JP5155020B2 - Radiographic imaging apparatus and radiographic imaging method - Google Patents

Description

  The present invention relates to a radiation image capturing apparatus and a radiation image capturing method for irradiating a subject with radiation, detecting the radiation transmitted through the subject, converting the detected radiation into an electrical signal, and generating a radiation image based on the converted electrical signal. is there.

  Radiographic imaging devices are used in various fields including, for example, medical diagnostic images and industrial nondestructive inspection. As a radiation detector for detecting radiation (X-rays, α-rays, β-rays, γ-rays, electron beams, ultraviolet rays, etc.) transmitted through a subject in a radiographic imaging device, at present, a flat panel that converts radiation into electrical signals There is a type using a flat panel detector (FPD).

  In a radiographic radiation imaging apparatus using an FPD, a subject is irradiated with radiation from a radiation source, the radiation transmitted through the subject is converted into an electrical signal by the FPD, and an electrical signal corresponding to the image data of the subject is read from the FPD. Generate an image.

  Here, the FPD has a configuration in which a plurality of subareas divided in at least one of a readout unit or a detection element unit are connected, and a readout circuit for image data is provided for each subarea. However, there is an individual difference in the sensitivity of each sub-area and the amplification characteristics of the readout circuit, causing an offset at the boundary between adjacent sub-areas, resulting in a step (artifact) in the signal intensity (luminance) of the radiographic image. was there.

  On the other hand, for example, Patent Document 1 evaluates image data of a plurality of sensor elements for forming image data, a reading unit related to an image subarea, and a concatenated image area of adjacent subareas, and correction data. An analysis unit arranged to generate a non-linear image, and a correction unit arranged to correct illegal image data by the correction data, so that the user does not need a new X-ray administration or an intermediary, and is non-linear. An image forming apparatus is described that allows regular and accurate recalibration or correction of amplification features.

  Further, Patent Document 2 includes radiation irradiating means for irradiating a subject with radiation and radiation detecting means for detecting radiation transmitted through the subject, and based on a radiation detection signal detected from the radiation detecting means. A radiographic imaging device that obtains a radiographic image and calculates a statistic regarding the distribution of the signal level of the pixel based on the detected radiation detection signal, and signals between the two regions divided by the configuration of the radiation detection means When a level difference appears, a statistic calculation means for calculating a statistic for each region, and a correction amount related to the difference in the statistic between the two regions so as to cancel the signal level difference A radiation imaging apparatus is provided that includes a pixel correction means that corrects each pixel by reducing the signal level difference.

JP 2001-197313 A JP 2006-296722 A

  Here, in the apparatus described in Patent Document 1, in order to analyze the amplification characteristic for each reading unit, a correction amount is calculated by obtaining a histogram for each sub-area. Therefore, this apparatus can correct the step of the image signal caused by the uniform fluctuation in the sub-area, but the step of the signal strength of the image caused by the non-uniform fluctuation in the sub-area differs depending on the boundary position of the sub-area. Therefore, there is a problem that the step of the signal intensity of the image due to non-uniform fluctuation in the sub-area cannot be eliminated.

  Further, even in an apparatus that calculates a statistical value of each sub-area and calculates a correction value from the difference between the statistical values as in the apparatus described in Patent Document 2, the statistical amount is based on the vicinity of the division boundary. Since the entire sub-area pixel in the direction perpendicular to the division boundary is not a statistical quantity but a population, it cannot be said that the step amount can be calculated appropriately. In addition, since the continuity of the correction amount regarding the direction parallel to the division boundary is not taken into consideration, there may be a streak-like artifact perpendicular to the division boundary in the corrected image.

Furthermore, if the correction value is calculated according to the step amount at the boundary and the pixel density, the edge pattern area composed of edge patterns where the pixel density of the radiation image changes abruptly overlaps the boundary. There is also a problem that a change in pixel density due to the edge pattern is detected and corrected as a step amount.
As described above, if the difference in pixel density caused by the edge pattern is detected as a step amount and corrected, the edge pattern disappears from the radiographic image, resulting in an image different from the actual image.

Even if the edge pattern area can be recognized and erroneous correction to the edge pattern can be prevented, the edge pattern area and the boundary overlap with each other unless the continuity with the boundary of the non-edge pattern area is taken into consideration (inappropriate) On the streak extending in the direction perpendicular to the boundary of the sub-area at the boundary between the edge pattern area and the part where the edge pattern area and the boundary do not overlap (part where the step is corrected with an appropriate value) There is also a problem that this artifact occurs.
That is, as shown in FIGS. 8A and 8B, when a part of the edge pattern ed of the image substantially overlaps the boundary b of the sub area and extends substantially in parallel, the edge There is a problem in that streak-shaped artifacts af extending in the direction perpendicular to the edges are generated at both ends (or one end portion) of the overlapping region where the pattern ed and the sub-area boundary b overlap. 8A is a schematic diagram conceptually showing the relationship between the artifact, the boundary, and the edge pattern, and FIG. 8B is an enlarged schematic diagram showing the periphery of the boundary in FIG. 8A in an enlarged manner. FIG.

  An object of the present invention is to solve the problems based on the above prior art, prevent the occurrence of artifacts at the boundary between subareas and prevent radiation from occurring, and a radiographic image capturing apparatus and radiographic image capable of capturing radiographic images with high accuracy It is to provide an imaging method.

In order to solve the above-described problems, the present invention is provided with a detection unit in which a plurality of subareas each including a plurality of detection elements for detecting radiation are arranged and corresponding to each subarea, and a detection element in a corresponding subarea. A radiation detector having a plurality of reading circuits for reading the generated detection signals, imaging data processing means for generating image data from the detection signals respectively read by the plurality of reading circuits, and the imaging data processing means Step correction means for correcting a step of pixel density generated at the boundary between the sub-area of the image data and a sub-area adjacent to the sub-area based on image data in the vicinity of the boundary, The correction means includes a step amount detection unit that detects a step amount of pixel density that occurs at a boundary between the sub area and a sub area adjacent to the sub area. An edge region determination unit that determines whether an image at a position corresponding to the boundary is an image of an edge pattern region extending substantially parallel to the boundary from the step amount detected by the difference amount detection unit, and the step amount detection A step correction unit that calculates a correction amount based on the step amount detected by the unit and corrects the detection signal based on the calculated correction amount, wherein the step correction unit corrects the region determined to be the edge pattern region A quantity is calculated based on a correction amount of an area not determined to be the edge pattern area in contact with an edge of the area determined to be the edge pattern area, and image data in the edge pattern area is corrected. An image pickup apparatus is provided.

Here, when the one end of the area determined as the edge pattern area coincides with the edge of the image, the step correction unit only corrects the correction amount of the area in contact with the other end of the area determined as the edge pattern area. It is preferable to calculate the correction amount of the area determined as the edge pattern area based on the above.
In addition, when the both ends of the area determined as the edge pattern area do not coincide with the edges of the image, the step correction unit determines the edge pattern based on the correction amount of the area contacting each of the both ends of the area determined as the edge pattern area. It is also preferable to calculate the correction amount of the area determined as the area.
Further, it is preferable that the step correction unit calculates a correction amount by linear interpolation using the step amounts at both ends.
The step correction unit preferably calculates the correction amount by nonlinear interpolation using the step amounts at both ends.

Further, the step amount detection unit calculates a difference between the pixel density of the image data near the boundary of the sub area and the pixel density of the image data near the boundary of the sub area adjacent to the sub area in a direction parallel to the boundary. It is preferable to calculate the level difference by smoothing.
The smoothing process is at least one of a process using a median filter and a moving average process.

Moreover, it is preferable that the said edge area | region determination part determines an edge pattern area | region by threshold value determination of the amount of steps.
Furthermore, it is preferable that the threshold value is a value near the maximum value of the step amount not caused by the edge pattern.
Further, it is preferable that the step amount detection unit calculates a statistic of pixel density of image data in the vicinity of the boundary of each sub-area, calculates a difference from the statistic value, and calculates a step amount.
Moreover, it is preferable that the level difference detection unit calculates the statistic by a median filter.
Moreover, it is preferable that the level difference detection unit calculates an average value of pixel densities of image data in the vicinity of the boundary of each sub-area as a statistic.
The statistic is preferably a value calculated from detection signals respectively detected from a plurality of detection elements arranged in a direction orthogonal to the boundary.
Further, it is preferable that the step amount detection unit calculates a step amount from a difference between pixels adjacent to another sub area.
Moreover, it is preferable that the level | step difference correction | amendment part calculates the correction amount which attenuated correction intensity | strength as it left | separated from the said boundary.

The step correction unit preferably calculates a correction amount according to the pixel density of the image data before correction.
The step correction unit preferably calculates a correction amount that does not cause an underflow due to overcorrection.
Further, the step correction unit calculates the correction amount by addition when the pixel density of the image data before correction is larger than the pixel density of the image data used for calculating the step amount, and used for calculating the step amount. When it is smaller than the pixel density of the image data, the correction amount is preferably calculated by multiplication.

It is preferable that the detection element accumulates a charge corresponding to the radiation dose, and the reading circuit reads the charge of the detection element as the detection value.
The detection element includes a fluorescent element that fluoresces according to an irradiation dose of radiation, and a charge storage element that accumulates charges according to light emission of the fluorescent element,
The reading circuit preferably reads the charge of the charge storage element as the detection value.

In order to solve the above-described problem, the present invention provides a detection unit in which a plurality of subareas each including a plurality of detection elements for detecting radiation are arranged and corresponding to each subarea, and detection of the corresponding subarea. A radiation image reading step for reading a radiation image with a radiation detector having a plurality of reading circuits for reading detection signals generated by the elements, and image data generation for generating image data from the detection signals respectively read by the plurality of reading circuits A step amount of pixel density generated at a boundary between the step and the sub area and a sub area adjacent to the sub area is detected, and an image at a position corresponding to the boundary is substantially parallel to the boundary from the detected step amount. determining whether the image of the extending edge pattern regions, and calculates a correction amount based on the detected step amount, based on the calculated correction amount A step correction step for correcting the output signal, wherein the step correction step includes a correction amount of the area determined to be the edge pattern area and the edge pattern area in contact with an end of the area determined to be the edge pattern area. The present invention provides a radiographic imaging method characterized in that it is calculated based on a correction amount of a region that has not been determined, and image data in an edge pattern region is corrected.

  According to the present invention, even when the edge pattern region, which is a region where the pixel value of the radiation image is abruptly changed, and the boundary of the sub area overlap, the edge pattern of the radiation image is retained and the boundary of the sub area is maintained. Generation of a step can be prevented. As a result, the occurrence of artifacts can be prevented and the edge pattern can also be retained, and a highly accurate radiation image can be taken.

  A radiographic image capturing apparatus and a radiographic image capturing method according to the present invention will be described in detail based on embodiments shown in the accompanying drawings.

Figure 1 is a block diagram showing the schematic configuration of one embodiment of a radiation imaging apparatus of the present invention using the radiation image capturing method of the present invention, FIG. 2 is a schematic of an FPD (Flat Panel Detector) as shown in FIG. 1 It is a schematic diagram which shows a structure.

  A radiographic image capturing apparatus (hereinafter simply referred to as “imaging apparatus”) 10 shown in FIG. 1 irradiates a subject with radiation, and obtains a detection signal for detecting the radiation irradiated (transmitted) to the subject. A photographing data processing means 14 for digitally converting a detection signal (a signal corresponding to image data) acquired by the imaging means 12, a step correction processing means 15 for performing a step correction process on the digitally converted image data, and a step correction. Image processing means 16 for performing image processing on the processed image data, output means 18 for outputting the image processed image data, photographing instruction means 20 for inputting a photographing instruction to the imaging means 12 and the like, and control for controlling each part And means 22. The imaging apparatus 10 irradiates a subject (subject) H with radiation, detects radiation transmitted through the subject H with a detection element, and based on a signal detected by the detection element, radiation obtained by imaging the subject H Generate an image.

  The imaging unit 12 includes an irradiation control unit 24, a radiation source 26, an imaging table 28, and a radiation detection unit 30. The imaging unit 12 irradiates the subject H with radiation and detects the radiation transmitted through the subject H, thereby detecting the subject. A radiographic image of H is taken. The imaging means 12 outputs data (analog data) of a radiographic image obtained by imaging the subject H.

The radiation source 26 is a radiation irradiation mechanism that generates radiation and irradiates the imaging table 28 with radiation. Radiation emitted from the radiation source 26 passes through the subject H on the imaging table 28 and enters the radiation detection unit 30.
Here, as the radiation source 26, various radiation irradiation mechanisms used as the radiation source of the imaging apparatus can be used. In the present invention, X-rays, α rays, β rays, γ rays, electron beams, ultraviolet rays, and the like can be used as radiation.
The irradiation control unit 24 drives the radiation source 26 to control the irradiation amount so that the radiation with the intensity set according to the imaging mode is irradiated for the set time.
The imaging table 28 is a support table that is disposed at a position facing the radiation source 26 and on which the subject H is placed on the surface on the radiation source 26 side. The imaging table 28 is a normal imaging table used in various types of radiographic imaging devices.

  The radiation detection unit 30 has an FPD 32, detects the radiation transmitted through the subject H with the FPD 32, and converts it into an electrical signal (detection signal, radiation image data). The radiation detection unit 30 outputs analog radiation image data (hereinafter, also simply referred to as “image data”) obtained by photographing the subject H.

As shown in FIG. 2, the FPD 32 is configured by connecting five subareas 42a to 42e, and image data readout circuits 44a to 44e are provided for the respective subareas 42a to 42e. To 42e and read circuits 44a to 44e are connected by read lines 46a to 46e. In the following, the direction parallel to the boundaries 48a to 48d of the two subareas (that is, the tangent line between the subarea and the adjacent subarea) is defined as the main direction (the left-right direction in FIG. 2), and the direction orthogonal to the main direction. Is the sub-direction (vertical direction in FIG. 2).

  The sub-areas 42a to 42e are configured by a plurality of reading pixels (also referred to as detection elements) arranged in an array in a two-dimensional direction (that is, each of the main direction and the sub-direction). Each reading pixel has a radiation-sensitive film and generates a detection signal corresponding to the radiation dose.

  The read circuits 44a to 44e are circuits for reading the detection signals generated in the corresponding subareas 42a to 42e, respectively. For example, a semiconductor integrated circuit (ASIC) designed and manufactured for a specific application can be used. .

  In FIG. 2, only one line is representatively displayed. However, a plurality of read lines 46a to 46e are arranged in the sub direction for each sub area, and the read lines corresponding to the sub areas 42a to 42e and each sub area are read out. The circuits 44a to 44e are connected to each line extending in a direction parallel to the main direction of the sub areas 42a to 42e.

  When image data is read from the FPD 32, image data for one row in the sub-direction is obtained by all the read circuits 44a to 44e via all the read lines 46a to 46e of all the sub areas 42a to 42e. Read simultaneously. Further, the image data for one screen is read by repeating the reading of the image data for one row over all the pixels in the sub direction while changing the position in the main direction, the number of times corresponding to the number of rows.

  Here, as the FPD 32, a direct type FPD that directly converts radiation into electric charge and generates a detection signal, or radiation is once converted into light, and the converted light is further converted into an electric signal to generate a detection signal. Both indirect FPDs that are generated can be used. Further, an optical reading type FPD that detects radiation by a method different from the known direct method and indirect method proposed by the present applicant in Japanese Patent Application No. 2007-218816 and generates a detection signal can also be used. That is, various methods can be used as a method of sensing radiation by the subarea and the readout circuit and generating a detection signal from the sensed radiation.

  Here, the direct type FPD includes, for example, a photoconductive film such as amorphous selenium, a capacitor, a TFT (Thin Film Transistor) as a switch element, and the like. For example, when radiation such as X-rays is incident, electron-hole pairs (e-h pairs) are emitted from the photoconductive film. The electron-hole pair is accumulated in the capacitor, and the electric charge accumulated in the capacitor is read out as an electric signal through the TFT.

  On the other hand, the indirect FPD is composed of, for example, a scintillator layer made of a phosphor such as “CsI: Tl”, a photodiode, a capacitor, and a TFT. For example, when radiation is incident, the scintillator layer emits light (fluoresce). Light emitted by the scintillator layer is photoelectrically converted by a photodiode and accumulated in a capacitor, and the electric charge accumulated in the capacitor is read out as an electrical signal through the TFT.

  Furthermore, an optical reading type FPD can be summarized as follows: a recording photoconductive layer that exhibits conductivity when irradiated with radiation (recording light) and a charge pair when irradiated with reading light. Thus, a photoconductive layer for reading exhibiting conductivity, a substrate having transparency to the reading light, and the like are laminated in this order. Further, the optical reading type FPD is provided with a line light source for sequentially irradiating the substrate side with reading light for one line when the accumulated charge is read out.

  In an optical reading type FPD, when line-shaped reading light is irradiated from a line light source, recording is performed on a portion irradiated with reading light in image information recorded as a latent image polarity charge accumulated in a power storage unit. The image information for one line is read out as an electric signal of a level corresponding to the amount of latent image polarity charge for each pixel through the transparent linear electrode. By performing this process for all lines, image information for one screen is read as image data.

The imaging unit 12 is configured as described above.
The radiation source 26 and the radiation detection unit 30 may be configured to be capable of reciprocating along the longitudinal direction (left and right direction in FIG. 1) of the imaging table 28 for, for example, long imaging. . The imaging unit 12 may further include a moving unit for the radiation source 22, a lifting unit for the imaging table 28, a moving unit for moving in the horizontal direction, a tilting unit for tilting the imaging table 28, and the like.

  Next, the imaging data processing unit 14 is a part that performs data processing such as A / D (analog / digital) conversion on the radiation image data supplied from the imaging unit 12. The photographic data processing means 14 outputs digital image data after data processing.

  The step correction processing means 15 includes a step amount calculation unit 52, a smoothing processing unit 54, a ½ multiplication unit 56, an edge region determination unit 58, and a step correction unit 60. The step correction processing unit 15 performs a step correction process on the radiographic image data after the data processing output from the imaging data processing unit 14. The step correction processing means 15 outputs the data of the radiographic image after the step correction.

  The step amount calculation unit 52 compares the pixel density (that is, the signal strength of the image) at a specified position in each sub-area of the radiographic image data output from the imaging data processing unit 14, and compares the adjacent two sub-areas. The amount of step at the boundary of the area is calculated. Here, the step amount calculation unit 52 calculates the step amount at each position of the boundary between two adjacent subareas, thereby absorbing the locality at each position of the boundary between two adjacent subareas with high accuracy. A step amount at each position is calculated.

  The smoothing processing unit 54 smoothes the step amount calculated by the step amount calculating unit 52 using a median filter, and removes noise (more specifically, high-frequency noise) from the step amount.

  The 1/2 multiplier 56 multiplies the step amount smoothed by the smoothing processor 54 by 1/2 (that is, multiplies by 0.5). As described above, the step amount to be allocated to each of two adjacent subareas is calculated by halving the step amount by the ½ multiplier 56.

  Based on the step amount calculated by the ½ multiplier 56, the edge region determination unit 58 has an image at a position corresponding to the boundary between two adjacent subareas substantially parallel to the boundary between two adjacent subareas. It is determined whether the edge pattern image extends. That is, the edge area determination unit 58 determines whether the image at the position corresponding to the boundary is an image of the edge pattern area extending substantially parallel to the boundary.

The step correction unit 60 corrects the step of the image data based on the calculated step amount.
Specifically, the step amount that is halved by the ½ multiplier 56 from the corresponding pixel from the image data of each pixel of the readout line of one of the adjacent subareas (the one with the higher signal strength of the image) Occurs at the boundary by subtracting and adding the step amount halved by the ½ multiplier 56 from the pixel corresponding to the image data of each pixel of the other line (the lower signal strength of the image) The level difference of the image signal to be corrected is corrected.
Further, the level difference correcting unit 60 corrects the level difference of the image data of the adjacent subarea using the correction amount that is changed so that the ratio of the correction amount to the signal intensity of the image decreases as the distance from the boundary increases. Specifically, the step correction unit 60 uses a correction coefficient that changes according to the distance from the boundary (the value decreases as the distance from the boundary decreases), and applies the correction coefficient to the step amount for each position from the boundary. A correction amount is calculated from the calculated value, and the level difference of the image data is corrected using the calculated correction amount.
By adjusting the correction amount according to the distance from the boundary, it is possible to more reliably prevent the occurrence of image unevenness.
The level difference correcting means 15 is basically configured as described above.

  The image processing unit 16 is a part that performs various types of image processing including contrast enhancement and edge enhancement processing on the image data in which the level difference is corrected by the level difference correction processing unit 15. The image processing means 16 is configured by a program (software) operating on a computer, dedicated hardware, or a combination of both. Image data after image processing is output from the image processing means 16.

  The output unit 18 is a part that outputs image data after image processing supplied from the image processing unit 16. The output means 18 is, for example, a monitor that displays a radiation image on a screen, a printer that prints out a radiation image, a storage device that stores radiation image data, and the like.

  Here, in the imaging apparatus 10, predetermined imaging conditions are set in advance in addition to a manual imaging mode in which imaging conditions such as radiation intensity and irradiation time (irradiation amount) are manually set as imaging modes. Multiple shooting modes are provided.

  The photographing instruction means 20 is a part that sets photographing conditions and a photographing mode and instructs photographing of the subject H. As the shooting instruction means 20, an input key for setting shooting conditions and a shooting mode and a shooting button of a two-step push type are used for shooting instructions. When the shooting button is pressed down to the first level, it is ready for shooting, and when it is pressed down to the second level, shooting starts. The shooting instruction means 20 outputs shooting instruction signals indicating shooting conditions, shooting modes, and shooting button states.

The control unit 22 is a part that controls the operation of the imaging apparatus 10 in accordance with the imaging instruction signal supplied from the imaging instruction unit 20. For example, the control unit 22 performs shooting control in the imaging unit 12, image processing control in the image processing unit 16, and output control in the output unit 18.
The imaging device 10 is basically configured as described above.

Next, the operation of the imaging apparatus 10 will be described.
Here, FIG. 3 is a flowchart showing the operation of the step correction means 15 of the imaging apparatus, and FIG. 4 is a flowchart showing the operations of the edge region determination unit 58 and the step correction unit 60. In the following embodiments, a case where a pixel value (or LSB) is used as a pixel density (intensity of a detection signal) will be described as an example.

  First, in the shooting instruction unit 20, when a shooting condition or shooting mode is set by the user and the shooting button is pushed down to the first level, the imaging apparatus 10 enters a shooting ready state under the control of the control unit 22.

  Subsequently, shooting is started when the shooting button is pushed down to the second level. When imaging is started, the imaging unit 12 emits radiation with intensity set according to imaging conditions or imaging modes from the radiation source 26 for a set time. The irradiated radiation passes through the subject H on the imaging table 28 and enters the FPD 32 of the radiation detection unit 30, and the radiation transmitted through the subject H is converted into an electrical signal (radiation image data).

  Subsequently, the captured radiographic image data is read from the FPD 32, and data processing such as A / D conversion is performed by the imaging data processing unit 14 to generate digital image data.

Next, the step correction process is performed on the image data generated by the photographing data processing unit 14 by the step correction unit 15.
Hereinafter, this will be specifically described with reference to FIG.

First, the image data generated by the photographing data processing unit 14 is divided for each position of the sub area, and the step amount calculation unit 52 reads out the read line (boundary at a predetermined position) with reference to the boundary between the sub area and the adjacent sub area. The step amount at each position in the main direction is calculated from the signal intensity of the image (image data) at each position (line parallel to) (step S10). The relationship between each position and the step amount is collectively used as step amount data.
Here, the step amount is calculated by subtracting the signal intensity of the image in the sub area below the boundary from the signal intensity of the image in the sub area above the boundary. Therefore, when the signal strength of the image in the sub-area above the boundary is higher than the signal strength of the image in the sub-area below the boundary, the step amount becomes a positive value, and in the opposite case, the step amount Is negative. In the present embodiment, the step amount is a value obtained by subtracting the signal intensity of the image in the sub area below the boundary from the signal intensity of the image in the sub area above the boundary. The reference sub-area) is not particularly limited, and a value obtained by subtracting the signal intensity of the sub-area image above the boundary from the signal intensity of the sub-area image below the boundary may be used as the step amount.
Next, the level difference calculated is smoothed by the smoothing processing unit 54. (Step S12).
In other words, a median filter is used for the calculated step amount, and smoothing processing is performed using the step amount at the target position and the step amounts at a plurality of positions adjacent to the target position in a direction parallel to the main direction.
Next, the smoothed level difference data is multiplied by 1/2 by the 1/2 multiplier 56 (step S14).
Next, using the step amount multiplied by 1/2, the edge area determination unit 58 and the step correction unit 60 create correction data for the sub-area image above the boundary (step S16).
Specifically, the edge area determination unit 58 determines whether or not the image at the boundary is an image of the edge pattern area extending substantially parallel to the boundary between the two sub-areas based on the step amount (that is, the boundary and the image). Whether or not the edge pattern overlaps with a certain length or more), and a correction amount is calculated based on the determination result. The calculation method of each correction amount will be described in detail later.
Such calculation for calculating the correction amount is repeated, and the correction amount at each position in the sub-area above the boundary is calculated to obtain correction data for the sub-area above the boundary.
Next, the step amount data halved in step S14 is multiplied by -1 (step S18).
Next, using the step amount data multiplied by −1, the edge area determination unit 58 and the step correction unit 60 create correction data for a sub-area image below the boundary (step S20).
Specifically, as in step S16, it is determined whether or not the boundary and the edge pattern overlap by a certain length or more, and the correction amount is calculated based on the determination result.
Next, the step amount of the image data is corrected using the correction data for the sub-area above the boundary calculated in step S16 and the correction data for the sub-area below the boundary calculated in step S20 ( Step S22).
Specifically, among adjacent subareas, the correction value of each corresponding pixel is subtracted from the image data in each pixel in the main direction of the readout line closest to the boundary of the subarea with the higher signal strength of the image, The correction value of each pixel corresponding to the image data in each pixel in the main direction of the readout line closest to the boundary of the subarea with the lower image signal intensity is added.
Subsequently, the intensity is set so that the signal strength of the image becomes weaker as it moves away from the boundary in the sub-direction (away from the boundary in the upper sub-area and away from the boundary in the lower sub-area). By multiplying the intensity correction coefficient of the correction table, the image data at all positions in the adjacent upper and lower subareas is corrected according to the signal intensity of the image of each pixel in the main direction and the sub direction.

  In other words, the image data at all positions in the adjacent upper and lower sub-areas are changed and corrected by the step correction means 15 so that the correction value decreases as the distance from the boundary in the sub-direction increases. Thereby, the level | step difference of the image signal in a boundary can be corrected smoothly.

  According to the above procedure, the same processing is performed on all the borders of adjacent subareas, so that even in the case of an image in which the signal strength of the image changes greatly at the border between adjacent subareas, the level difference of the image signal Can be corrected accurately.

Next, the image processing means 16 performs image processing such as contrast enhancement and edge enhancement on the image data whose level difference is corrected.
Further, the image data after the image processing is output by the output means 18. Specifically, when the output means is a monitor, it is displayed on the monitor, and when the output means is a printer, it is printed out.

Next, a method of creating correction data by the edge area determination unit 58 and the step correction unit 60 will be described with reference to FIGS. 4 and 5.
Here, FIG. 4 is a flowchart showing the operations of the edge region determination unit 58 and the step correction unit 60. FIG. 5 is a front view showing the coordinates of each position in two adjacent subareas. In this embodiment, as shown in FIG. 5, the main direction (direction parallel to the boundary) is the x direction, and the sub direction is the y direction.
In this embodiment, the y coordinate of the boundary between two adjacent subareas is set to iMid, and the reference line of the subarea above the boundary used for calculation of the step amount is moved iSm above the boundary. A reference line used to calculate a step amount and used to calculate a step amount in a sub-area below the boundary is a line moved iSp above the boundary.
Further, the difference between one reference line and the other reference line of two adjacent subareas is taken, and a step amount obtained by halving the difference is defined as step [x]. Also, the pixel value at each position is Dt [x] [y].
In the following description, a correction value is calculated at a position (iM, iS) that is a fixed distance away from the reference line.
The correction value is repeatedly calculated from one end of the sub-area toward the other end in the direction parallel to the boundary.

  First, it is determined whether the pixel value Dt [iM] [iS] at a position spaced apart from the reference line by a certain distance (hereinafter referred to as “target position”) is greater than or equal to the pixel value Dt [iM] [iMid + iSm] at the step calculation position. (Step S30). That is, it is determined whether or not Dt [iM] [iS] ≧ Dt [iM] [iMid + iSm] is satisfied.

If Dt [iM] [iS] ≧ Dt [iM] [iMid + iSm], the correction amount is set to step [iM] (step S32), and the process proceeds to step S36.
That is, the step amount calculated at the reference position is set as the correction amount.
If Dt [iM] [iS] <Dt [iM] [iMid + iSm], the correction amount is Dt [iM] [iS] × step [iM] / D [iM] [iMid + iSm] (step S34), and step S36. Proceed to

Here, FIG. 6A is a graph showing the relationship between the pixel value before correction and the pixel value after correction when the step amount is positive, and FIG. 6B shows the case where the step amount is negative. It is a graph which shows the relationship between the pixel value before correction | amendment, and the pixel value after correction | amendment. In FIGS. 6A and 6B, the horizontal axis is the pixel value before correction, and the vertical axis is the pixel value after correction.
In FIGS. 6A and 6B, a thick line is a line indicating a relationship between a pixel value before correction and a pixel value after correction.
As described above, the calculation method of the correction amount is switched depending on the size of the pixel value in step S30, step S32, and step S34, and as shown in FIG. 6A and FIG. Can prevent underflow from occurring by reducing the correction amount proportionally.

After calculating the correction amount, it is determined whether or not the absolute value of the step amount step [iM] is equal to or greater than the threshold value f _UPPER (step S36). That is, it is determined whether | step [iM] |> f _UPPER is satisfied.
Here, the threshold value f _UPPER is a value near the maximum value calculated as the step amount between the sub-areas. By setting the threshold value to such a value, the step amount step [iM] exceeding the threshold value is not a step caused by the step in the sub area, but a step caused by the edge pattern of the image (here, It can be determined that the pixel density difference of the image itself.

If | step [iM] |> f _UPPER, it is determined that the boundary serving as a reference for the step correction of the target position overlaps the edge pattern of the image, and the process proceeds to step S38.
If it is determined that the reference boundary overlaps the edge pattern of the image, it is determined whether or not there is the other end of the edge region (step S38). That is, it is determined whether or not there is a region that is not an edge pattern in the region opposite to the end portion on the region side where the correction value has already been calculated.
If it is determined that there is the other end portion of the edge pattern region, a correction amount is calculated based on the correction amounts of the regions that are not two edge patterns adjacent to the respective end portions of the edge pattern region (step S40). Proceed to step S44.
If it is determined that there is no other end portion of the edge pattern region, the correction amount is based on the correction amount of one end portion of the edge pattern region (the non-edge pattern region adjacent to the already detected edge pattern region). Is calculated (step S42), and the process proceeds to step S44.
If | step [iM] | ≦ f _UPPER, it is determined that the boundary serving as a reference for correcting the step at the target position does not overlap the edge pattern of the image, and the process proceeds to step S44.
In step S44, the correction amount calculated in any of step S32, step S34, step S40, or step S42 is multiplied by the correction intensity table value, and the calculated correction amount is stored in the correction data.
Here, the correction intensity table value is a coefficient stored for each distance from the boundary, and is a coefficient that gradually decreases as the distance from the boundary increases.
If the correction amount at the target position is calculated in step S44, it is determined whether the calculation of the correction values for all the image data has been completed (step S46).
If the calculation of the correction amount for all the image data in the sub-area has not been completed, the process proceeds to step S30, and the correction amount is calculated using the position where the correction amount is not calculated as the target position.
On the other hand, when calculation of the correction amount of all image data in the sub-area has been completed, the step to be corrected is moved to the next step and the sub-area adjacent to the step (that is, the cause of the step). The correction amount of the image data in the two sub-areas constituting the boundary is calculated in the same manner. This is performed for all the steps, and the correction amount of the entire image is calculated.

  In this way, it is determined whether the edge pattern of the image overlaps the boundary between two adjacent subareas. If it is determined that the edge pattern overlaps the boundary, the portion of the edge pattern that is not in contact with the edge pattern By calculating the correction amount based on the correction value, it is possible to correct the step generated at the boundary of the sub-area while preventing the edge pattern from being corrected as a step. In addition, by using a correction amount of a portion that is not an edge pattern that is in contact with the edge pattern, a streak-like artifact (unevenness) that extends in a direction perpendicular to the boundary of the sub-area at the boundary between the edge pattern and the portion that is not the edge pattern Can also be prevented.

Here, FIG. 7A is a graph showing an example of the calculation result of the step amount at each position of the boundary, and FIG. 7B is a graph showing an example of the calculation result of the correction amount at each position of the boundary. FIG. 7C is a graph showing an example of a calculation result obtained by calculating the correction amount based only on the immediately preceding correction amount in the edge pattern.
Specifically, as shown in FIG. 7A, a region where the step amount exceeds the threshold value f _UPPER is detected as an edge pattern region, and there are regions other than the edge pattern region at both ends of the edge pattern region. As shown in FIG. 7B, the level difference is calculated while maintaining the edge pattern region of the image by calculating the correction amount of the edge pattern region from the correction amount of the position in contact with both ends of the edge pattern region. The amount can be corrected.
In addition, by calculating the correction amount of the edge pattern area from two correction amounts at positions adjacent to both ends of the edge pattern area, the correction amount rapidly increases at the boundary between the edge pattern area and the non-edge pattern area. The change can be prevented, and the occurrence of artifacts in the image can be prevented.
That is, when the correction amount of the edge pattern region is calculated from only one end (for example, calculated using the previous correction amount), the correction amount at the boundary with the other end as shown in FIG. The correction amount may change abruptly at the boundary between the two, but it can be seen that the correction amount changes abruptly by adding the correction amounts at the positions adjacent to both ends as shown in FIG. Can be prevented. As a result, since the correction amount can be prevented from changing suddenly between adjacent pixels in the main direction, it is possible to prevent a constant discontinuous change from occurring in the same position in the main direction, and the sub direction. It is possible to more reliably prevent the occurrence of an artifact extending to

  Here, the correction value of the edge pattern region may be calculated by linear interpolation from two correction amounts at positions adjacent to both ends of the edge pattern region as shown by a solid line in FIG. As indicated by a dotted line in B), it may be calculated by nonlinear interpolation such as spline interpolation from two correction amounts at positions adjacent to both ends of the edge pattern region. Here, since the correction amount can be changed more smoothly, the correction amount is preferably calculated by nonlinear interpolation such as spline interpolation.

Further, when the edge pattern area does not have the other end (the boundary between the edge pattern area and the non-edge pattern area) as in the imaging device 10, that is, the edge pattern area is positioned at the edge of the image even in the main direction. In this case, the step amount can be calculated more appropriately by calculating the correction amount based on only one end.
As described above, even when the edge pattern region does not have the other end, an appropriate correction value can be calculated by calculating the correction amount based on the step amount of the edge pattern region adjacent to the edge pattern region. it can.

  Further, as described above, the correction amount is adjusted according to the pixel density. Specifically, multiplication is performed at a pixel density lower than a certain density (in this embodiment, the pixel density at the position where the step amount is calculated). By calculating the correction amount and calculating the correction amount by addition at a pixel density higher than a certain density, it is possible to prevent underflow due to overcorrection. In this way, the amount of correction is changed in accordance with the pixel density. Specifically, when the pixel density is low, the amount of correction is reduced to prevent underflow, thereby preventing image unevenness. it can.

Further, the line position (that is, the pixel position) used for calculating the step amount between the sub-areas may be in the vicinity of the boundary, but the position that is in contact with the adjacent sub-area (that is, is in contact with the boundary). It is preferable to use the pixel at position. In addition, it is preferable to use pixels at positions symmetrical with respect to the boundary for the calculation of the step amount.
Furthermore, the level difference between the sub-areas is not limited to detecting the pixel density of one pixel in each sub-area, and may be calculated by comparing the pixel density statistical values of a plurality of pixels.
Here, as the statistical value, a value calculated from the pixel density of a plurality of pixels using a median filter may be used, or an average value of pixel densities of a plurality of pixels may be used.
Thus, by using a statistical value, high frequency noise can be removed and erroneous detection can be prevented.
Here, the statistic is preferably calculated from the pixel density of a plurality of pixels in the sub-direction (that is, the direction orthogonal to the boundary). By calculating statistics from multiple pixels in the sub-direction, that is, multiple pixels with the same position in the main direction, even if there is noise such as horizontal stripes in the image, more accurate steps that eliminate the effect The amount can be calculated.

In the smoothing processing unit 54 , the smoothing process is performed by the median filter. However, the present invention is not limited to this, and various smoothing processes can be used. For example, smoothing may be performed by moving average processing.
Here, from the viewpoint of step amount for calculating is not affected by noise such as horizontal stripes, as in the image pickup apparatus 10, it is preferable to provide a smoothing processing section 54, the present invention is not limited thereto, The smoothing processing unit may not be provided.

In addition, since the imaging device 10 can reliably correct the step between the sub-areas and appropriately calculate the edge pattern, it is determined whether or not the calculated step amount overlaps the edge pattern of the image. As the threshold value, a value in the vicinity of the maximum value of the step amount not caused by the edge pattern is used. However, the present invention is not limited to this, and may be an arbitrary value set by an operator, for example.
Also, the value near the maximum value of the step amount not caused by the edge pattern is one of the maximum value of the step amount not caused by the edge pattern, a value smaller by a certain amount than this maximum value, or a value larger by a certain amount than the maximum value. It is preferable that the step amount is not caused by the edge pattern, or is a certain amount larger than the maximum value or the maximum value. By setting the threshold value to the maximum value of the step amount not caused by the edge pattern or a value larger than the maximum value by a certain amount, it is possible to reliably prevent the step amount caused by the step between the sub-areas from being determined as the edge pattern region. be able to.

  The radiographic image capturing apparatus and radiographic image capturing method of the present invention have been described in detail above. However, the present invention is not limited to the above-described embodiment, and various improvements and modifications can be made without departing from the gist of the present invention. Of course it is also good.

  For example, in the imaging apparatus 10, one FPD 32 is configured by five subareas, but the number of subareas is not limited to this, and one FPD 32 may be configured by two or more subareas. Moreover, in the said embodiment, although it was set as the structure which piled up the subarea in the up-down direction in FIG. 2, it is not limited to this, You may comprise so that a subarea may be piled up in the left-right direction in FIG.

In the imaging device 10, after calculating the correction amount from the step amount, it is determined whether the boundary and the edge pattern overlap. However, the present invention is not limited to this, and after determining whether the boundary and the edge pattern overlap. The correction amount may be calculated from the step amount. After determining whether the boundary and the edge pattern overlap, by calculating the correction amount from the step amount, the region where the boundary and the edge pattern overlap can be calculated without calculating the correction amount from the step amount. As described above, the correction amount can be calculated from the correction amounts at both ends of the edge pattern.
Thereby, the correction amount can be calculated without waste, the calculation amount can be reduced, and the step correction process can be performed in a shorter time.

It is a block diagram which shows schematic structure of one Embodiment of the radiographic imaging apparatus of this invention. It is a schematic diagram which shows schematic structure of FPD shown in FIG. It is a flowchart which shows operation | movement of the level | step difference correction | amendment means of an imaging device. It is a flowchart which shows operation | movement of an edge area | region determination part and a level | step difference correction | amendment part. It is a front view which shows the coordinate of each position in two adjacent subareas. (A) is a graph showing the relationship between the pixel value before correction and the pixel value after correction when the step amount is positive, and (B) is the pixel value before correction when the step amount is negative. It is a graph which shows the relationship with the pixel value after correction | amendment. (A) is a graph showing an example of the calculation result of the step amount at each position of the boundary, (B) is a graph showing an example of the calculation result of the correction amount at each position of the boundary, (C) The edge pattern is a graph showing an example of a calculation result obtained by calculating a correction amount based only on the immediately preceding correction amount. (A) is a schematic diagram conceptually showing the relationship between an artifact, a boundary, and an edge pattern, and (B) is an enlarged schematic diagram showing an enlarged periphery of the boundary of (A).

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 Radiation imaging device 12 Imaging means 14 Imaging | photography data processing means 15 Level | step difference correction means 16 Image processing means 18 Output means 20 Imaging instruction means 22 Control means 24 Irradiation control part 26 Radiation source 28 Imaging stand 30 Radiation detection part 32 FPD (flat panel) Detector)
42a to 42e Subarea 44a to 44e Read circuit 46a to 46e Read line 48a to 48d Boundary 52 Step amount calculation unit 54 Smoothing processing unit 56 1/2 multiplier 58 Edge region determination unit 60 Step correction unit

Claims (19)

A plurality of subareas each including a plurality of detection elements for detecting radiation, a plurality of reading circuits provided corresponding to each subarea, and reading detection signals generated by the detection elements in the corresponding subarea; A radiation detector having
Photographing data processing means for generating image data from detection signals respectively read by the plurality of reading circuits;
Step correction for correcting a step in pixel density generated at the boundary between the sub-area and the sub-area adjacent to the sub-area of the image data generated by the photographing data processing unit based on image data in the vicinity of the boundary Means,
The step correction means includes a step amount detection unit that detects a step amount of pixel density generated at a boundary between the sub area and a sub area adjacent to the sub area.
An edge region determination unit that determines whether an image at a position corresponding to the boundary is an image of an edge pattern region extending substantially parallel to the boundary from the step amount detected by the step amount detection unit;
A correction amount is calculated based on the step amount detected by the step amount detection unit, and a step correction unit corrects the detection signal based on the calculated correction amount.
The step correction unit calculates a correction amount of the region determined as the edge pattern region based on a correction amount of the region not determined as the edge pattern region in contact with an end portion of the region determined as the edge pattern region. A radiographic image capturing apparatus for correcting image data in an edge pattern region.   When the one end of the area determined to be the edge pattern area coincides with the edge of the image, the step correction unit is based only on the correction amount of the area in contact with the other end of the area determined to be the edge pattern area. The radiographic image capturing apparatus according to claim 1, wherein a correction amount of an area determined as the edge pattern area is calculated.   When the both ends of the area determined as the edge pattern area do not coincide with the edge of the image, the step correction unit determines whether the edge pattern area and the edge pattern area based on the correction amount of the area contacting each of the both ends of the area determined as the edge pattern area. The radiographic image capturing apparatus according to claim 1, wherein a correction amount of the determined area is calculated. The radiographic imaging apparatus according to claim 1, wherein the step correction unit calculates a correction amount by linear interpolation using step amounts at both ends. The radiographic imaging apparatus according to claim 1, wherein the step correction unit calculates a correction amount by nonlinear interpolation using a step amount at both ends. The step amount detection unit smoothes the difference between the pixel density of the image data near the boundary of the sub area and the pixel density of the image data near the boundary of the sub area adjacent to the sub area in a direction parallel to the boundary. The radiographic image capturing apparatus according to claim 1 , wherein the step amount is calculated by performing the conversion process.   The radiographic image capturing apparatus according to claim 6, wherein the smoothing process is at least one of a process using a median filter and a moving average process.   The radiographic image capturing apparatus according to claim 1, wherein the edge area determination unit determines an edge pattern area by determining a threshold value of a step amount.   The radiographic image capturing apparatus according to claim 8, wherein the threshold value is a value near a maximum value of a step amount that is not caused by an edge pattern.   The level difference detection unit calculates a level difference by calculating a statistical amount of pixel density of image data in the vicinity of the boundary of each sub-area and calculating a difference from the statistical value. The radiographic imaging apparatus of description.   The radiographic image capturing apparatus according to claim 10, wherein the step amount detection unit calculates the statistic by a median filter.   The radiographic image capturing apparatus according to claim 10, wherein the step amount detection unit calculates an average value of pixel densities of image data in the vicinity of the boundary of each sub-area as a statistical amount. The radiographic image capturing apparatus according to claim 10 , wherein the statistic is a value calculated from detection signals respectively detected from a plurality of detection elements arranged in a direction orthogonal to the boundary.   The radiographic image capturing apparatus according to claim 1, wherein the step amount detection unit calculates a step amount from a difference between pixels adjacent to another sub area. The radiographic image capturing apparatus according to claim 1, wherein the step correction unit calculates a correction amount in which a correction intensity is attenuated as the distance from the boundary increases. The radiographic image capturing apparatus according to claim 1, wherein the step correction unit calculates a correction amount according to a pixel density of image data before correction. The radiographic imaging apparatus according to claim 1, wherein the step correction unit calculates a correction amount that does not cause an underflow due to overcorrection. The step correction unit calculates the correction amount by addition when the pixel density of the image data before correction is larger than the pixel density of the image data used to calculate the step amount, and the image data used to calculate the step amount The radiographic image capturing apparatus according to claim 1, wherein a correction amount is calculated by multiplication when the pixel density is smaller than the pixel density. A plurality of subareas each including a plurality of detection elements for detecting radiation, a plurality of reading circuits provided corresponding to each subarea, and reading detection signals generated by the detection elements in the corresponding subarea; A radiation image reading step for reading a radiation image with a radiation detector comprising:
An image data generation step for generating image data from detection signals read by each of the plurality of reading circuits;
A step amount of pixel density generated at a boundary between the sub area and a sub area adjacent to the sub area is detected, and an image at a position corresponding to the boundary extends substantially parallel to the boundary from the detected step amount. A step correction step of determining whether the image is an edge pattern region image to be calculated , calculating a correction amount based on the detected step amount, and correcting the detection signal based on the calculated correction amount;
In the step correction step, the correction amount of the area determined as the edge pattern area is calculated based on the correction amount of the area not determined as the edge pattern area in contact with the edge of the area determined as the edge pattern area. A radiographic image capturing method comprising correcting image data of an edge pattern region.