates corresponding to the pixel position on the approximated function by the projection information generation means. Generate the attached projection information.
Then, the MTF measuring device uses the projected pixel value averaging means to project the pixel value of each pixel in the rotated ROI image onto the other coordinate axis of the Cartesian coordinates, and averages the pixel values in sub-pixel units. By doing so, an edge profile showing the characteristics of the edge is generated.

Further, the MTF measuring device obtains the line spread function by differentiating the edge profile by the frequency characteristic calculation means, and calculates the MTF by Fourier transforming the line spread function.

The MTF measuring device is an MTF measuring program for causing the computer to function as the ROI setting means, region extraction means, edge extraction means, edge inclination detecting means, image rotation means, edge profile generating means, and frequency characteristic calculation means. Can be operated with.

The present invention has the following excellent effects.
According to the present invention, since the ROI can be set in any shape, the ROI can be set in any place on the chart so as not to overlap with other edges. Thereby, the present invention can efficiently measure MTF from one chart.
Further, conventionally, the MTF can be measured only at an edge inclined in the horizontal direction, the vertical direction, or an angle of 45 degrees to several degrees. However, according to the present invention, the MTF can be measured no matter how many times the edge is tilted except in the horizontal direction, the vertical direction, or the oblique direction of 45 degrees. Therefore, according to the present invention, it is not necessary to accurately capture a slightly tilted edge image on the chart as in the conventional case. Thereby, the present invention can enhance the convenience of MTF measurement.

It is a block block diagram which shows the structure of the MTF measuring apparatus which concerns on embodiment of this invention. In the figure which shows the setting example of ROI, (a) shows a quadrangle, (b) shows a triangle, and (c) shows an ellipse ROI. In the figure for demonstrating the Hough transform, (a) shows Cartesian coordinates (xy coordinates), (b) shows polar coordinates, and (c) shows a Hough transform matrix. It is a figure which shows the rotation of the ROI image by the image rotation means, (a) shows the ROI image before rotation, (b) shows the ROI image after rotation. It is explanatory drawing for demonstrating the method of generating the projection information in a projection information generation means. It is explanatory drawing for demonstrating the method of generating an edge profile in the projected pixel value averaging means. It is a graph which shows the measurement result of MTF displayed by the measurement result display means. It is a flowchart which shows the operation of the MTF measuring apparatus which concerns on embodiment of this invention. It is an inmega cycle chart which is an example of a conventional chart. It is a figure which shows the example of the edge image used in the conventional slanted-edge method. It is a figure for demonstrating ROI in a chart image, (a) is a vertical ROI (vertical edge image), (b) is a horizontal ROI (horizontal edge image). It is a figure which shows the example of the yaw wheel chart including the edge in a multi-direction. It is a figure which shows the setting area of the conventional ROI, (a) shows an appropriate ROI including one edge, (b) shows an inappropriate ROI including a plurality of edges.

Hereinafter, embodiments of the present invention will be described with reference to the drawings.
[Configuration of MTF measuring device]
First, the configuration of the MTF measuring device 1 according to the embodiment of the present invention will be described with reference to FIG.

The MTF measuring device 1 measures the MTF representing the spatial frequency characteristic of the imaging system 2. When the measurement is performed by the MTF measuring device 1, the imaging system 2 and the display device 3 are connected and used.

The imaging system 2 is a video camera or still camera to be measured by the MTF, a camera including a lens to be measured by the MTF, or the like. The imaging system 2 may be a video processing device (for example, a down converter, an up converter [super-resolution]) that affects the MTF characteristics. Further, the object to be measured may be an image whose MTF changes depending on the DETAIL (detail) control of the camera.
The imaging system 2 outputs an image obtained by capturing the chart CH to the MTF measuring device 1.

The chart (chart for MTF measurement) CH is a chart having different contrasts at the boundary. Here, an arrow wheel chart will be described as an example of the chart CH, but any chart CH may be used as long as it includes edges having different contrasts at the boundary.

The display device 3 provides a user interface for operating the MTF measuring device 1, and displays a chart image captured by the imaging system 2, a graph as a measurement result, and the like. For example, the display device 3 is a liquid crystal display, an organic EL display, or the like.
The display device 3 may be provided with a display device for displaying the chart image captured by the imaging system 2 and a display device for displaying a graph or the like as a measurement result.

Hereinafter, the configuration of the MTF measuring device 1 for measuring the MTF of the imaging system 2 based on the images captured by the imaging system 2 will be described in detail.
As shown in FIG. 1, the MTF measuring device 1 includes a chart image storage means 10, an ROI setting means 11, an ROI image extracting means 12, an edge profile generating means 13, a frequency characteristic calculation means 14, and a measurement result display. Means 15 and.

The chart image storage means 10 stores an image (chart image) obtained by capturing the chart CH for MTF measurement in the imaging system 2. In the chart image storage means 10, a chart image is input from the imaging system 2 via a video input means (not shown), and the input chart image is stored. When a new chart image is input from the imaging system 2, the chart image storage means 10 overwrites the already stored chart image with the new chart image. The chart image storage means 10 is, for example, a general storage device such as a hard disk or a memory.
The chart image stored in the chart image storage means 10 is output to the display device 3 via a video output means (not shown) and is read out by the ROI image extraction means 12.

The ROI setting means 11 sets a region of interest (ROI) including a boundary (edge) in the chart image captured by the imaging system 2.
The ROI setting means 11 sets the closed area as the ROI by designating a closed area having an arbitrary shape by the measurer in the chart image displayed by the display device 3.

For example, as shown in FIG. 2A, the ROI setting means 11 specifies an arbitrary four points in the chart image to form a quadrangle (trapezoid, rhombus, etc.) having the designated four points as vertices. Set as ROI.
Further, as shown in FIG. 2B, the ROI setting means 11 sets an arbitrary three points in the chart image to set a triangle having the designated three points as vertices as the ROI. May be good.
Further, as shown in FIG. 2C, the ROI setting means 11 sets the designated ellipse as the ROI by designating any two points in the chart image as the vertices of the long axis of the ellipse. May be.
The ROI setting means 11 is not limited to the shape illustrated in FIG. 2, as long as the shape is set so as to include only one edge in the region, and may be any polygon or circle.

Further, the ROI setting means 11 displays a frame having a predetermined shape (for example, a quadrangle) at a designated position, and changes the position of the apex by an instruction means (not shown) such as a mouse to change the position of the region. The shape may be deformed. Further, the ROI setting means 11 may increase or decrease the vertices by the instruction means.

As a result, the ROI setting means 11 can set the ROI as a region including only one edge even when the chart CH includes a plurality of edges. Further, the ROI setting means 11 can set the ROI as a region including only one edge by changing the shape even when a plurality of edges are included in the region having a predetermined shape.
The ROI setting means 11 outputs the ROI information (for example, the coordinates of four points in the example of FIG. 2A) that specifies the region of the ROI to the ROI image extracting means 12.

The ROI image extraction means 12 extracts an ROI image from a chart image stored in the chart image storage means 10 and rotates the ROI image so that the direction of the edge is perpendicular to the projection axis. Here, the ROI image extraction means 12 includes a region extraction means 12a, an edge extraction means 12b, an edge inclination detection means 12c, and an image rotation means 12d.

The area extraction means 12a extracts an image (ROI image) of the area indicated by the ROI information set by the ROI setting means 11 from the chart image stored in the chart image storage means 10. That is, the area extraction means 12a extracts the ROI image in which the coordinate position inside the area indicated by the ROI information and the pixel value are paired from the chart image. The region extraction means 12a outputs the extracted ROI image to the edge extraction means 12b.

The edge extraction means 12b detects and extracts an edge from the ROI image extracted by the region extraction means 12a. As the edge extraction method in the edge extraction means 12b, a known method can be used. For example, the edge extraction means 12b applies a sharpening filter such as a sobel filter, a Laplacian filter, and a canny filter that emphasizes a change in the brightness value in the image to the ROI image, and binarizes the edge to extract the edge. Generate an extracted image.

When the ROI image has an arbitrary shape and the ROI boundary is extracted as an edge, the edge extraction means 12b deletes the edge of the ROI boundary by mask processing. For example, the edge extraction means 12b excludes the edge included in the region where the contour line of the ROI is thickened, based on the ROI information that identifies the shape of the ROI.
The edge extraction means 12b outputs the edge extraction image from which the edges are extracted to the edge inclination detection means 12c.

The edge tilt detecting means 12c regards the edge as a straight line from the edge extracted image obtained by extracting the edge by the edge extracting means 12b, and determines the tilt angle of the orthogonal coordinates with respect to one predetermined coordinate axis (here, the y-axis). It is to detect. A known method can be used for edge inclination detection in the edge inclination detecting means 12c. For example, the edge inclination detecting means 12c detects the inclination of the edge by the Hough transform.

Here, with reference to FIG. 3, edge inclination detection using the Hough transform performed by the edge inclination detecting means 12c will be described.
As shown in FIG. 3A, when the image coordinates of the edge extracted image E are Cartesian coordinates (x, y), the length r and the angle θ of the normal line from the origin to the edge (straight line) e are as follows. It can be expressed by the equation (1) of.

Here, θ is set to 0 degrees on the vertical axis (y-axis) of the orthogonal coordinates (x, y) and positive in the counterclockwise direction of the coordinate origin O.
At this time, there are innumerable straight lines passing through each coordinate point of the edge e. However, when the coordinates (x, y) of each coordinate point of the edge e are substituted into the above equation (1) and r and θ are plotted in polar coordinates (r, θ) as shown in FIG. 3 (b). , The plot points are concentrated at the position of the inclination angle θ ₁ with respect to the vertical axis of the edge e.

The edge inclination detecting means 12c calculates the length r from the xy coordinate value of each coordinate point of the edge e in the edge extracted image E by the above equation (1) in the range of the angle θ from −90 degrees to 90 degrees. .. Note that r and θ may be appropriate sampling intervals according to the resolution of the ROI image. For example, r is a pixel interval and θ is 0.1 degrees as the minimum unit.
Then, as shown in FIG. 3C, the edge inclination detecting means 12c generates a Hough transform matrix A having the values of r and θ as the elements of the matrix from the edge e.
Then, the edge inclination detecting means 12c searches for the peak value of the angle θ in the Hough transform matrix A and sets it as the inclination angle of the edge e.

Since the edge inclination detecting means 12c calculates the inclination angle of the edge at a predetermined interval such as 0.1 degree unit, the exact inclination angle is not obtained. However, in the edge profile generating means 13 described later, the projection of the pixel on the projection axis in the slanted-edge method is performed along the edge, so that there is no problem.

Returning to FIG. 1, the configuration of the MTF measuring device 1 will be described.
The edge tilt detecting means 12c outputs the detected tilt angle of the edge to the image rotating means 12d.

The image rotating means 12d rotates the ROI image so that the direction of the edge of the ROI image is perpendicular to the projection axis based on the tilt angle of the edge detected by the edge tilt detecting means 12c.
When the edge tilt detecting means 12c detects that the edge is tilted by an angle θ with respect to the vertical axis (y-axis), the image rotating means 12d rotates the ROI image by −θ.
As a result, the ROI image extracting means 12 can convert the ROI image so that the direction of the edge is perpendicular to the projection axis regardless of the shape of the ROI.
The image rotating means 12d outputs the rotated ROI image to the edge profile generating means 13.

Since the ROI image extracting means 12 tilts the ROI image tilted by −θ, the edge can be made substantially perpendicular to the projection axis. In this case, as shown in FIG. 4A, for example, even if the inclination θe of the edge e is −30 degrees, as shown in FIG. 4B, each pixel of the ROI image rotated by 30 degrees The centers are offset with respect to the x-axis subpixels. Therefore, the pixel value of the edge is distributed to a plurality of sub-pixels, and the Slanted-edge method can be applied.

On the other hand, when the angles θ are 0 degrees, ± 45 degrees, and ± 90 degrees, that is, when the edges are in the vertical direction, the diagonal 45 degree direction, and the vertical direction, the pixel values of the edges change with the rotation of the ROI image. It is not suitable for the Slanted-edge method because it will be projected on the same subpixel. However, it is easy to avoid these angles by tilting the chart CH or the imaging system 2. If the angle θ is 0 degree, ± 45 degree, ± 90 degree, the MTF measuring device 1 may display an error message or the like, and the measurer may change the angle of the chart CH.

The edge profile generating means 13 generates an edge profile showing the pixel distribution shape of the edge of the ROI image extracted and rotated by the ROI image extracting means 12. Here, the edge profile generation means 13 includes a function approximation means 13a, a projection information generation means 13b, and a projection pixel value averaging means 13c.

The function approximating means 13a approximates the edge in the ROI image with a predetermined non-linear function such as a quadratic function or an exponential function from the pixel value distribution of the ROI image generated by the ROI image extracting means 12. That is, the edge function approximating means 131 approximates the edge when the ROI image is arranged at coordinates (xy coordinates) having two axes in the horizontal direction and the vertical direction as a function. As a method of approximating this edge with a function, a known method (see Patent Document 3) may be used.
The function approximation means 13a outputs a parameter (coefficient) for specifying the function to the projection information generation means 13b.

The projection information generation means 13b sets the pixel position of each pixel of the ROI image along the slope (curve) of the edge specified by the function obtained by the function approximation means 13a to the projection axis (the other coordinate axis of Cartesian coordinates; here. Then, the projection information associated with the coordinates (referred to as the x-axis) is generated.
The coordinate system of the projection axis is a sub-pixel unit smaller than the coordinate system of the pixel unit of the ROI image, and is, for example, 1/4 or 1/8 of one pixel.
The projection information generation means 13b outputs the projection information of the ROI image to the projection pixel value averaging means 13c.

Here, with reference to FIG. 5, a method in which the projection information generating means 13b associates the pixels of the ROI image with the projection axis (x-axis) by using an approximated function will be described.
In FIG. 5, in order to make the explanation easy to understand, the edge e specified by the nonlinear function is superimposed on the ROI image R. Further, the white point or the black point of each pixel of the ROI image R indicates the pixel position in the xy coordinates of the ROI image R.

As shown in FIG. 5, the projection information generating means 13b associates the pixel position on the ROI image R with the coordinates in subpixel units (x _sub coordinates) of the x coordinates along the inclination of the edge e.
For example, projection information generating unit 13b is a coordinate pixel _{g 1} on the edge e, the function of the edge e, associated with _{x 1} of _{x sub} coordinates corresponding to y = 0. The projection information generating unit 13b is a pixel g ₂ does not exist on the edge e, (in the figure, dotted line) function passing through the pixel g ₂ in which slide the edge e in the horizontal direction, corresponding to y = 0 x associates and _{x 2} of the _sub coordinate.

Returning to FIG. 1, the configuration of the MTF measuring device 1 will be described.
The projection pixel value averaging means 13c sets the pixel value of each pixel of the ROI image rotated by the ROI image extraction means 12 for each sub-pixel of the projection axis based on the projection information generated by the projection information generation means 13b. It is projected and averaged. That is, as shown in FIG. 6, the projected pixel value averaging means 13c averages the pixel values of the ROI image in sub-pixel units of the projection axis (x-axis) along a function that approximates the edges. Generate an edge profile showing the pixel distribution shape of the edge.
The projected pixel value averaging means 13c outputs the generated edge profile to the frequency characteristic calculation means 14.

The frequency characteristic calculation means 14 calculates the MTF from the edge profile generated by the edge profile generation means 13.
The frequency characteristic calculation means 14 obtains a line spread function (LSF) by differentiating the edge profile generated by the edge profile generation means 13, and Fourier transforms the LSF. As a result, the frequency characteristic calculation means 14 can obtain the MTF.
The frequency characteristic calculation means 14 outputs the calculated MTF to the measurement result display means 15.

The measurement result display means 15 displays the measurement result of the MTF calculated by the frequency characteristic calculation means 14. The measurement result display means 15 graphs the MTF calculated by the frequency characteristic calculation means 14.
For example, as shown in FIG. 7, the measurement result display means 15 plots data corresponding to a graph in which the horizontal axis is frequency (cycles / pixel) and the vertical axis is MTF.
The measurement result display means 15 outputs this graph to the display device 3 and displays it. As a result, the measurer can recognize the MTF of the imaging system 2 to be measured.

By configuring the MTF measuring device 1 as described above, the MTF measuring device 1 can detect the inclination of the edge from the ROI of an arbitrary shape and measure the MTF.
As a result, the MTF measuring device 1 can set the edge portion existing at an arbitrary position on the chart CH as the ROI, and can efficiently measure the MTF from one chart CH.

In addition, the MTF measuring device 1 detects the inclination of the edge from the ROI and converts the slanted-edge method into an applicable edge inclination. It is not necessary to take an edge image that is tilted several degrees from the angle. Therefore, the MTF measuring device 1 can measure the MTF no matter how many times the edge is tilted except in the horizontal direction, the vertical direction, or the angle of 45 degrees, and the convenience of the MTF measurement can be enhanced.
The MTF measuring device 1 can operate a computer (not shown) with an MTF measuring program for functioning as the above-mentioned means.

[Operation of MTF measuring device]
Next, the operation of the MTF measuring device 1 according to the embodiment of the present invention will be described with reference to FIG. 8 (see FIG. 1 for the configuration as appropriate).

First, in step S1, the MTF measuring device 1 inputs the chart image captured by the imaging system 2, stores it in the chart image storage means 10, and displays it on the display device 3.
Then, in step S2, the MTF measuring device 1 sets the ROI including the boundary (edge) by the ROI setting means 11. For example, the ROI setting means 11 is a pointing device (not shown) operated by the measurer in the chart image displayed by the display device 3, and a region having an arbitrary shape is designated to form the shape of the ROI. Set the ROI information to specify.
After that, in step S3, the MTF measuring device 1 uses the region extraction means 12a to use the chart image stored by the chart image storage means 10 as an ROI image of the region specified by the ROI information set in step S2. Extract.

Then, in step S4, the MTF measuring device 1 extracts the edge from the ROI image extracted in step S3 by the edge extraction means 12b.
Further, in step S5, the MTF measuring device 1 detects the inclination angle of the edge from the edge extracted in step S4 by the edge inclination detecting means 12c by using the Hough transform.

After that, in step S6, the MTF measuring device 1 rotates the ROI image extracted in step S3 in the direction of making the tilt angle detected in step S5 0 degrees by the image rotating means 12d.
Then, in step S7, the MTF measuring device 1 approximates the edge of the ROI image rotated in step S6 by a function approximation means 13a with a non-linear function such as a quadratic function.

Then, in step S8, the projection information generation means 13b sets the pixel position of each pixel of the ROI image along the slope (curve) of the edge specified by the function approximated in step S7 to the coordinates of the projection axis (x-axis). Generates projection information associated with.
Further, in step S9, the MTF measuring device 1 sets the pixel value of each pixel of the ROI image for each sub-pixel of the projection axis based on the projection information generated in step S8 by the projection pixel value averaging means 13c. An edge profile is generated by projecting and averaging.

After that, in step S10, the MTF measuring device 1 obtains a line spread function (LSF) by differentiating the edge profile generated in step S9 by the frequency characteristic calculation means 14, and Fourier transforms the LSF. Calculate the MTF.

Then, in step S11, the MTF measuring device 1 displays the MTF measurement result calculated in step S10 by the measurement result display means 15. For example, as shown in FIG. 7, the measurement result display means 15 plots data corresponding to a graph in which the horizontal axis is frequency (cycles / pixel) and the vertical axis is MTF, and displays the data on the display device 3.

By the above operation, the MTF measuring device 1 can measure the MTF at an arbitrary position on the chart set by the measurer and an ROI of an arbitrary shape.
As a result, the MTF measuring device 1 can efficiently measure the MTF from one chart. In addition, the MTF measuring device 1 detects the inclination of the edge in the ROI and rotates the ROI image. It can be measured, and the convenience of MTF measurement can be enhanced.

[Modification example]
Although the embodiment of the present invention has been described above, the present invention is not limited to this embodiment.

(Modification example 1)
Here, the MTF measuring device 1 detects the inclination angle of the edge in the ROI image by the edge extraction means 12b and the edge inclination detecting means 12c.
However, since the inclination angle of this edge may be any angle other than the horizontal direction, the vertical direction, or the oblique 45 degrees, the measurer may manually set the inclination.

In that case, the MTF measuring device 1 may include edge inclination specifying means (not shown) that accepts the designation of the edge inclination from the measurer instead of the edge extracting means 12b and the edge inclination detecting means 12c. Here, in the edge inclination designating means, for example, the measurer specifies two points on the edge of the ROI image, and the inclination of the straight line specified by the two points is set as the edge inclination. Further, for example, the edge inclination specifying means draws a line segment on the ROI image, and by operating the mouse of the measurer or the like, the positions of both ends of the line segment are moved to match the edge, so that the edge is adjusted from the inclination of the line segment. You may find the slope of.

(Modification 2)
Further, here, the MTF measuring device 1 approximates the edge on the ROI image from the ROI image by a function such as a quadratic function by the function approximating means 13a.
However, the function approximation means 13a may function-approximate the edges from the edge extraction image generated by the edge extraction means 12b.

In that case, the function approximation means 13a rotates the edge extraction image generated by the edge extraction means 12b in the direction of eliminating the inclination detected by the edge inclination detection means 12c (the direction in which the inclination angle is set to 0 degrees) and rotates. Function approximation may be performed from the data string of the pixel positions of the edges in the later edge extraction image.

1 MTF measuring device 10 Chart image storage means 11 ROI setting means 12 ROI image extraction means 12a Area extraction means 12b Edge extraction means 12c Edge inclination detection means 12d Image rotation means 13 Edge profile generation means 13a Function approximation means 13b Projection information generation means 13c Projected pixel value averaging means 14 Frequency characteristic calculation means 15 Measurement result display means 2 Imaging system 3 Display device CH chart (MTF measurement chart)

Claims (6)

An MTF measuring device that measures an MTF that represents the spatial frequency characteristics of an imaging system using charts with different contrasts at the boundaries.
An ROI setting means for setting a region having an arbitrary shape including the boundary in a chart image obtained by capturing the chart by the imaging system.
An area extraction means for extracting an image of an area set by the ROI setting means as an ROI image from the chart image,
An edge extraction means that extracts an edge from the ROI image and generates an edge extraction image,
An edge tilt detecting means for detecting the tilt angle of an edge with respect to one of the predetermined coordinate axes of Cartesian coordinates from the edge extracted image.
An image rotating means that rotates the ROI image by the tilt angle in a direction that eliminates the tilt of the edge with respect to the one coordinate axis.
A function approximation means that approximates the edge after rotation with a function,
A projection information generating means that generates projection information in which the pixel position of the ROI image after rotation is associated with the position of each sub-pixel in the other coordinate axis of the Cartesian coordinates corresponding to the pixel position on the function. ,
Using the projection information, the pixel value of each pixel in the rotated ROI image is projected onto the other coordinate axis and averaged in units of the sub-pixels to generate an edge profile showing the characteristics of the edge. Projected pixel value averaging means and
A frequency characteristic calculation means for calculating MTF from the edge profile,
An MTF measuring device comprising. The MTF measuring device according to claim 1, wherein the ROI setting means sets a polygon, a circle, or an ellipse having an arbitrary shape as the region. The ROI setting means sets a shape specified by the plurality of points as the region by designating a plurality of points, and changes the shape of the region by designating the movement of the points. The MTF measuring device according to claim 1 or 2, characterized in that. The MTF measuring device according to any one of claims 1 to 3, wherein the edge inclination detecting means calculates the inclination angle of the edge by a Hough transform in the edge extracted image. The function approximation means rotates the edge extraction image generated by the edge extraction means in a direction that eliminates the inclination of the edge detected by the edge inclination detection means, and approximates the edge in the rotated edge extraction image by a function. The MTF measuring apparatus according to any one of claims 1 to 4, wherein the MTF measuring device is characterized in that. An MTF measurement program for causing a computer to function as each means of the MTF measuring device according to any one of claims 1 to 5.