JPS6385877A - X-ray image processor - Google Patents

Abstract

PURPOSE:To repair an image having comparatively small blurring and to attain the variaty of repairing by executing prescribed arithmetic processing including an approximately medium blurring component in a device for correcting data. CONSTITUTION:At first, X-rays are radiated to an object having prescribed width and prescribed X-ray transmissivity and applying an edge response and image data including a scattered X-ray component and a comparatively small blurring component inherent in the device are collected and sent to an arithmetic processing part 10. In such a case, approximately medium blurring component is shielded by the object having the prescribed width. The arithmetic processing part 100 executes the processing of the image data to find out corrected data obtained by correcting the scattered X-ray component and executes the prescribed arithmetic processing of the corrected data to execute operation including the approximately medium blurring component, so that corrected image data including only the comparatively small blurring component inherent in the device can be formed. Consequently, influences caused by the scattered X-rays and a veiling flare component are removed and a blur reparing filter based upon the corrected image data including only the small blurring component inherent in the device can be formed.

Description

[Detailed Description of the Invention] [Objective of the Invention] (Industrial Application Field) The present invention provides an X-ray method for creating an X-ray blur correction filter based on image data collected by irradiating a subject with an edge response with X-rays. The present invention relates to an image processing device.

(Prior Art) In general, in a device that processes and displays X-ray image data obtained by irradiating an object with X-rays, there is a possibility that an edge response may occur due to various factors in the device. A spatial digital filter is applied to correct blurring of various sizes in the image.

By the way, various sizes of blur that occur in the X-ray image of the subject can be caused by the A relatively small blur due to inherent factors, a moderate blur due to Bering glare that occurs when the collected data converted to an optical signal passes through the glass part provided at the final stage of the image intensifier, and the subject. The relatively large blur due to scattered X-rays in the image can be mentioned.

One way to find a spatial digital filter to correct blurring of various sizes is to create an ideal image based on image data obtained by irradiating a subject with edge response with X-rays. A known method is to obtain a blur repair filter using a gradient method using the sum of squares of differences between each image data and ideal image data as an index. A case in which, for example, a one-dimensional blur repair filter is obtained using this method will be briefly described below.

Density B of blurred image data; Let f k(T) be.

Here, i and j represent the addresses of each image data, and k represents the relative position of each filter coefficient of the estimated repair filter from the center filter coefficient.

Therefore, if the filter size is 2+1, -≦≦.

Then, Fi is the density of image data obtained by applying the estimated repair filter f k (T) to image data including blur.

j(season (processed image data density)) is expressed by the following equation (1).

The ideal image data Ir,j and the processed image data Fi,j
(T) and the difference for each image data (difference image data) is Ei,
j(T), this difference image data E i,j (season) can be expressed by the following equation (2).

E i,j(T) = F i,j(T) −I i,j
...(2) And ideal image data I i
. can be expressed as (3) below.

The reason why Ei,j(T) is divided by Ii,j here is to take normalization into consideration.

If the gradient method is adopted as the image data convergence method, the estimated repair filter fk(T+1) at the next stage of the estimated repair filter f k(T) can be expressed by the following equation (4).

By substituting the above equations (1), (2), and (3) into equation (4), the following equation (5) is obtained.

...(5) Here, δ is expressed as an acceleration coefficient.

Therefore, it is possible to obtain a blur repair filter based on equation (5).

However, in conventional devices that create blur repair filters using this method, for example, moderate blur due to Bering glare originating from image intensifiers is not taken into account, and large blur due to scattered X-ray components in the subject is not considered. There is a problem in that even if the component can be removed, it is not possible to separate the medium blur component and the relatively small blur component that remain and create a blur repair filter that includes only the relatively small blur component.

(Problems to be Solved by the Invention) As described above, the conventional apparatus has the problem that it is not possible to perform blur restoration on images with relatively small blur, and it is not possible to achieve diversity in blur restoration work. Rarely.

SUMMARY OF THE INVENTION Accordingly, an object of the present invention is to provide an X-ray image processing apparatus that is capable of performing blur correction on images with relatively small blur and that can achieve a wide variety of blur correction.

[Structure of the Invention] (Means for Solving the Problems) The X-ray image processing device of the present invention eliminates image blur based on image remains data collected by irradiating X-rays onto a subject that gives an edge response. An X-ray image processing device that creates a spatial digital filter to be repaired and applies the spatial digital filter to another image data collected by this device to obtain a deblurred image. Obtain correction data that corrects the scattered X-ray components of the object itself included in the image data obtained by irradiating the object with transmittance and X-rays. The apparatus includes an arithmetic processing unit that performs predetermined arithmetic processing that takes into account blur components, obtains corrected image data that includes only relatively small blur components unique to the device, and uses this corrected image data to create a spatial digital filter. Ru.

(for production) Next, the operation of the device having the above configuration will be explained.

First, this device irradiates X-rays onto a subject that has a predetermined width, a predetermined X-ray transmittance, and provides an edge response, and produces an image containing scattered X-ray components and a relatively small blur component unique to the device. It collects data and sends this image data to the arithmetic processing section. At this time, the medium blur component is blocked by the object having a predetermined width.

The arithmetic processing unit first performs processing on this image data to obtain correction data in which scattered X-ray components are corrected, and then performs predetermined arithmetic processing on the correction data to perform calculations that take into account medium blurring components. Corrected image data containing only relatively small blur components unique to the device is created. Then, the arithmetic processing section roughly uses this modified image data to create a spatial digital filter.

(Example) Examples of the present invention will be described in detail below.

The embodiment shown in FIG. 1 includes an X-ray tube 1 that emits X-rays toward a subject, and an X-ray aperture 2 that determines the X-ray irradiation field.
, an X-ray transparent member 3 that supports the subject and the X-ray shielding material (to be replaced with a subject having bones or organs as necessary), which will be described later, and the I-1 that takes in radiation information and forms an X-ray image; a TV camera 37 that transmits the X-ray image formed on this I-I and converts it into an electrical signal; and aliasing for the output signal of the TV camera 37. (a phenomenon in which high-frequency components fold back into low-frequency components), a predetermined frequency (for example, A/
Nyquist frequency of D converter 8 (-5 MHz as an example)
) a filter 38 that reduces transmission of frequency components below;
A converting the output signal of the filter 38 into a digital signal
/D conversion unit 8, a data storage unit 40 that stores the output data of this A/D conversion unit 8, and image data is taken in from this data storage unit 40, calculation of the average value of image density, scattering
A blur repair filter that performs various operations such as line component correction, predetermined linear operations, and subtraction between multiple image data, and also repairs relatively small blurs based on the results of these operations, as will be described in detail later. an arithmetic processing unit 100 that creates
A D/A converter 27 that converts the output data of the arithmetic processing section 100 into analog data, a display means 28 such as a CRT that takes in and displays the analog data, and a CPU.
It is configured to include an arithmetic control section 29 that controls the entire device, and a bus 30 that transfers various data and control signals.

The data storage unit 40 includes first to sixth memories 9 to 14, and stores the output data of the A/D conversion unit 8 and the image intensifier I as will be described in detail later.
Basing glare component two-point image distribution function at -1
P SF: Po1nt 5predFuncti
It is designed to store measurement data such as on).

The arithmetic processing unit 100 includes first to fifth arithmetic units 22 to 26. Sixth calculation unit 35. 7th to 1st 3rd memories 15 to 21. 14th to 17th memories 31 to 34
．． It is configured to include an 18th memory 36.

The first calculation unit 22 takes in image data from the third to sixth memories 11 to 14, calculates the average value of a specific density area of each of these image data, and sends the result to the seventh memory 15. Send to. It looks like this.

Note that this seventh memory 15 is preliminarily stored with coefficients (α, β) for one-dimensional scattered X-ray data correction, etc.
Parameters necessary for the calculation in step 3 are also stored.

The second calculation unit 23 extracts the coefficient (α
, β) and the third. Image data input from the fourth memory 11.12 (this is referred to as "xi j")
A linear operation expressed as αX:1β is performed on the data, and the results are stored in the eighth memory 16, the twelfth memory 20 . 16th
memory 33° to a seventeenth memory 34.

The third calculation unit 24 takes in the image data taken in from the first memory 9 or the second memory 10 and the image data taken in from the eighth memory 16 and subjected to linear calculation, and processes these images. A subtraction process is executed between the data and the result is stored in the ninth memory 17. Tenth memory 18. The data is sent to one of the eleventh memories 19, and the ninth memory 1
7. The image data in the tenth memory 18 is taken in.

Further, the image data in the eleventh memory 19 is stored in the first calculation unit 22. It is designed to be taken into the second calculation section 23.

The fourth calculation unit 25 takes in image data from the twelfth memory 20 and the eighteenth memory 36, performs predetermined data processing on the image data as will be described later, creates a blur repair filter, and removes the blur. The data of the repair filter is sent to and stored in the thirteenth memory 21, and the data is sent to the eighteenth memory 36.
It is designed to import image data from.

The fifth calculation unit 26 performs a filter calculation using data of the blur correction filter stored in the thirteenth memory 21 on the input image data that is taken in separately from the image data for creating the blur correction filter described above. is executed, and the result is sent to the D/A converter 27, and the fifteenth memory 3
Image data is sent to the fourteenth memory 31, and image data is taken in from the fourteenth memory 31.

The sixth arithmetic unit 35 is configured to take in image data from the sixteenth memory 33 and the seventeenth memory 34, perform addition processing on these data, and send the resultant data to the eighteenth memory 36.

The D/A converter 27 converts the image data subjected to the filter operation into analog data and sends it to the display means 28.

Next, a subject including the X-ray transmitting member 3 will be explained.

The object to be photographed is a thin thin plate-shaped X-ray absorbing member 4 made of aluminum, copper, iron, etc., as shown in Figure 2 (A).
．． 5. An X-ray shielding member made of a thin thin plate of lead or the like to obtain scattered X-ray data in one-dimensional direction as shown in FIG. 2(B). In order to correct the scattered X-ray data as shown in FIG. 2(C), a small piece-shaped X-ray shielding member 6 that can obtain scattered X-ray data in a minute area is used.

Then, the X-ray absorbing member 4 and the X-ray shielding member 5.6 are placed on the X-ray transmitting member 3 in FIGS.
By using only the X-ray transparent member 3 as shown in FIG. 3(D), the objects 70a to 70c, 70e, 70f, 70d.

Now, considering the width of the subject, it is desirable that the width of the subject be larger than the spread of the Bering flare based on the presence of the glass plate offset by δ2 on the output surface of I-1. This is because, as will be explained in detail later, this device obtains a complete edge image of the subject, and based on this, creates an image blurred by Bering glare through simulation, so a perfect edge image is not affected by Bering glare. By being required to be something.

Also, considering the X-ray transmittance of the subject, this
It is desirable that the radiation transmittance approximates the X-ray transmittance of a subject having soft tissues such as bones and organs. In other words, when X-rays are irradiated to a subject with bones or soft tissue, there is no such sudden change in the X-ray transmittance even at these boundaries, and the X-ray transmittance is significantly different from that of such a subject. Using an object is not suitable for creating a spatial digital filter for deblurring.

To give an example of such a subject, as shown in FIG. 6, it is preferable to use a subject that has an X-ray transmittance that is approximately 1/2 that of the O level and the operating level.

Next, the operation of the device having the above configuration will be explained.

First, a description will be given of the steps from the stage of collecting X-ray image data using each of the subjects 70a to 70f to the stage of determining correction data, which is a prerequisite for determining a blur correction filter.

First, X-rays are irradiated from the X-ray tube 1 to the multiplexed objects 70a to 70f shown in FIGS. 3(A) to 3(F), respectively.
The X-ray image is formed by the image intensifier I, and the X-ray image is further shadowed by the TV camera 37 to convert it into an electrical signal, and then filtering is performed by the filter 38.
When the A/D converter 8 performs conversion into a digital signal and collects X-ray image data, the Bering glare components generated on the output surface of the image intensifier I-I are Be cut off. Then, the obtained X-ray image data of the combined photographic objects 70a to 70f are stored in a second memory 110, a fourth memory 12.
Sixth memory 14. First memory9. Third memory 11
．． The data is stored in the fifth memory 13.

Next, the first arithmetic unit 22 reads the X-ray image data of the subject 70f from the fifth memory 13 under the control of the arithmetic control unit 29 and addresses the The image data (this will be referred to as "scattered X-ray component data f") is obtained, and the obtained image data and its address are sent to the seventh memory 15 and stored therein.

Furthermore, under the control of the calculation control unit 29, the first calculation unit 22 calculates the scattered X-ray component data C using the same procedure for the image data of the subject 70c stored in the sixth memory 14. The scattered X-ray component data C and its address are sent to the seventh memory 15 and stored therein.

Next, the first arithmetic unit 22 executes the process as shown in FIG. 3(E) stored in the third memory 11 under the control of the arithmetic control unit 29.
The image data of the subject 70e shown in FIG.
1 line of direction e.

-el is selected, and the image data at the address closest to the addresses of the two X-ray shielding members 6, 6 in the scattered X-ray component data f in this line e□-el (this is referred to as "scattered X-ray data e') and send them to the seventh memory 15 to be stored therein.

Further, the first calculation unit 22 also refers to the address of the scattered X-ray component data C for the image data of the subject 70b shown in FIG. Obtain X-ray component data b.

Now, regarding the image data stored in the third memory 11 and the fourth memory 12, the image data of each line e□ -el, b(1-bl) in FIGS. 3(E) and (B) is For example, the profile is shown in Figure 4 (A>, (B))
It becomes as shown in .

In FIG. 4(A), Fo and Fl are two image data (X-ray intensity) of scattered X-ray component data f, and E2
, E3 are two image data (X
line strength).

Similarly, in FIG. 4(b), Co and C1 are two image data (X-ray intensity) of scattered X-ray component data C,
81 and B2 are two image data (X-ray intensity) of scattered X-ray component data b.

As shown in FIG. 4 (A>), Fo and Fl are E2.
From E3, as shown in FIG. 4(B), Go,
Ox has a larger value than 81.82.

This means that for the subject 70e shown in FIG. 3(E),
X measured at the point corresponding to E2 on eo-C1
Since the radiation shielding member 5 occupies a larger area within the X-ray irradiation field than the X-ray shielding member 6, the scattered X that should originally be measured
This is due to part of the line being cut.

Therefore, it is not appropriate to use the scattered X-ray component data for one line of EO-81 as is, and correction thereof is required.

Such correction is performed as follows.

Ratio between scattered X-ray component data f and scattered X-ray component data b,
That is, FO/E2 in FIGS. 4(A) and (B)
, Fl /E3, Go /81, C1/B2
Calculate the average value of FO/E2 and F1/E3 (this is
Correction coefficient E'. ) and Go /B1 and 01/8
2 (this will be referred to as "correction coefficient B"), and these values are stored in the seventh memory 15.
゛Then, FIG. 3 (E
) The image data of the subject 70e shown in
, the correction coefficient E is taken into the α, and O is taken into the β.
is given, linear density conversion is performed on the captured image data based on these values, and the result is stored in the eighth memory 16.

Next, the image data of the subject 70d shown in FIG. 3(D) stored in the first memory 9 and the corrected one line of scattered The image data is taken into the arithmetic unit 24 of No. 3, where a subtraction process is executed between both image data, and the result is stored in the ninth memory 17.

Similarly, the image data of the subject 70b shown in FIG.
is given O, linear density conversion is performed on the captured image data based on these values, and the result is stored in the eighth memory 16.

Next, FIG. 3(A) stored in the second memory 10
The image data of the subject 70a shown in and the corrected one-line scattered X-ray component data stored in the eighth memory 16 are taken into the third calculation unit 24, and subtraction processing between both image data is executed here. and stores the result in the tenth memory 18.

Through these series of operations, the subject 7 shown in Figure 3 CD) can be photographed.
The image data obtained by removing one line of scattered X-ray components from the image data of 0d and the image data obtained by removing one line of scattered X-ray components from the image data of the subject 70a shown in FIG. 9. 10th memory 17.18
is memorized.

Roughly speaking, the third arithmetic unit 24 is connected to the ninth memory 17 and the first
The image data stored in the 11th memory 18 is taken in, a subtraction process is performed between the two image data, and the result is stored in the 11th memory 19. By this operation, it is possible to eliminate the influence of the positional dependence of the sensitivity of the X-ray detector 3 on the image data.

By the way, among the image data stored in the eleventh memory 19, only the data corresponding to one line is valid. The profile of this effective one-line image data is shown in FIG. 5 (A>).

Further, FIG. 1B shows a profile of image data corresponding to one valid line among the image data stored in the tenth memory 18.

When using multiple lines as image data for determining the blur correction filter, move the X-ray shielding member 3 forward.
By collecting image data again by spatially moving the X-ray shielding members 5, 6 and the X-ray absorbing member 4 while fixed, and performing the series of operations described in (a) above,
Since another line of valid data can be obtained, this process can be repeated as many times as necessary to obtain multiple lines of valid image data. At this time, in the second and subsequent image data convergence, there is no need to collect the image data of the subjects 70a and 70b shown in FIG. 3 (A>) and FIG. 3 (B).

Next, correction of image density non-uniformity based on fluctuations in voltage and current applied to the X-ray tube 1 will be explained.

Even if the X-ray tube conditions (voltage, current) are set the same, the subject is not necessarily exposed to a constant amount of X-rays due to fluctuations in the applied voltage and current, and as a result, the overall displayed image Immediately after the image data is collected, the first arithmetic unit 22 is used to detect a specific density area where it is thought that variations in density between image data will not occur. Then, a certain density based on one image of the first detected image data is used as a standard, and a second density is set for all remaining image data to match this standard density.
The correction can be made by executing density conversion in the calculation unit 23 of.

Now, the image data with one or more valid lines is the first one.
When the information is stored in the memory 19 of No. 1, the next step is performed.

The first calculation unit 22 takes in the image data stored in the eleventh memory 19, and calculates R1 and R2 in FIG. 5 (Δ).
, R3, image densities are determined in areas where there is almost no change in density on both sides of the edge of the object.

That is, if the concentrations in the regions R1, R2, and R3 are respectively Di, Dl, and D3, then the concentration D1
．． The average value of the average value of D3 and the density D2 is determined.

Since Dl and Dl are almost equal, L may be the average value of the density D1 and the density D2.

Alternatively, a region including all the etches of the object may be assumed, and the DC component within that region may be averaged.

Next, the second arithmetic unit 23 takes in the image data stored in the eleventh memory 19, and for α=1. β=
-1 is given, image data is obtained by subtracting the average value from each image data, and this result is stored in the twelfth memory 20 as correction data. Among the image data stored in the twelfth memory 20, an effective one-line profile is shown in FIG. 5(C).
It has positive and negative densities bordering on the level.

When performing this operation, it is sufficient that the densities D1 and D3 are positive and negative, and the average value is not required to be very precise.

Through the above-described operations, the correction data that is a prerequisite for determining the blur correction filter is stored in the twelfth memory 20.

Next, a description will be given of an operation for obtaining corrected image data containing only relatively small blur components from this correction data.

First, complete edge image data of the subject is obtained from the correction data. That is, the fourth calculation section 25 takes in the image data 7 having the profile shown in FIG. The addresses of the portions are determined, the average value of each region R1, R2, and R3 is determined, and the results are sent to the fourteenth memory 31 and stored as complete edge image data. The profile of one line of this complete edge image data is shown in FIG. 5(D).

Next, the principle of creating image data blurred by Bering glare based on this complete edge image data will be briefly explained below.

In advance, the ratio VG to the direct (primary) component in the Bering component on the output side of I-I and the PSF of Bering glare (hereinafter simply referred to as "P") are measured, and this data is stored in the data storage unit 40. Store it in one of the memories.

Then, if the above-mentioned complete edge image data is represented by IE, the image in which this complete edge image data IE causes verifiability glare (this is referred to as "ideal image data IvJ") is expressed by the following equation (6). can be expressed.

...(6) 10, where ■ means a convolution operation.

That is, complete edge image data IE is convolved with the Bering rare PSF given as a filter, and the resulting image data is VG/1+V.
Multiply by G. In addition, complete edge image data■ε is 1/
Multiply by 1+VG. Then, by performing addition processing on one piece of image data obtained as a result of this calculation, ideal image data Iv that takes Bering glare components into consideration can be obtained.

The operation for obtaining this ideal image data Iv will be explained with reference to FIG.

The complete edge image data IE stored in the fourteenth memory 31 is taken into the fifth calculation unit 26, and the calculation control unit 29 reads out the Bering glare PSF from the data storage unit 40, which is also the fifth calculation. Send to Department 26.

Then, the result data of both convolution operations is stored in the fifteenth memory 32. Roughly speaking, this result data and the ratio V stored in the data storage section 40
G is sent to the second arithmetic unit 23, and the second arithmetic unit 23 calculates VG/
An operation of multiplying by 1+VG is executed and the result is stored in the sixteenth memory 33.

Further, the complete edge image data IE stored in the fourteenth memory 31 and the ratio VG stored in the data storage section 40 are sent to the second calculation section 23, and the IE/1+
The calculation VG is executed and the result is stored in the 17th memory 37.
to be memorized. Next, the data stored in the 16th memory 33 and the data stored in the 17th memory 37 are sent to the calculation unit 35 shown in FIG. 18th memory 3 as IV
6 to be memorized.

Through the above operations, the ideal image data Iv including the Bering rare component is obtained.

Next, an operation for finding a blur repair filter for repairing relatively small blur based on this ideal image data Ivy will be explained.

The arithmetic unit 25 in FIG. 4 takes in the ideal image data Iv from the eighteenth memory 36, performs arithmetic processing on the ideal image data Iv by, for example, the gradient method, and converts the ideal image data Iv into which the Bering rare component has been added. A blur correction filter is created using only the corrected image data from which this Bering rare component is removed and includes only small blur components specific to the device, and this result is sent to the third memory 21. This blur correction filter is stored in the thirteenth memory 21 under the control of the arithmetic control section 29.
is memorized.

The blurring repair filter obtained in this way is used for all image data collected by this device. In other words, the target general image data and the brush repair 1! The filter is taken into the fifth arithmetic unit 26, where filter processing is executed, and the result is sent to the D/A converter 27, where it is converted into an analog signal, and then sent to the display unit 28 for display. be done. This makes it possible to obtain high-quality images without blur.

The present invention is not limited to the embodiments described above, and various modifications can be made within the scope of the invention.

For example, since each memory of the arithmetic processing unit 100 is valid for one line of image data, it can be replaced with a memory having a storage capacity for one line.

Furthermore, in the above-described embodiment, a case has been described in which a blur correction filter that includes only small blur components specific to the device is created, but scattered X-ray components are removed based on the correction data stored in the twelfth memory 20. Of course, it is also possible to create a blur correction filter.

[Effects of the Invention] According to the present invention described in detail above, the effects of scattered X-rays and Bering rare components are removed, and a blur repair filter can be created using corrected image data that includes only small blur components unique to the device. , it is possible to provide an X-ray image processing device that can achieve fineness and diversity in blur restoration.

[Brief explanation of the drawing]

Fig. 1 is a block diagram showing an example device of the present invention, and Fig. 2 (A>, (B), and (C) respectively show an X-ray absorbing member and an X-ray shielding member constituting a subject used in the same device. A perspective view, and FIGS. 3(A) to 3(F) are plan views showing the configuration of a subject in the same apparatus, and FIG. 4(A) is an X-ray intensity profile obtained from the subject shown in FIG. 3(E). Figure 4 (
B) is the X-ray intensity profile obtained from the subject shown in Figure 3 (B), Figures 5 (A> and (B) are the density profiles of the image data sent to the subtraction processing section of the same device, respectively),
FIG. 5(C) is the density profile of correction data obtained by the arithmetic processing section of the same device, FIG. 5(D) is the density profile of complete edge image data obtained by the arithmetic processing section of the same device, and FIG. FIG. 2 is an explanatory diagram showing the state of transmission of X-rays to a subject. 40... Data storage unit, 70a to 70f... Subject, 100... Arithmetic processing unit. Agent: Patent attorney Nori Chika, Kanshu, Ogo, Norio (A) (B) Figure 2 (A) Figure 4 CB) (C) Figure 5

Claims (4)

[Claims] (1) Create a spatial digital filter to repair image blur based on image data converged by irradiating X-rays on a subject that gives an edge response, and use the spatial digital filter as another image data collected by this device. An X-ray image processing device that obtains a deblurred image by applying X-rays to a subject having a predetermined width and a predetermined X-ray transmittance, and image data obtained by exposing this subject to In addition to obtaining correction data that corrects the scattered X-ray components of the included object itself, a predetermined calculation process that takes into account the medium blur component of the device is performed on this correction data, and only the relatively small blur component unique to the device is included. An X-ray image processing device comprising: an arithmetic processing unit that obtains corrected image data and uses the corrected image data to create a spatial digital filter. (2) The X-ray image processing apparatus according to claim 1, wherein the width of the subject is greater than the spread of Bering glare that causes moderate blur in this apparatus. (3) The X-ray image processing apparatus according to claim 1, wherein the X-ray transmittance of the subject is approximate to the X-ray transmittance of the tissue of the subject providing the other image data. (4) Predetermined arithmetic processing by the arithmetic processing unit, arithmetic processing to obtain complete edge image data of the subject based on the maternal data, and an ideal image with a moderate blur component added based on this complete edge image data. The X-ray system according to claim 1, which includes arithmetic processing for obtaining data and, based on this ideal image data, arithmetic processing for obtaining corrected image data that includes only relatively small blur components caused by the apparatus. Image processing device.