JPH09160149A - Image processing method and device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To simplify the constitution of an image processing device removing a grid image from an image including the grid image and to shorten processing time for high-speed processing. SOLUTION: A reduced image signal SS is obtained by performing filtering processing by a grid removing filter to an image signal S expressing an original image by a filtering means 31. The signal SS is enlarged to nearly the same size as the original image by an interpolating and enlarging means 32, and subtracted from the image signal S so as to obtain a detail image signal SD. In the case enlargement ratio is >=1 in an enlarging and reducing means 40, an already processed image signal Sproc is generated, based on the signals SS and SO, and in the case it is <1, the signal Sproc is generated, based on the signal SS.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention reads an image at a predetermined sampling interval from a recording sheet on which an image and a striped pattern corresponding to the grid are recorded by photographing using a grid and enlarging the image. Further, the present invention relates to an image processing method and apparatus for reducing, and a grid removal filter for removing a grid image.

[0002]

2. Description of the Related Art A recorded radiographic image is read to obtain an image signal, and the image signal is subjected to appropriate image processing.
Reproduction and recording of images are performed in various fields.
In addition, the applicant of the present application, radiation (X-ray, α-ray, β-ray, γ
Ray, electron beam, ultraviolet ray, etc.) causes a part of this radiation energy to be accumulated, and then irradiation with excitation light such as visible light causes stimulated emission depending on the accumulated energy. Radiation image of a subject such as a human body is temporarily photographed and recorded on a sheet-shaped stimulable phosphor by using a fluorescent phosphor, and this stimulable phosphor sheet is scanned with excitation light such as laser light to stimulate emission. Image data is obtained by generating light and photoelectrically reading the resulting stimulated emission light, and based on this image data, the radiation image of the subject is output as a visible image on a recording material such as a photographic light-sensitive material or a CRT. A radiation image recording / reproducing system has already been proposed (JP-A-55-12429).
56-11395, 55-163472, 56-104645,
55-116340, etc.).

On the other hand, when a radiation image of a subject is photographed and recorded on a recording sheet such as an X-ray film or a stimulable phosphor sheet, the radiation scattered by the subject is not irradiated to the recording sheet by about 4.0 lines / mm. In some cases, a grid is arranged between the subject and the recording sheet, in which a grid in which radiation that does not transmit radiation and that easily transmits radiation such as aluminum and wood are alternately arranged at a fine pitch. When the image is captured using the grid, the radiation scattered by the subject is less likely to irradiate the recording sheet, so the contrast of the radiation image of the subject can be improved, but on the other hand, the recording sheet has a fine striped pattern along with the subject image. A grid image is recorded.

On the other hand, a radiation image reading apparatus that obtains an image signal from a recording sheet on which a radiation image is recorded usually reads photoelectrically the light carrying the radiation image information to obtain the highest spatial frequency (hereinafter , And a sampling interval Δx = 1 / (2 · fss) corresponding to the required highest spatial frequency is represented by fss) to obtain a digital image signal. The image signal obtained in this way is not only useful information representing the radiation image of the object, but is caused by the grid image even if the spatial frequency of the grid image is higher than the required highest spatial frequency fss. It may also contain noise.

FIG. 8 is a graph showing a spatial frequency characteristic of a radiation image (including a grid image) recorded on a recording sheet in a direction orthogonal to the striped pattern of the grid image.

It is assumed here that imaging is performed using a grid of 4.0 lines / mm, and the spatial frequency of the grid image is 4 cycles / mm, and the maximum spatial frequency fss required to reproduce the radiation image of the subject is 2.5 cycles / mm. It will be described as mm.

FIG. 9 shows a sampling interval Δx = corresponding to fss = 2.5 (cycle / mm) in order to obtain information on a spatial frequency band below fss = 2.5 (cycle / mm) necessary for reproducing a radiation image of a subject. 1 / (2 · fss) = 0.2 (mm), that is, a diagram for explaining a mixed state of noise when sampling is performed 5 times per 1 mm.

In this case, the so-called aliasing noise is mixed in at the position where the graph shown by the solid line in the figure (the same as the graph in FIG. 8) is folded back at fss = 2.5 (cycle / mm). Therefore, in this case, the spatial frequency of the grid image is 4 cycl.
Aliasing of e / mm occurs at 1 cycle / mm.

FIG. 10 shows that the sampling interval Δx = 1 / (2
-The spatial frequency characteristic of the radiographic image carried by the image signal obtained by sampling at fss) = 0.2 (mm) is shown.

Noise corresponding to the grid image is mixed in at a position of 1 cycle / mm, and when a visible image is reproduced and recorded, it appears as a striped pattern of 1 cycle / mm in the visible image.
When observing a radiation image, even if the grid image is in a spatial frequency band that is not so noticeable, a stripe pattern appears in the spatial frequency band that is easily noticeable when sampling and obtaining an image signal, and is reproduced based on this image data. Sometimes the visible image was very hard to see.

[0011] Therefore, the applicant has applied a filtering process to the image data by a grid removal filter that removes the grid image from the image containing the grid image to reduce or remove the stripe pattern noise caused by the grid. An image processing method for obtaining a visible image that is easy to observe has been proposed (JP-A-3-14039).

[0012]

SUMMARY OF THE INVENTION
In the method described in 3-114039, although the grid image can be removed, it is necessary to prepare a grid removal filter of a different size according to the scaling ratio of the image, and multiple grid removal filters are stored. Since the storage means for doing so is required, the device becomes expensive. Furthermore, since it is necessary to perform filtering processing with grid removal filters of different sizes depending on the scaling ratio, especially when a grid removal filter of a relatively large size is used, it takes a long time to perform the filtering operation and high-speed processing is required. Could not be done.

In view of the above circumstances, it is an object of the present invention to provide an image processing method and apparatus which can be constructed at low cost and can perform high-speed processing.

Another object of the present invention is to provide a grid removal filter used in the above image processing method and apparatus.

[0015]

In the image processing method according to the present invention, an original image composed of a striped grid image and a radiation image corresponding to the grid is recorded by capturing an image using the grid. An image processing method for obtaining a processed image by reading the original image at a predetermined sampling interval from a recording sheet, and enlarging or reducing the original image at a predetermined enlargement / reduction ratio, wherein: A filtering process is performed by a grid removal filter that removes the grid image at a predetermined pixel interval to obtain a reduced image in which the grid image is removed from the original image and the original image is reduced, and the reduced image is the original image. The image is enlarged to substantially the same size as the image, and the enlarged reduced image is subtracted from the original image to obtain a detailed image of the original image and the predetermined enlargement / reduction ratio. In the case of 1 or more, the reduced image is enlarged to substantially the same size as the original image, the enlarged reduced image is added to the detailed image to obtain an added image, and the added image is added to the enlargement / reduction ratio. When the predetermined scaling ratio is less than 1, the processed image is obtained by enlarging according to
It is characterized in that the processed image is obtained by enlarging or reducing the reduced image according to the enlargement / reduction ratio.

The image processing apparatus according to the present invention is for carrying out the above-mentioned image processing method, and is higher than the highest spatial frequency of the desired spatial frequency band by taking an image using a grid. The original image is read at a predetermined sampling interval from a recording sheet on which an original image consisting of a striped grid image corresponding to the grid and a radiation image having a spatial frequency is recorded, and the original image is enlarged by a predetermined amount. An image processing apparatus for obtaining a processed image by enlarging or reducing according to a reduction ratio, wherein the original image is
Filtering means for obtaining a reduced image in which the grid image is removed from the original image and the original image is reduced by performing a filtering process by a grid removal filter that removes the grid image at a predetermined pixel interval, and the reduced image. To a size substantially equal to that of the original image, and a detailed image calculation means for obtaining a detailed image of the original image by subtracting the enlarged reduced image from the original image; In the case of 1 or more, the reduced image is enlarged to substantially the same size as the original image, the enlarged reduced image is added to the detailed image to obtain an added image, and the added image is added to the enlargement / reduction ratio. The processed image is obtained by enlarging according to the above-mentioned processing, and when the predetermined enlargement / reduction rate is less than 1, the processing is performed by enlarging or reducing the reduced image according to the enlargement / reduction rate. It is characterized in that a scaling means for obtaining an image.

Furthermore, the grid removal filter according to the present invention is used in the above-mentioned image processing method and apparatus, and has a response of about 1 at a frequency of 2.5 cycle / mm or less and a frequency of 3.4 cycle / mm. 4.0cy
In cle / mm, it has a frequency characteristic that the response is substantially zero.

Specifically, the mask size is 7 × 7, and the filter value is

[0019]

(Equation 3)

However, each value is characterized by having an error of ± 10%.

[0021]

According to the image processing method and apparatus of the present invention, the original image is filtered by the grid removing filter at a predetermined pixel interval to obtain a reduced image smaller in size than the original image. Next, the reduced image is enlarged to have substantially the same size as the original image by interpolation calculation or the like, and the enlarged reduced image is subtracted from the original image. Since the image obtained by enlarging the reduced image here has the detailed information in the original image missing, the image obtained by subtracting the enlarged reduced image from the original image is The detailed image represents only the detailed information of the original image.
Then, the enlargement / reduction of the original image is performed based on the reduced image and the detailed image according to the enlargement / reduction ratio.
When the enlargement / reduction ratio is 1 or more, that is, when enlargement is performed, the reduced image is enlarged to have substantially the same size as the original image, the detailed image is added to the enlarged image, and the original image is almost completely reproduced. Based on the scale. This is because the reduced image lacks detailed information compared to the original image, so if the image is enlarged based only on the reduced image, the resolution of the enlarged image will deteriorate, and When enlarging, the reason is that the moire due to the grid image becomes inconspicuous. On the other hand, when the enlargement / reduction ratio is less than 1, that is, when the reduction is performed, the moire due to the grid image becomes conspicuous.

As described above, the image processing method and apparatus according to the present invention can deal with various enlargement / reduction ratios by performing the filtering process by the grid removing filter only once, and therefore, a plurality of types of grid removing filters can be applied. It becomes unnecessary to have the above, and the structure of the device can be simplified. Further, since the filtering process only needs to be performed once, it is possible to shorten the calculation time for performing the filtering process and perform the process at high speed.

Also, the grid normally used is 34 lines / cm.
Alternatively, since it is 40 lines / cm, a grid image can be removed almost completely by using a grid removal filter having a frequency characteristic with a response of almost 0 at 3.4 cycle / mm and 4.0 cycle / mm. It will be possible.

Furthermore, 3.4 cycle / mm and 4.0 cycl
At e / mm, as a grid removal filter having a frequency characteristic with a response of almost 0,

[0025]

(Equation 4)

By using a relatively small size filter of 7 × 7 size, the calculation time for the filtering process can be shortened as compared with the case of performing the filtering process by the large size filter.

Furthermore, since the original image is divided into the reduced image and the detailed image, it is extremely advantageous when compressing the image to reduce the file size.

[0028]

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is a diagram showing an outline of an example of the radiation image capturing apparatus. Here, an example in which the above-described stimulable phosphor sheet is used as a recording sheet will be described.

The radiation 2 emitted from the radiation source 1 passes through the subject 3 and further through the grid 4 and illuminates the stimulable phosphor sheet 11. In the grid 4, lead 4a and aluminum 4b are alternately arranged at a pitch of 4 lines / mm.
The radiation 2 is blocked by the lead 4a and transmitted through the aluminum 4b to irradiate the sheet 11. Therefore, a striped grid image of 4 lines / mm or 3.4 lines / mm is accumulated and recorded on the sheet 11 together with the subject image. The radiation 2a scattered in the subject 3 is obliquely incident on the grid 4 and thus is shielded by the grid or reflected by the grid 4 and is not irradiated to the sheet 11, and therefore the sheet 11 is clear with less irradiation of scattered radiation. Various radiation images are accumulated and recorded.

FIG. 2 shows a subject image (shaded areas in the figure) and a striped grid image (vertical stripes in the figure) stored and recorded in the stimulable phosphor sheet 11 by photographing using a grid. It is a figure showing. In this manner, a radiation image on which the subject image 5 and the striped pattern 6 are superimposed is recorded on the sheet 11.

FIG. 3 is a perspective view of an embodiment of a radiation image reading apparatus including the image processing method of the present invention.

The stimulable phosphor sheet 11 on which a radiation image is set at a predetermined position of the reading unit 10, which is an example of a reading unit, is conveyed by a sheet conveying unit 15 such as an endless belt driven by a driving unit (not shown). , Are conveyed (sub-scanning) in the arrow Y direction. On the other hand, the light beam 17 emitted from the laser light source 16 is reflected and deflected by a rotary polygon mirror 18 driven by a motor 24 and rotating at a high speed in the arrow direction, and fθ
After passing through the focusing lens 19 such as a lens, the optical path is changed by the mirror 20 and the light enters the sheet 11, and the main scanning is performed in the arrow X direction substantially perpendicular to the sub-scanning direction (arrow Y direction). From the portion of the sheet 11 irradiated with the light beam 17, a stimulating luminescent light 21 having an amount corresponding to the radiation image information stored and recorded is diverged, and the stimulating luminescent light 21 is guided by a light guide 22, Photoelectrically detected by a photomultiplier (photomultiplier tube) 23. The light guide 22 is made by molding a light guide material such as an acrylic plate, and has a linear incident end surface.
22a is arranged so as to extend along the main scanning line on the stimulable phosphor sheet 11, and the light receiving surface of the photomultiplier 23 is coupled to the emission end surface 22b formed in an annular shape. The stimulated emission light 21 that enters the light guide 22 from the incident end face 22a is
Repeatedly undergoes total internal reflection inside the light guide 22, and exits
It is emitted from 22b and received by the photomultiplier 23,
Photomultiplier is photostimulated emission light 21 that represents a radiation image
It is converted into an electric signal by 23.

The analog output signal S _A output from the photomultiplier 23 includes the highest first space of the spatial frequency bands in a desired range necessary for reproducing and outputting a good radiation visible image. Information on the spatial frequency band higher than the frequency fss = 2.5 cycles / mm, particularly information on the grid image 6 shown in FIG. 2 is also included. The information on the grid image 6 is one of the causes of making the visible image difficult to see when observing the visible image, and is information that needs to be removed.

The analog output signal S _A is logarithmically amplified by the log amplifier 26, sampled at a predetermined sampling interval by the A / D converter 28, and digitized.
A digital image signal S is obtained. The image signal S is temporarily stored in the storage unit 29.

FIG. 4 is a diagram showing each sampling point on the stimulable phosphor sheet 11. The horizontal direction (x direction) in the figure corresponds to the X direction (main scanning direction) in FIG. 3, and the vertical direction (y direction).
Corresponds to the Y direction (sub-scanning direction) in FIG. In the figure
The mark represents each sampling point in the initial image signal, and the mark ◯ represents each sampling point after resampling. In the present embodiment, these sampling points are set in both the x direction (main scanning direction) and the y direction (sub scanning direction).
Sampling interval 2 · Δ = 1 / (2 ・ corresponding to the highest spatial frequency fss = 2.5 cycle / mm of the desired spatial frequency band necessary for reading the radiation image recorded on the stimulable phosphor sheet 11. fss) = 0.2 (mm), half sampling interval Δ = 0.1 (mm), that is, twice the highest spatial frequency fss = 2.5 (cycle / mm), spatial frequency fsw = 5.
Sampling interval corresponding to 0 (cycle / mm) Δ = 1 /
Sampling is performed at (2 · fsw) = 0.1 (mm). The adjustment of the sampling interval is performed by adjusting the sampling time interval of the A / D converter 28 (see FIG. 3) in the x direction (main scanning direction) and the like in the y direction (sub scanning direction). This is performed by adjusting the transport speed of the sheet 11 by the sheet transport means 15 (see FIG. 3).

Although a grid removal filter of 5 lines / mm or more may be used, since the sampling interval is Δ = 0.1 mm as described above, the grid image is removed in the obtained image signal. It doesn't matter.

The image signal S thus obtained at the sampling interval (0.1 mm interval) shown by the dot in FIG. 4 is fsw
= 5.0 (cycle / mm) or less, the information of the striped pattern 6 shown in Fig. 2 (4.0cycle / mm or
It also carries 3.4 cycles / mm). Further, in this embodiment, aliasing of the striped pattern 6 does not occur.

The image signal S is temporarily stored in the storage unit 29 shown in FIG. 3 and then input to the image processing means 30 to be subjected to image processing as follows.

FIG. 5 is a block diagram showing the outline of an apparatus for carrying out the image processing method according to the present invention. As shown in FIG. 5, the apparatus for carrying out the image processing method according to the present invention performs a filtering process on the image signal S every four pixels by a grid removal filter to be described later to obtain ¼ of the original image. The filtering means 31 for obtaining the reduced image signal S _D representing the reduced image of the size and the reduced image signal S _S obtained by the filtering processing means 31 are interpolated to obtain an image of substantially the same size as the original image. obtain 'an interpolation enlarging means 32 to obtain an image signal S _S from the image signal S' image signal S _S representing a by subtracting the detailed image signal S _D representing the detailed image carrying only detailed information of the original image Subtracting means 33, reduced image signal S _S and detailed image signal S
A memory 34 for storing _D , a magnifying power judgment means 35 for judging whether the magnifying power inputted from an input means (not shown) is 1 or more and less than 1, and a reduced image signal S _S according to the magnifying power.
And a scaling means 40 for scaling the detailed image signal S _D.

The enlarging / reducing means 40 enlarges the reduced image signal S _S into an image signal S _S ′ representing an image having substantially the same size as the original image when the enlarging ratio is 1 or more, and an image enlarging means 41. The image signal S is obtained by adding S _S ′ and the detailed image signal S _D.
And the image signal S obtained by the adding means 42 is subjected to interpolation calculation processing to obtain the processed image signal S.
The interpolation means 43 obtains proc, and the interpolation means 44 obtains a processed image signal Sproc by subjecting the reduced image signal _Ss to an interpolation calculation process when the enlargement ratio is less than one.

Next, the operation of the image processing apparatus according to the embodiment of the present invention will be described. First, the filtering means 31 performs a filtering process on the image signal S every 4 pixels, that is, every 1 pixel in the vertical direction and every 1 pixel in the horizontal direction, and a reduced image signal S representing a reduced image having a size of ¼ of the original image. _S is obtained. This filtering process is performed by the following 7 × 7 filter.

[0043]

(Equation 5)

The frequency characteristic of this filter is shown in FIG.
As shown in FIG. 6, the frequency characteristic of this filter is 3.4.
No response at cycle / mm and 4.0 cycle / mm
It is something that has become. Therefore, the reduced image represented by the reduced image signal S _S obtained by performing the filtering process with this filter has the grid image removed. The reduced image signal S _S is stored in the memory 34.

Further, the reduced image signal S _S is inputted to the interpolation enlarging means 32 and subjected to an interpolation operation to obtain an image signal S _S ′ representing an image having substantially the same size as the original image. Examples of this interpolation calculation include Cubic spline interpolation calculation and B-spline interpolation calculation.

The Cubic spline interpolation calculation and the B spline interpolation calculation will be described in detail below.

Successive pixels X _k-2 , X _k-1 , X _k , X _{k + 1} , X obtained by digitally reading from the image
_As shown in FIG. 7, the image data of _{k + 2} , ...
_{Let k-2} , Y _k-1 , Y _k , Y _{k + 1} , Y _{k + 2} , .... Here, the cubic spline interpolation function is applied to each section X _{k-2 to} X.
_k−1 , X _{k−1 to} X _k , X _{k to} X _{k + 1} , X _{k + 1 to} X _{k + 2} are respectively set, and spline interpolation functions corresponding to the respective sections are f _k−2 , f Let _k-1 , f _k , f _{k + 1} , f _{k + 2} . Each of these interpolation functions is a cubic function having the position of each section as a variable.

First, the point to be interpolated (hereinafter,
A case will be described in which X _{p (referred} to as an interpolation point) is in the range of section X _{k to} X _{k + 1} . The spline interpolation function f _k corresponding to the sections X _{k to} X _{k + 1} is expressed by the following equation (1).

F _k (x) = A _k x ³ + B _k x ² + C _k x + D _k (1) In the Cubic spline interpolation calculation, the spline interpolation function f _k passes through the original sample point (pixel) and It is necessary that the first-order differential coefficient be continuous between each section, and from these conditions, it is necessary to satisfy the following equations (2) to (5).

F _k (X _k ) = Y _k (2) f _k (X _{k + 1} ) = Y _{k + 1} (3) f _k ′ (X _k ) = f _k−1 ′ (X _k ) (4 ) F _k ′ (X _{k + 1} ) = f _{k + 1} ′ (X _{k + 1} ) (5) Note that f _k ′ is the first derivative (3A _k x ² +2) of the function f _k .
B _k x + C _k ). Strictly speaking, the Cubic spline interpolation calculation also includes the continuation condition of the second derivative, but since the formula becomes complicated, it is generally used by simplifying it as described above.

[0051] Also in the Cubic spline interpolating operation, the first-order differential coefficient at the picture element X _k is the pixel X _k
For the X _k-1 and X _{k + 1} is the preceding and succeeding pixels, these image data _{Y k-1, Y k +} 1 gradient _{(Y k + 1 -Y k-} 1)
Since the condition is that it matches / (X _{k + 1} −X _k−1 ), it is necessary to satisfy the following expression (6).

F _k ′ (X _k ) = (Y _{k + 1} −Y _k−1 ) / (X _{k + 1} −X _k−1 ) (6) Similarly, the first derivative of the pixel X _{k + 1} . For coefficients X _k and X _{k + 2} that are the pixels before and after the pixel X _{k + 1} ,
Gradient (Y _{k + 2} − of these image data Y _k , Y _{k + 2}
Since the condition is that Y _k ) / (X _{k + 2-} X _k ), the following formula (7) must be satisfied.

F _k ′ (X _{k + 1} ) = (Y _{k + 2} −Y _k ) / (X _{k + 2} −X _k ) (7) Here, each section X _{k−2 to} X _k−1 , X _{k-1 to} X _k , X _k to
The interval between X _{k + 1} , X _{k + 1 to} X _{k + 2} (referred to as lattice interval) is 1
And then, if the position of the interpolation point X _p in the pixel X _{k + 1} directions from the pixel X _k and t (0 ≦ t ≦ 1), equation (1) to (4) and (6), (7) Therefore, f _k (0) = D _k = Y _k f _k (1) = A _k + B _k + C _k + D _k = Y _{k + 1} f _k ′ (0) = C _k = (Y _{k + 1} −Y _{k −1} ) / 2 f _k ′ (1) = 3A _k + 2B _k + C _k = (Y _{k + 2} −
Y _k ) / 2 Therefore, A _k = (Y _{k + 2} -3Y _{k + 1} + 3Y _k -Y _k-1 ) / 2 B _k = (-Y _{k + 2} + 4Y _{k + 1} -5Y _k + 2Y _k-1 ) / 2 C _k = (Y _{k + 1} −Y _k-1 ) / 2 D _k = Y _k Note that the spline interpolation function f _k (x) is X as described above.
Since the variable conversion of = t is performed, f _k (x) = f _k (t). Therefore, the interpolation image data Y at the interpolation point X _p
p can be represented by Y _p = f _k (t) = A _k t ³ + B _k t ² + C _k t + D _k (8). Here, the respective coefficients A _k , B _k , C
Substituting _k and D _k into the equation (8), Y _p = {(Y _{k + 2} −3Y _{k + 1} + 3Y _k −Y _k−1 ) / 2}
_{^{t 3 + {(- Y k}} + 2 + 4Y k + 1 -5Y k + 2Y k-1) /
2} t ² + {(Y _{k + 1} −Y _k−1 ) / 2} t + Y _k , which is the image data Y _k−1 , Y _k , Y _{k + 1} , Y.
_{If k + 2} is arranged, it can be expressed by the following equation (9).

[0054] ^{_{Y p = {(- t 3}} + 2t 2 -t) / 2} Y k-1 + {(3t 3 -5t 2 +2) / 2} Y k + {(- 3t 3 + 4t 2 + t) / 2 } Y _{k + 1} + {(t ³ −t ² ) / 2} Y _{k + 2} (9).

Here, the image data Y _k-1 , Y _k ,
Interpolating coefficients Y _{k + 1} and Y _{k + 2} with interpolation coefficients a _k-1 , a _k , and a
_They are called _{k + 1} and _{ak + 2} . That is, the interpolation coefficients a _k-1 , a _k , a _{k + 1} , a _{k + 2} corresponding to the image data Y _k-1 , Y _k , Y _{k + 1} , Y _{k + 2} in the equation (9) are _{, a k-1 = (-} t 3 + 2t 2 -t) / 2 a k = (3t 3 -5t 2 +2) / 2 a k + 1 = (- 3t 3 + 4t 2 + t) / 2 a k + 2 = (T ³ −t ² ) / 2.

The above calculation is performed for each section X _{k-2 to} X _k-1 , X
By repeating for _{k−1 to} X _k , X _{k to} X _{k + 1} , and X _{k + 1 to} X _{k + 2} , interpolated image data having a different interval from the image data can be obtained for the entire image data.

By the way, the Cubic spline interpolation calculation is
Passing through the original sample points (pixels) as described above,
Obtaining interpolated image data for reproducing a sharp secondary image (image obtained by interpolation) having a relatively high sharpness because the first-order differential coefficient is required to be continuous between each section. On the other hand, B-spline interpolation calculation for obtaining interpolated image data for reproducing a smooth secondary image with relatively low sharpness is also known. This B-spline interpolation operation does not require passing through the original sample points (pixels), but instead, the first derivative and the second derivative (represented by f ″ (X)) are continuous between the sections. Need to be done.

That is, in f _k (x) = A _k x ³ + B _k x ² + C _k x + D _k (1), f _k ′ (X _k ) = f _k−1 ′ (X _k ) (10) f _k ′ (X _{k + 1} ) = f _{k + 1} ′ (X _{k + 1} ) (11) f _k ″ (X _k ) = f _k−1 ″ (X _k ) (12) f _k ″ (X _{k + 1} ) = F _{k + 1} ″ (X _{k + 1} ) (13). However, X _k-1 first-order differential coefficient at the picture element X _k is with respect to the picture elements X _k and X _{k + 1}
For and, the condition is that they match the gradient (Y _{k + 1} −Y _k−1 ) / (X _{k + 1} −X _k−1 ) of these image data Y _k−1 , Y _{k + 1.} , It is necessary to satisfy the following formula (14).

F _k ′ (X _k ) = (Y _{k + 1} −Y _k−1 ) / (X _{k + 1} −X _k−1 ) (14) Similarly, the first derivative of the pixel X _{k + 1} . For coefficients X _k and X _{k + 2} that are the pixels before and after the pixel X _{k + 1} ,
Gradient (Y _{k + 2} − of these image data Y _k , Y _{k + 2}
Since the condition is that Y _k ) / (X _{k + 2-} X _k ), the following formula (15) must be satisfied.

F _k ′ (X _{k + 1} ) = (Y _{k + 2-} Y _k ) / (X _{k + 2-} X _k ) (15) Further, the function f (X) is generally expressed by the following equation (16). It is approximated by things.

F (X) = f (0) + f ′ (0) X + {f ″ (0) / 2} X ² (16) Here, each section X _{k−2 to} X _k−1 , X _{k− 1} ~ X _k , X _k ~
The interval between X _{k + 1} , X _{k + 1 to} X _{k + 2} (referred to as lattice interval) is 1
And then, if the position of the interpolation point X _p in the pixel X _{k + 1} directions from the pixel X _k and t (0 ≦ t ≦ 1) , the equation (10) to (13) and (16), f _{k '(0) = C k =} (Y k + 1 -Y k-1) / 2 f k' (1) = 3A k + 2B k + C k = (Y k + 2 -
_{Y k) / 2 f k "} (0) = Y k + 1 -2Y k + Y k-1 = 2B _{Therefore, A k = (Y k +} 2 -3Y k + 1 + 3Y k -Y k-1) / 6 B _k = (Y _{k + 1} −2Y _k + Y _k−1 ) / 2 C _k = (Y _{k + 1} −Y _k−1 ) / 2 Since D _k is unknown, D _k = (D ₁ Y _{k + 2} + D ₂ Y _{k + 1} + D ₃ Y _k + D ₄ Y
_k-1 ) / 6. Further, as described above, the spline interpolation function f _k (x) performs the variable transformation of X = t, so that f _k (x) = f _k (t). Therefore, f _k (t) = ｛(Y _{k + 2} −3Y _{k + 1} + 3Y _k −Y _k−1 )
/ 6｝ t ³ + ｛(Y _{k + 1} -2Y _k + Y _{k -1} ) / 2｝ t ²
+ ｛(Y _{k + 1} −Y _k−1 ) / 2｝ t + (D ₁ Y _{k + 2} + D ₂
Y _{k + 1} + D ₃ Y _k + D ₄ Y _k-1 ) / 6, which is the image data Y _k-1 , Y _k , Y _{k + 1} , Y
_By rearranging _{k + 2} , it can be expressed by the following equation (17).

F _k (t) = {(− t ³ + 3t ² −3t + D ₄ ) / 6} Y _k−1 + {(3t ³ −6t ² + D ₃ ) / 6} Y _k + {(− 3t ³ + 3t ² + 3t + D ₂ ) / 6} Y _{k + 1} + {(t ³ + D ₁ ) / 6} Y _{k + 2} (17) Here, if t = 1, then f _k (1) = {(D ₄ − 1) / 6} Y _k-1 + {(D ₃ −
3) / 6｝ Y _k + ｛(D ₂ +3) / 6｝ Y _{k + 1} + ｛(D
₁ +1) / 6} Y _{k + 2} Next, the equation (17) for the sections X _{k + 1 to} X _{k + 2} is f _{k + 1} (t) = {(− t ³ + 3t ² −3t + D ₄ ) / _{6} Y k + {(3t} 3 -6t 2 + D 3) / 6} Y k + 1 + {(- 3t 3 + 3t 2 + 3t + D 2) / 6} Y k + 2 + {(t 3 + D 1) / 6 } Y _{k + 3} (18) wherein, if put as _{t = 0, f k + 1} (0) = (D 4/6) Y k + (D 3/6) Y k + 1
_{+ (D 2/6) Y} k + 2 + (D 1/6) Y k + 3 continuity conditions _{(f k (1) = f} k + 1 (0)), and coefficients corresponding to the image data Under the condition that they are equal to each other, D ₄ −1 = 0, D ₃ −3 = D ₄ , D ₂ + 3 = D ₃ ,
D ₁ + 1 = D ₂ , D ₁ = 0, and thus D _k = (Y _{k + 1} + 4Y _k + Y _k-1 ) / 6. _{Therefore, Y p = f k (t} ) = {(- t 3 + 3t 2 -3t + 1) / 6} Y k-1 + {(3t 3 -6t 2 +4) / 6} Y k + {(- 3t 3 + 3t ^{2 + 3t + 1) / 6} } Y k + 1 + {t 3/6} Y k + 2 (19) Therefore, the image data _{Y k-1, Y k,} Y k + 1, Y k + 2
Interpolation coefficients b _k−1 , b _k , b _{k + 1} , b respectively corresponding to
_{k + 2 is, b k-1 = (-} t 3 + 3t 2 -3t + 1) / 6 b k = (3t 3 -6t 2 +4) / 6 b k + 1 = (- 3t 3 + 3t 2 + 3t + 1) / 6 b a ^{_{k + 2 = t 3/6}} .

The above calculation is performed for each section X _{k-2 to} X _k-1 , X
By repeating for _{k−1 to} X _k , X _{k to} X _{k + 1} , and X _{k + 1 to} X _{k + 2} , interpolated image data having a different interval from the image data can be obtained for the entire image data.

The image signal S _S ′ obtained by such interpolation calculation is input to the subtracting means 33 and subtracted from the image signal S. Here, since the image obtained by enlarging the reduced image lacks the detailed information in the original image, the image obtained by subtracting the image signal S _S ′ from the image signal S The signal is a detailed image signal S _D representing a detailed image representing only detailed information of the original image. Then, the detailed image signal S _D obtained in this way is stored in the memory 34 similarly to the reduced image signal S _S.

Then, the enlargement ratio input means (not shown) determines the enlargement ratio determination means 35, and the enlargement / reduction processing is performed as follows.

First, when the enlargement ratio determination means 35 determines that the enlargement ratio is 1 or more, the enlargement / reduction means 40
From the memory 34, the detailed image signal S _D and the reduced image signal S _S
Is read. The reduced image signal S _S is supplied to the interpolation enlargement means 41.
And is subjected to interpolation calculation processing, and an image signal S _S ′ representing an image having substantially the same size as the original image is obtained. The processing performed by the interpolation enlarging means 41 is the same as that of the interpolation enlarging means 32 described above. The image signal S _S ′ thus obtained is added to the detailed image signal S _{D by the} adding means 42. The image signal obtained by adding the reduced image signal S _S ′ thus enlarged and the detailed image signal S _D becomes the image signal S representing the original image. The reason why the image signal S is reproduced when the enlargement ratio is 1 or more is as follows. That is, since the reduced image lacks detailed information as compared with the original image, if the enlargement is performed only based on the reduced image signal S _S , the resolution of the enlarged image will deteriorate. Also, when enlarging the original image, moire due to the grid image becomes inconspicuous.

The image signal S thus obtained is inputted to the interpolating means 43 and outputted as a processed image signal Sproc representing an image enlarged according to the enlargement ratio judged by the enlargement ratio judging means 35, It is reproduced as a visible image by the reproducing means.

On the other hand, when the enlargement ratio judging means 35 judges that the enlargement ratio is less than 1, the enlarging / reducing means 40 reads only the reduced image signal S _S from the memory 34. Then, the reduced image signal S _S read out is reduced in the interpolating means 44 according to the enlargement ratio determined in the enlargement ratio determining means 35 (reduction because the enlargement ratio is less than 1), and in the enlargement ratio determining means 35. It is output as a processed image signal Sproc representing an image reduced according to the determined enlargement ratio, and is reproduced as a visible image by the reproducing means. However, since the reduced image signal S _S has a size that is ¼ of that of the original image, the enlargement ratio in the interpolation means 44 is twice the input value. The reason why the reduced image signal S _S is used when the enlargement ratio is less than 1 is that when an image is reduced, moire due to a grid image becomes noticeable.

As described above, the image processing method and apparatus according to the present invention can deal with various enlargement / reduction ratios by performing only one filtering process by the grid removing filter in the filtering means 31.
It is not necessary to have a plurality of types of grid removal filters, and the configuration of the device can be simplified. Further, since the filtering process only needs to be performed once, it is possible to shorten the calculation time for performing the filtering process and perform the process at high speed.

In the above-described embodiment, the 7 × 7 size filter described above is used as the grid removal filter, but the filter size may be any number. However, if the size of the filter is too large, it takes a long time for the filtering process, so a filter of about 7 × 7 size is most suitable. Further, the filter value of the above filter is not limited to the above value and may have an error of about ± 10%.

[Brief description of the drawings]

FIG. 1 is a diagram showing an image capturing apparatus that captures a radiographic image using a grid.

FIG. 2 is a diagram showing a radiographic image obtained by using a grid.

FIG. 3 is a diagram showing a radiation image reading device.

FIG. 4 is a diagram showing sampling points on a stimulable phosphor sheet.

FIG. 5 is a diagram for explaining an image processing method according to the present invention.

FIG. 6 is a diagram showing details of image processing means.

FIG. 7 is a diagram for explaining an interpolation calculation.

FIG. 8 is a diagram showing spatial frequency characteristics in a direction orthogonal to the striped pattern of the grid image.

9 is a graph showing the graph of FIG. 8 and aliasing of an image signal sampled at a sampling interval corresponding to 2.5 cycle / mm.

10 is a diagram showing a spatial frequency characteristic of an image carried by an image signal sampled at the same sampling intervals as in FIG.

[Explanation of symbols]

 11 Accumulative phosphor sheet 17 Light beam 23 Photomultiplier 26 Log amplifier 27 Analog filter 28 A / D converter 29 Storage unit 30 Image processing means 31 Filtering means 32 Interpolation enlarging means 33 Subtracting means 34 Memory 35 Enlarging ratio judging means 40 Enlargement / reduction means 41 Interpolation enlargement means 42 Addition means 43, 44 Interpolation means

Claims (6)

[Claims] 1. A recording sheet on which an original image composed of a striped grid image corresponding to the grid and a radiation image is recorded by performing imaging using a grid,
An image processing method of reading the original image at a predetermined sampling interval, enlarging or reducing the original image at a predetermined enlargement / reduction ratio to obtain a processed image, wherein the original image is formed at a predetermined pixel interval. A filtering process is performed by a grid removal filter that removes the grid image to obtain a reduced image in which the grid image is removed from the original image and the original image is reduced, and the reduced image is substantially the same as the original image. A detailed image of the original image is obtained by enlarging the image to a size and subtracting the enlarged reduced image from the original image. When the predetermined enlargement / reduction ratio is 1 or more, the reduced image is converted into the original image. And the processed image by enlarging the added image by adding the enlarged reduced image to the detailed image to obtain an added image and enlarging the added image according to the enlargement / reduction ratio. Obtained, wherein when the predetermined scaling factor is less than 1, the image processing method characterized by obtaining the processed image by enlarging or reduced according to the reduced image in the enlarged reduction ratio. 2. A frequency characteristic in which the response is approximately 1 at a frequency of 2.5 cycle / mm or less, and the response is approximately 0 at frequencies of 3.4 cycle / mm and 4.0 cycle / mm. A grid removal filter used in the image processing method according to claim 1. 3. The mask size is 7 × 7, and the filter value is However, each value has an error of ± 10%, The grid removal filter according to claim 2. 4. A recording sheet on which an original image composed of a striped grid image corresponding to the grid and a radiation image is recorded by performing imaging using the grid,
An image processing apparatus for reading the original image at a predetermined sampling interval and enlarging or reducing the original image at a predetermined enlargement / reduction rate to obtain a processed image, wherein the original image is obtained at a predetermined pixel interval. A filtering unit that performs a filtering process by a grid removal filter that removes the grid image to obtain a reduced image in which the grid image is removed from the original image and the original image is reduced; and the reduced image as the original image. A detailed image calculating means for obtaining a detailed image of the original image by enlarging the image to a substantially same size and subtracting the enlarged reduced image from the original image; and when the predetermined enlargement / reduction rate is 1 or more, , The reduced image is enlarged to have substantially the same size as the original image, the enlarged reduced image is added to the detailed image to obtain an added image, and the added image is enlarged or reduced. To obtain the processed image, and when the predetermined enlargement / reduction ratio is less than 1, obtain the processed image by enlarging or reducing the reduced image according to the enlargement / reduction ratio. An image processing apparatus comprising an enlarging / reducing unit. 5. A frequency characteristic in which the response is approximately 1 at a frequency of 2.5 cycle / mm or less, and the response is approximately 0 at frequencies of 3.4 cycle / mm and 4.0 cycle / mm. A grid removal filter used in the image processing apparatus according to claim 4. 6. The mask size is 7 × 7, and the filter value is However, each value has an error of ± 10%, The grid removal filter according to claim 5.