JP2009011566A - Image processing apparatus, method and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To improve the detection accuracy of an adjacent defect line. <P>SOLUTION: This image processing apparatus applies different thresholds to detect defect lines respectively (108 and 110) using offset compensation data which are acquired by reading image data a plurality of times from a radiographic image detection in a radiation non-irradiating state, and defect line detection image data which are acquired by uniformly irradiating radiation of a given level to the whole face of the radiographic image detector, reading the radiographic data from the radiographic image detector a plurality of times, averaging the plurality of pieces of image data (106) and executing an offset compensation and shading compensation, composes defect line maps acquired by the defect line detections respectively (112) and provides a defect line map for correcting a defect line. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to an image processing apparatus, method, and program, and in particular, an image processing apparatus that detects a defect line of an image sensor in which photoelectric conversion units are arranged two-dimensionally, an image processing method applicable to the image processing apparatus, and a computer The present invention relates to an image processing program for causing a computer to function as the image processing apparatus.

  In radiography for the purpose of medical diagnosis, radiation transmitted through the subject is irradiated to a radiation image detector equipped with a photoelectric conversion layer that is sensitive to radiation, and radiation image detection is performed according to the radiation dose applied to the radiation image detector. There is known a system that obtains a digital radiation image by reading out electric charges accumulated in a device as an electric signal for each pixel and converting the read electric signal into digital data. In a radiation image detector in this type of system, for example, a pixel (referred to as a defective pixel) from which an output corresponding to the radiation dose cannot be obtained due to deterioration due to radiation irradiation, poor contact of electrical wiring, or the like may occur. Pixel defects of imaging elements such as radiation image detectors are pixel defects in which a single or a small number of defective pixels are distributed in a dot pattern (called point defects), and pixels in which a plurality of defective pixels are distributed in a straight line. Although it is roughly classified into defects (referred to as line defects), among these, line defects are easily visually recognized even if the difference in brightness and the like from surrounding pixels is small compared to point defects.

  For this reason, in Patent Document 1, a pixel of interest is selected from image information of an input screen, a sampling region is defined so as to include the pixel of interest, and luminance is maximized for each pixel column from pixel data of the sampling region. Or, extracting the smallest feature pixel, and setting a weighted value according to the distance between the predicted position of the line defect obtained from the position information of the feature pixel and the target pixel to the target pixel, A technique is disclosed in which a feature value is set for all pixels by repeatedly performing for each pixel of interest, and a line defect on the screen is determined based on the feature value.

Further, in Patent Document 2, a first defective pixel whose difference from the average value of sensitivity characteristics is larger than a first threshold among defective pixels on a flat panel X-ray detector is first extracted, and subsequently And extracting second defective pixels having a difference from the average value of the sensitivity characteristics larger than the second threshold value (<first threshold value) out of the pixels adjacent to the first defective pixel. A technique for obtaining an X-ray image by performing luminance correction on all defective pixels is disclosed, and the same processing is performed on an offset image that is captured without irradiating X-rays, and the result is synthesized. Is also disclosed.
JP 2006-47077 A JP 2006-234557 A

  In an image sensor such as a radiation image detector, a line adjacent to a line in which a line defect occurs (defect line) may be affected by the defect line and change to a characteristic different from that of a normal line. When the line changes to a characteristic different from that of the normal line due to the influence of the line, the characteristic of the line (adjacent defect line) is different from the characteristic of the defective line and is not constant. The techniques described in Patent Documents 1 and 2 have a problem that the adjacent defect line cannot be detected depending on the characteristics of the adjacent defect line.

  That is, the inventor of the present application has confirmed through experiments that the adjacent defect line may exhibit the characteristics as shown in FIGS. In this experiment, a radiographic image detector was used as an imaging device, and the signal output in the defect line and the surrounding lines at the time of radiation non-irradiation and radiation irradiation was examined. FIGS. 11 (A) and 11 (B) In the example shown, the signal output value of the defective line (line n) is the same as the offset (dark output value) of the normal line (lines n−2, n + 2, etc.) both when radiation is not irradiated and when radiation is irradiated. On the other hand, the signal output values of adjacent defect lines (lines n−1, n + 1) are lower than the offset value of the normal line when radiation is not irradiated, and the output slightly increases when radiation is irradiated. When the variation in offset (dark output) for each pixel is corrected by offset correction and the variation in sensitivity for each pixel is corrected by shading correction with respect to the signal output value pattern shown in FIG. A signal output value pattern as shown in (C) is obtained. In FIG. 11C, the signal output value of the defective line is greatly increased because the signal output value of the defective line does not change regardless of radiation irradiation. This is because the value obtained by dividing the value is set to 0.

  In addition, the signal output value of each line in FIG. 11 represents the average of the signal output values from all the pixels of the line. In addition, FIG. 11 shows an example in which the lines on both sides of the defect line are adjacent defect lines having different characteristics from the normal line. However, when only one of the lines on both sides is the adjacent defect line, It is also possible that all of these lines are normal lines.

  In the pattern shown in FIG. 11C, the difference between the signal output values of the adjacent defect line and the normal line is very small through the offset correction and the shading correction. The techniques described in Patent Documents 1 and 2 correct the value variation caused by factors other than pixel defects with respect to the signal corresponding to FIG. 11C (the signal obtained when light is incident on the image sensor). When the adjacent defect line shows the characteristics as shown in FIGS. 11A and 11B, the adjacent defect line and the normal line in the defect line detection target signal are detected. As shown in FIG. 11C, the difference in signal output value between and the adjacent defect lines cannot be detected as a defect line.

  Patent Document 2 describes that a defect line is also detected for an offset image and the result is synthesized, but the technique described in Patent Document 2 is based on a first threshold value. After detecting one defective pixel (defective line), a second threshold value smaller than the first threshold value is applied to a pixel adjacent to the first defective pixel to generate a second defective pixel ( When the adjacent defect line shows the characteristics as shown in FIGS. 11A and 11B even if the above-described defect line is detected for the offset image, As shown in FIG. 11A, since there is no difference in signal output value between the defective line and the normal line on the offset image, the defective line cannot be detected, and accordingly, the adjacent defective line cannot be detected.

  The present invention has been made in consideration of the above-described facts, and an object thereof is to obtain an image processing apparatus, an image processing method, and an image processing program that can improve the detection accuracy of adjacent defect lines.

  In order to achieve the above object, an image processing device according to the first aspect of the present invention is an image pickup device in which photoelectric conversion units are arranged two-dimensionally, and the photoelectric conversion unit has sensitivity, and is not irradiated with light in a predetermined wavelength range. First detection means for detecting a defective line of the image sensor for the entire image represented by the first image signal using a first threshold value based on the first image signal read from, and light in the predetermined wavelength range Offset correction that corrects variations in dark output for each photoelectric conversion unit and shading that corrects variations in sensitivity for each photoelectric conversion unit for image signals read from the image sensor in a uniformly irradiated state Second detection means for detecting a defective line of the image sensor for the entire image represented by the second image signal using a second threshold value based on the second image signal obtained by performing the correction; and Inspection It is configured to include a synthesizing means for synthesizing the detection result of the defective line by the detection result and the second detection means defective line by means.

  The invention described in claim 1 includes first detection means and second detection means. The first detection means is configured to receive light in a predetermined wavelength range in which photoelectric conversion units are arranged two-dimensionally and the photoelectric conversion units are sensitive. Based on the first image signal read from the image sensor in an unirradiated state, the first threshold is used to detect a defect line of the image sensor for the entire image represented by the first image signal, and the second detection means For image signals read from an image sensor that is uniformly irradiated with light in a predetermined wavelength range, offset correction that corrects variations in dark output for each photoelectric converter and sensitivity for each photoelectric converter Based on the second image signal obtained by performing the shading correction for correcting the variation, a defective line of the image sensor is detected for the entire image represented by the second image signal using the second threshold value.

  Specifically, the detection of the defective line using the first threshold by the first detection unit is specifically performed based on, for example, the first image signal for each line (for example, all photoelectric conversion units on the same line). The average value of the signal values, etc.), and the deviation between the signal value of each line and the first reference value (for example, the average value of the signal values of all the lines obtained from the first image signal, or a predetermined value) This can be done by determining whether or not is greater than or equal to the first threshold. For the detection of the defective line using the second threshold by the second detection means, specifically, for example, as described above, a signal value is obtained for each line based on the second image signal, This can be done by determining whether or not the deviation between the signal value and the second reference value (for example, the average value of the signal values of all lines obtained from the second image signal) is equal to or greater than the second threshold value.

  Thereby, when the defect line and the adjacent defect line exist in the image sensor, the signal values of the defect line and the adjacent defect line are set to the second reference value on the second image signal subjected to the offset correction and the shading correction. If they are different by the second threshold value or more, these lines are detected as defective lines by the second detection means. In addition, when the adjacent defect line existing in the image sensor exhibits the characteristics as shown in FIGS. 11A and 11B, the adjacent defect line is not detected as a defect line by the second detection unit. When the signal value of the adjacent defect line is different from the first reference value by a first threshold or more on one image signal, the adjacent defect line is detected as a defective line by the first detection means. According to the first aspect of the present invention, since the detection result of the defective line by the first detection means and the detection result of the defect line by the second detection means are synthesized by the synthesis means, at least one of the first detection means and the second detection means. A line detected as a defective line by one is treated as a defective line.

  Thus, according to the first aspect of the present invention, the first image signal read from the imaging element that has not been irradiated with the light in the predetermined wavelength region and the imaging in which the light in the predetermined wavelength region is uniformly irradiated. For the second image signal obtained by performing offset correction and shading correction read from the element, the defect line is detected for the entire image without distinguishing between the defect line and the adjacent defect line. When there is a defect line, the probability that the adjacent defect line is detected as a defect line by the first detection means or the second detection means can be improved regardless of the characteristics, and the detection accuracy of the adjacent defect line can be improved. Improvement can be realized.

  In the first aspect of the present invention, as the imaging device, for example, as described in claim 2, a radiation image detector including a photoelectric conversion layer having sensitivity to radiation is preferable. In this case, the imaging device The image captured by the radiation image detector is an image representing the spatial distribution of radiation incident on the radiation image detector. However, the image sensor according to the present invention is not limited to the above-described radiation image detector, and may be another image sensor such as a CCD sensor.

  More specifically, the radiological image detector according to claim 2 as the image pickup device is provided with a plurality of gate lines and a plurality of data lines, respectively, as described in claim 3, for example. Corresponding to the photoelectric conversion unit, switching elements respectively connected to any of the plurality of gate lines and any of the plurality of data lines are two-dimensionally arranged, and a signal input via the gate line Accordingly, a configuration in which the signal of the photoelectric conversion unit corresponding to the turned on switching element is read through the data line by turning on the switching element in accordance with the switching element is suitable, but is not limited to this, for example, constant A plurality of readout electrodes arranged along the direction are provided, and the irradiated radiation is converted into electric charges to be accumulated and held, and the light is irradiated so that it is held in the irradiated area. Charges corresponding to the amount of charge that is may be configured to be output as a current through the corresponding readout electrodes.

  According to the first aspect of the present invention, offset correction (and shading correction) is performed only for the second image signal, and the noise component to be corrected by the offset correction (darkness for each photoelectric conversion unit) is applied to the first image signal. Considering that the noise component corresponding to the output variation) is superimposed, the first threshold value and the second threshold value are set to a value larger than the second threshold value as described in claim 4, for example. It is preferable to set.

  Further, in the first aspect of the invention, the first detection means and the second detection means may be a median for removing a noise component from an image signal used for detection of a defective line, for example, as described in claim 5. It is preferable that the defect line is detected after the processing. Specifically, for example, as described in claim 6, a median process in the first direction is performed on an image signal used for detection of a defective line, and the image signal that has undergone the median process in the first direction is used to perform the first median process. A defect line extending in a second direction intersecting with one direction is detected, a median process in a second direction is performed on the image signal, and the first direction is performed using the image signal that has undergone the median process in the second direction. It is possible to configure so as to detect a defect line extending in the direction. Thus, by making the direction of the median processing different from the direction of the defect line to be detected thereafter, the noise component of the image signal can be removed or reduced by the median processing without adversely affecting the detection of the defective line. This can further improve the detection accuracy of the defective line.

  In the first aspect of the invention, the detection result of the defective line synthesized by the synthesizing means can be used for correction of the defective line on the image signal by, for example, interpolation calculation, but on the image represented by the image signal. When a predetermined number or more of defective lines are continuous, it becomes difficult to correct the defective lines properly by correction using interpolation calculation or the like. In view of this, the synthesis unit, for example, as described in claim 7, combines the defect line detection result by the first detection unit and the defect line detection result by the second detection unit, and then the synthesized defect line. In this detection result, a warning may be output when the detected defective lines continue for a predetermined number or more. As a result, when a predetermined number or more of defective lines continue, it is possible to take another measure other than correction by interpolation calculation or the like, such as determining that the image sensor is defective and replacing it.

  The image processing method according to an eighth aspect of the present invention is the first image signal read out from the image sensor in which the photoelectric conversion units are arranged two-dimensionally and the photoelectric conversion unit has sensitivity and the light in the predetermined wavelength region is not irradiated. Based on the above, the image sensor in a state in which the defect line of the image sensor is detected for the entire image represented by the first image signal using the first threshold and the light in the predetermined wavelength range is uniformly irradiated The second image signal obtained by performing offset correction for correcting dark output variation for each photoelectric conversion unit and shading correction for correcting sensitivity variation for each photoelectric conversion unit on the image signal read from And detecting a defect line of the image sensor for the entire image represented by the second image signal, using the second threshold, and detecting a defect line detection result based on the first image signal and the second image Since synthesizing the detection result of the defective line based on the item, as with the first aspect of the present invention, it is possible to realize improvement of detection accuracy of adjacent defective line.

  The image processing program according to the ninth aspect of the invention is a computer that reads a computer from an image pickup device in which photoelectric conversion units are arranged two-dimensionally, and the photoelectric conversion unit has sensitivity, and is not irradiated with light in a predetermined wavelength range. First detection means for detecting a defective line of the image sensor for the entire image represented by the first image signal using a first threshold value based on one image signal, light in the predetermined wavelength range is uniformly irradiated Obtained by performing offset correction for correcting variations in dark output for each photoelectric conversion unit and shading correction for correcting variations in sensitivity for each photoelectric conversion unit for the image signal read from the image sensor in the state of Second detection means for detecting a defective line of the image sensor for the entire image represented by the second image signal using the second threshold value based on the second image signal, and the first detection To function detection result of the defective line by the detection result and the second detection means defective line by means as combining means for combining.

  The image processing program according to the invention described in claim 9 is a program for causing a computer to function as the first detecting means, the second detecting means, and the synthesizing means. Therefore, the computer according to the invention according to claim 9. By executing the image processing program, the computer functions as the image processing apparatus according to claim 1, and as in the invention according to claim 1, improvement in detection accuracy of adjacent defect lines can be realized. it can.

  As described above, the present invention detects a defect line of an image sensor for the entire image using the first threshold value based on the first image signal read from the image sensor not irradiated with light in a predetermined wavelength range. And using the second threshold value based on the second image signal obtained by performing offset correction and shading correction on the image signal read from the imaging element in a state where light in a predetermined wavelength region is uniformly irradiated, Since the defect line of the image sensor is detected for the entire image and the detection result of the defect line based on the first image signal and the detection result of the defect line based on the second image signal are combined, detection of the adjacent defect line It has an excellent effect of improving accuracy.

  Hereinafter, an example of an embodiment of the present invention will be described in detail with reference to the drawings. FIG. 1 shows a radiographic image capturing apparatus 10 according to the present embodiment. The radiographic imaging device 10 includes a radiation generation unit 12 that generates radiation such as X-rays (X-rays), and a radiation detection unit 14 that is provided at a distance from the radiation generation unit 12. Between the radiation generation unit 12 and the radiation detection unit 14 is an imaging position where the subject 16 is located at the time of imaging, and image information is transmitted by passing through the subject 16 emitted from the radiation generation unit 12 and positioned at the imaging position. The carried radiation is applied to the radiation detection unit 14.

  The radiation detection unit 14 includes a radiation image detector. The radiation image detector includes an electrostatic recording unit including a photoconductive layer that exhibits conductivity when irradiated with radiation, and receives image information on the electrostatic recording unit upon irradiation with radiation carrying image information. It records and outputs the image signal showing the recorded image information. Radiation image detectors include optical reading radiation image detectors that read image information recorded on the electrostatic recording unit using a semiconductor material that generates charges when irradiated with light, and charges generated when irradiated with radiation. There is a radiation image detector of a type (hereinafter referred to as a TFT method) that reads the stored charge by turning on and off a switching element such as a thin film transistor (TFT) one unit region at a time. Hereinafter, the configuration of the TFT type radiation image detector will be described as an example.

  As shown in FIGS. 2 and 3, the radiation image detector 20 according to the present embodiment exhibits electromagnetic wave conductivity, and charges according to the amount of incident radiation when radiation is incident with a bias voltage applied. The photoconductive layer 28 is provided as a charge conversion layer for generating the light. As the photoconductive layer 28, an amorphous material having a high dark resistance, good electromagnetic wave conductivity with respect to radiation irradiation, and capable of forming a large area at a low temperature by a vacuum deposition method is preferable. An amorphous Se (a-Se) film is used. A material obtained by doping As, Sb, and Ge into amorphous Se is also preferable because of its excellent thermal stability. A bias electrode 22 as an upper electrode for applying a bias voltage to the photoconductive layer 28 is provided on the upper side of the photoconductive layer 28. The bias electrode 22 is formed of a material such as gold (Au), for example, and constitutes an electrode layer 60.

  A plurality of charge collection electrodes 36A as lower electrodes are provided below the photoconductive layer 28. The plurality of charge collection electrodes 36A are arranged over the entire surface of the radiation image detector 20 (two-dimensionally), and the individual charge collection electrodes 36A are also arranged over the entire surface of the radiation image detector 20, respectively. The charge storage capacitor (capacitor) 36C and the switching element 36B are connected to each other (see FIG. 3). The charge collection electrode 36A, the switching element 36B, and the charge storage capacitor 36C form a charge detection layer 36, and the charge detection layer 36 forms an active matrix substrate 62 together with the glass substrate 38.

An intermediate layer is provided between the photoconductive layer 28 and the bias electrode 22. The intermediate layer is a layer existing between the upper electrode and the charge conversion layer, and may also serve as a charge injection blocking layer (including charge accumulation and diode formation). As a charge injection blocking layer, in addition to a resistance layer and an insulating layer, a hole injection blocking layer that blocks hole injection while being a conductor for electrons, and a conductor for holes There is an electron injection blocking layer that blocks electron injection. The hole injection blocking layer can be formed using, for example, CeO ₂ , ZnS, Sb ₂ S ₃ . Of these, ZnS is desirable because it can be formed at a low temperature. The electron injection blocking layer can be formed using, for example, Se, CdTe doped with Sb ₂ S ₃ , CdS, or Te, an organic compound, or the like. Sb ₂ S ₃ can function as a hole injection blocking layer or an electron injection blocking layer depending on the position where it is provided. In the present embodiment example, since the polarity of the bias electrode is positive, the hole injection blocking layer 24 is provided as an intermediate layer. An electron injection blocking layer 32 is provided between the photoconductive layer 28 and the charge collection electrode 36A.

Further, crystallization preventing layers 26 and 30 are provided between the hole injection blocking layer 24 and the photoconductive layer 28 and between the electron injection blocking layer 32 and the photoconductive layer 28, respectively. Anti-crystallization layer 26, 30 may for example GeSe, be formed by or _{GeSe 2, Sb 2 Se 3,} a-As 2 Se 3, Se-As, Se-Ge, a Se-Sb-based compounds .

  One pixel of the radiation image detector 20 shown in FIG. 4 and FIG. 5 has a size of about 0.1 mm × 0.1 mm to 0.3 mm × 0.3 mm. About 500 × 500 to 3000 × 3000 pixels are arranged in a matrix. As shown in FIG. 4, an active matrix substrate 62 is formed on a glass substrate 38 with a gate electrode 42, a charge storage capacitor electrode (hereinafter referred to as Cs electrode) 54, a gate insulating film 46, a drain electrode 44, a channel layer 48, A contact electrode 50, a source electrode 40, an insulating protective film 52, an interlayer insulating film 56, and a charge collecting electrode 36A are formed. The switching element 36B includes a gate electrode 42, a gate insulating film 46, a source electrode 40, a drain electrode 44, a channel layer 48, a contact electrode 50, etc., and the charge storage capacitor 36C has a Cs electrode 54, a gate insulating film 46, a drain. It is constituted by the electrode 44 and the like.

  As the glass substrate 38, for example, an alkali-free glass substrate (for example, # 1737 manufactured by Corning) can be used. Further, the gate electrode 42 and the source electrode 40 are electrode wirings arranged in a lattice pattern as shown in FIG. 5, the switching element 36B is formed at each intersection, and the source of the switching element 36B is connected to the source electrode 40. The drain of the switching element 36B is connected to the drain electrode 44 (see also FIG. 6). The source electrode 40 includes a linear portion as a signal line and an extended portion for constituting the switching element 36B, and the drain electrode 44 is provided so as to connect the switching element 36B and the charge storage capacitor 36C.

  The gate insulating film 46 is made of a material such as SiNX or SiOX, and is provided so as to cover the gate electrode 42 and the Cs electrode 54. The portion of the gate insulating film 46 located on the gate electrode 42 acts as a gate insulating film in the switching element 36B, and the portion located on the Cs electrode 54 acts as a dielectric layer in the charge storage capacitor 36C. That is, the overlapping region of the Cs electrode 54 and the drain electrode 44 formed in the same layer as the gate electrode 42 functions as the charge storage capacitor 36C. The gate insulating film 46 is not limited to SiNX or SiOX, and an anodic oxide film obtained by anodizing the gate electrode 42 and the Cs electrode 54 can be used in combination.

  The channel layer (i layer) 48 is a channel portion of the switching element 36 </ b> B, and is a current path connecting the source electrode 40 and the drain electrode 44. The contact electrode (n + layer) 416 makes contact between the source electrode 40 and the drain electrode 44. The insulating protective film 52 is formed over substantially the entire surface (substantially the entire region) on the source electrode 40 and the drain electrode 44, that is, on the glass substrate 38. As a result, the drain electrode 44 and the source electrode 40 are protected, and electrical isolation is achieved. Further, the insulating protective film 52 has a contact hole 58 formed in a part of the drain electrode 44 facing the Cs electrode 54. The charge collection electrode 36A is composed of an amorphous transparent conductive oxide film. The charge collection electrode 36 </ b> A is formed so as to fill the contact hole 58, and is stacked on the source electrode 40 and the drain electrode 44. The charge collection electrode 36A and the photoconductive layer 28 are electrically connected, and the charge generated in the photoconductive layer 28 is collected by the charge collection electrode 36A. The interlayer insulating film 56 is made of a photosensitive acrylic resin, and serves to electrically isolate the switching element 36B. A contact hole 58 passes through the interlayer insulating film 56, and the charge collection electrode 36 </ b> A is connected to the drain electrode 44. The contact hole 58 is shaped in a reverse taper shape as shown in FIG.

  A high voltage power supply (not shown) is connected between the bias electrode 22 and the Cs electrode 54. A voltage is applied between the bias electrode 22 and the Cs electrode 54 by the high voltage power source. As a result, an electric field can be generated between the bias electrode 22 and the charge collection electrode 36A via the charge storage capacitor 36C. At this time, since the photoconductive layer 28 and the charge storage capacitor 36C are electrically connected in series, if a bias voltage is applied to the bias electrode 22, charge is generated in the photoconductive layer 28. (Electron-hole pairs) are generated. Electrons generated in the photoconductive layer 28 move to the positive electrode side, and holes move to the negative electrode side. As a result, charges are stored in the charge storage capacitor 36C.

  In the TFT radiation image detector 20, a voltage is applied between the bias electrode 22 and the Cs electrode 54, that is, a voltage is applied to the photoconductive layer 28 via the bias electrode 22 and the Cs electrode 54. In this state, when the photoconductive layer 28 is irradiated with radiation, charges (electron-hole pairs) are generated in the photoconductive layer 28. Since the photoconductive layer 28 and the charge storage capacitor 36C are electrically connected in series, electrons generated in the photoconductive layer 28 move to the + electrode side, and holes move to the − electrode side. As a result, charges are stored in the charge storage capacitor 36C. The charge stored in the charge storage capacitor 36 </ b> C can be extracted to the outside through the source electrode 40 by turning on the switching element 36 </ b> B by an input signal to the gate electrode 42. Since the electrode wiring composed of the gate electrode 42 and the source electrode 40, the switching element 36B, and the charge storage capacitor 36C are all provided in a matrix, the gate electrodes 42 for inputting signals are sequentially switched, and the individual source electrodes are switched. By detecting the signal from 40 for each individual source electrode 40, image information representing the spatial distribution of the radiation incident on the radiation image detector 20 can be obtained.

  The radiation image detector 20 described above corresponds to the image sensor according to the present invention (more specifically, the image sensor according to claims 2 and 3).

  The radiation detection unit 14 includes a gate line driver 70 and a signal processing unit 72 shown in FIG. 6 in addition to the radiation image detector 20 described above, and each gate electrode 42 of the radiation image detector 20 is a gate line driver. 70, and each source electrode 40 of the radiation image detector 20 is connected to a signal processing unit 72. The gate line driver 70 is connected to a control device 74 (image read control unit 80: described later), and supplies a drive signal (a signal for turning on the switching element 36B) when reading an image from the radiation image detector 20. By sequentially switching the electrodes 42 in accordance with an instruction from the image readout control unit 80, pixels for one line connected to the same gate electrode (each pixel is composed of a switching element 36B, a charge storage capacitor 36C, and the like). ) As a unit, the charge stored in the charge storage capacitor 36C is sequentially output. The charges output from the individual pixels are input to the signal processing unit 72 as image signals through the source electrode 40. The signal processing unit 72 includes an amplifier and an A / D converter (not shown), amplifies the image signal input via the source electrode 40, converts the signal into digital image data, and converts the image signal into digital image data (an image processing unit of the control device 74). 82: Output to below. Thereby, all the image information for one image recorded in the radiation image detector 20 is read out as image data.

  As shown in FIG. 1, the radiation generation unit 12 and the radiation detection unit 14 are each connected to a control device 74. The control device 74 is connected to a computer having a non-volatile storage unit 74A (not shown except for the storage unit 74A) including a memory composed of a CPU, a RAM, etc., and an HDD (Hard Disk Drive). A predetermined program stored in the nonvolatile storage unit is executed by the CPU of the computer, and the computer and the peripheral circuit cooperate to generate radiation in the radiation generation unit 12. Functions as a radiation generation control unit 78 that controls the image data, and also functions as an image readout control unit 80 that controls readout of image information from the radiation image detector 20. Note that a display 76 for displaying a radiation image is also connected to the control device 74.

  Further, an image processing program is also stored in the non-volatile storage unit 74A of the control device 74, and the control device 74 can be used as the image processing unit 82 shown in FIG. 3 by executing the image processing program by the CPU. Function. The image processing program described above corresponds to the image processing program according to the present invention, and the image processing unit 82 realized by executing the image processing program corresponds to the image processing apparatus according to the present invention. Yes.

  Next, as an operation of the present embodiment, processing realized by the image processing unit 82 will be described. When the image data read from the radiation image detector 20 by the image reading control unit 80 is input, the image processing unit 82 sequentially performs offset correction, shading correction, and defective pixel correction on the input image data. By outputting the image data that has undergone each correction to the display 76, the radiation image represented by the image data is displayed on the display 76. Hereinafter, each correction performed by the image processing unit 82 will be described in order.

  The image data read from the radiation image detector 20 is a state in which the radiation image detector 20 is not irradiated with radiation (that is, no charge is stored in the charge storage capacitor 36C of each pixel and no image information is recorded). Even so, there is variation in the value of each pixel (the offset added to the value of each pixel varies). The offset correction is to correct this variation. Data for offset correction (see also FIG. 7) indicating the magnitude of the offset (dark output) for each pixel is generated in advance and stored in the storage unit 74A. When image data including image information is read out and input from the radiation image detector 20, the offset value for each pixel represented by the offset correction data is subtracted from the value of each pixel of the image data. Is done. Thereby, the image quality degradation of the radiographic image resulting from the dispersion | variation in the offset (dark output) for every pixel of the radiographic image detector 20 can be corrected.

  The above-described offset correction data is obtained by reading image data from the radiation image detector 20 in a state in which no radiation is irradiated a plurality of times (in FIG. 7, a plurality of image data obtained by the plurality of times of reading is shown in FIG. The average value of a plurality of image data obtained by reading a plurality of times is calculated for each pixel (shown by a plurality of blocks indicated as “0mR image data (X-ray non-irradiation)” in the upper left of FIG. 7 is also stored in the storage unit 74A as offset correction data. The offset correction data is generated periodically (for example, at the start of daily work).

  Further, in each pixel of the radiation image detector 20, for example, the relationship between the radiation dose and the amount of latent image polar charge accumulated in the power storage unit 30 due to deterioration caused by repeated irradiation of radiation ( The inclination of the photoelectric conversion characteristics may vary from pixel to pixel. The shading correction is to correct this variation in photoelectric conversion characteristics, and by previously generating shading correction data (see also FIG. 7) that defines a gain for correcting the variation in photoelectric conversion characteristics for each pixel. Stored in the storage unit 74A, the image data including the image information read from the radiation image detector 20 is subjected to the offset correction described above, and then the value of each pixel of the image data subjected to the offset correction is obtained. , By multiplying the gain for each pixel represented by the shading correction data. Thereby, the image quality degradation of the radiographic image resulting from the dispersion | variation in the photoelectric conversion characteristic for every pixel of the radiographic image detector 20 is correctable.

  The above-mentioned shading correction data is generated in a state where the subject 16 is not present at the photographing position, and the radiation generator 12 generates radiation of a certain level (for example, 10 mR), and the radiation image detector 20 emits radiation of a certain level over the entire surface. After uniformly irradiating, the image data is read from the radiation image detector 20 a plurality of times (in FIG. 7, a plurality of image data obtained by the plurality of times of reading are displayed in the upper center of FIG. 10mR image data (indicated by a plurality of blocks expressed as “X-ray uniform irradiation”), and an average value of a plurality of image data obtained by a plurality of readouts is calculated for each pixel (step of FIG. 7). 102), the above-described offset correction is performed on the image data obtained by calculation ("10mR average image data" shown in FIG. 7), and each image in the image data subjected to the offset correction is subjected to the offset correction. Based on the variation of the value for each pixel (the variation of the value at this time is caused by the variation of the photoelectric conversion characteristic for each pixel), a gain for correcting the variation of the photoelectric conversion characteristic for each pixel is calculated (FIG. 7). (See also step 104). The generation of shading correction data is performed periodically (for example, at the start of daily work) in the same manner as the generation of offset correction data.

  Further, in each pixel of the radiation image detector 20, for example, deterioration due to repeated irradiation of radiation, the charge storage capacitor 36C of the radiation image detector 20, the switching element 36B, the source electrode 40, the signal processing unit The amount of charge accumulated in the charge storage capacitor 36C due to poor contact of the electric circuit from 72 to the image processing unit 82 or from the gate line driver 70 to the switching element 36B through the gate electrode 42. A pixel that does not output a corresponding electrical signal, that is, a defective pixel may occur. In the defective pixel correction, the output (value) of the defective pixel is corrected on the image data.

  In the defective pixel correction, a defective pixel of the radiation image detector 20 is detected in advance, and defective pixel information (defective pixel map) indicating the position of each detected defective pixel is stored in the storage unit 74A, and the radiation image detector. After performing the above-described offset correction and shading correction on the image data including the image information read from 20 in order, recognition is performed based on the defective pixel information among the pixels of the image data subjected to the offset correction and the shading correction. Each defective pixel value is obtained by interpolating each of a plurality of non-defective pixel values existing around each defective pixel, and replacing each defective pixel value with the value obtained by interpolation. . Thereby, the image quality degradation of the radiographic image resulting from the defective pixel of the radiographic image detector 20 can be corrected.

  As described above, the image data including the image information read out from the radiation image detector 20 is sequentially subjected to offset correction, shading correction, and defective pixel correction, and then output to the display 76, whereby the radiation image detector It is possible to display on the display 76 a high-quality radiological image in which variation in offset (dark output) for each of the 20 pixels, variation in photoelectric conversion characteristics for each pixel, and image quality deterioration due to defective pixels are corrected. it can. In addition to the above corrections, image data that has been subjected to other processing such as median subtraction, which will be described later, may be output and displayed on the display 76. Further, instead of being displayed on the display 76, processing such as recording image data on a flash memory or other information recording medium or recording an image on a sheet-like recording material by a recording device such as a printer is performed. It may be.

  By the way, the pixel defect in the radiation image detector 20 includes a pixel defect (point defect) in which a single or a small number of defective pixels are distributed in a dot shape, and a pixel defect (in which a plurality of defective pixels are distributed in a straight line). Line defects are roughly classified into line defects. Of these, line defects are easier to be recognized even if the difference in brightness from surrounding pixels is smaller than that of point defects, and lines that have line defects (defect lines). The line adjacent to may be changed to a characteristic different from that of the normal line due to the influence of the defective line. For this reason, in the present embodiment, the defect line extraction process described below is performed, and the defect line information extracted by the process is stored in the storage unit 74A as defect pixel information (defective pixel map). The defect line extraction processing is performed periodically (for example, at the start of daily work) in the same manner as the generation of the offset correction data and shading correction data described above. Further, in the present embodiment, apart from the defect line extraction process described below, a point defect extraction process is also performed in which a point defect is detected and the detection result is stored in the storage unit 74A as defective pixel information (defective pixel map). Description of this point defect extraction process is omitted.

  In the defect line extraction processing according to the present embodiment, first, processing for acquiring image data used for detection of a defective line is performed. In this processing, in a state where the subject 16 does not exist at the imaging position, the radiation generation unit 12 emits radiation of a certain level (however, the radiation level is ½ compared to when shading correction data is acquired). After the radiation image detector 20 is uniformly irradiated with a certain level of radiation, image data is read out from the radiation image detector 20 a plurality of times (in FIG. 7, by the plurality of times of reading out) The obtained plurality of image data is indicated by a plurality of blocks indicated as “5mR image data (X-ray uniform irradiation)” in the upper right of FIG. 7), and a plurality of image data obtained by a plurality of readings. Is calculated for each pixel (see also step 106 in FIG. 7), and the above-described offset correction and shading are performed on the image data ("5mR average image data" shown in FIG. 7) obtained by the calculation. Ingu correction is performed sequentially. Thereby, defect line detection image data (see also FIG. 7) is obtained.

  In the defect line extraction process according to the present embodiment, the defect line detection image data (corresponding to the second image signal according to the present invention) obtained by the above process and the offset correction data for performing the offset correction (according to the present invention). A defect line detection process for detecting a defect line is performed on the target image (corresponding to the first image signal) (see also steps 108 and 110 in FIG. 7). In each defective line detection process, the process shown in FIG. 8 is performed. In FIG. 8, the defect line detection image data and the offset correction data are expressed as “defect line detection target image data”. The defect line detection process in step 108 of FIG. 7 is performed by the first detection means according to the present invention (more specifically, the first detection means according to claims 5 and 6). Corresponds to the second detection means according to the present invention (more specifically, the second detection means according to claims 5 and 6).

  In the defective line detection process shown in FIG. 8, a two-dimensional median process for removing noise is performed on the image data to be detected as a defective line (step 120 in FIG. 8). In this two-dimensional median processing, (2 × S_d + 1) pixels × (2 × S_d + 1) pixels when the allowable maximum size (width and length) of point defects in the radiation image detector 20 is S_d in terms of the number of pixels. A plurality of pixels ((2 × S_d + 1) × within the range of the median filter in which the value of each pixel of the image data of the defect line detection target is arranged so that the position of each pixel is the center. This is done by replacing with the median of (2 × S_d + 1) pixels). As a result, noise such as point defects is removed from the image data to be detected by the defective line.

Further, the value of a plurality of pixels corresponding to the central portion of the image is extracted from the image data that has undergone the two-dimensional median processing, and the average value of the extracted values of the plurality of pixels is calculated as the central portion average value Q_ave (step in FIG. 8). 122). In the present embodiment, a threshold deviation α for defect line detection with respect to the central average value Q_ave is preset, and the upper limit is obtained by reading this deviation α and adding the deviation α to the calculated central average value Q_ave. A threshold value (= Q_ave + α) is calculated, and a lower limit threshold value (= Q_ave−α) is calculated by subtracting the deviation α from the central average value Q_ave (step 124 in FIG. 8). Specifically, the deviation α is more specifically the deviation α ₁ (corresponding to the first threshold value according to the present invention) when the defect line detection target is offset correction data, and the defect line detection target is the defect line detection image data. Deviation α ₂ (corresponding to the second threshold value according to the present invention) is set separately, and deviation α ₁ > deviation α ₂ (the magnitude relationship between these deviations α ₁ and α ₂ is claimed in claim 4). Corresponding to the described invention). In the defect line detection processing for the offset correction data and defective line detection (step 108 in FIG. 7), the deviation alpha ₁ is computing the upper and lower thresholds is performed in read, the defect image data for detecting defective line in the defect line detection processing for the line detection (step 110 in FIG. 7), the deviation alpha ₂ is read the calculation of the upper and lower thresholds is performed.

  On the other hand, in the defect line detection process shown in FIG. 8, the horizontal direction (for example, the direction in which the gate electrode 42 extends, hereinafter, for the sake of convenience, the direction in which the gate electrode 42 extends is the horizontal direction and the source electrode 42 A one-dimensional median process is also performed along the direction in which the direction extends (referred to as the vertical direction) (step 126 in FIG. 8). This median processing in the horizontal direction uses a median filter with a size of (2 x S_d + 1) pixels x 1 pixel, directs the value of each pixel of the image data of the defect line detection target in the horizontal direction, and centers each pixel position This is performed by replacing with the median value of a plurality of pixels ((2 × S_d + 1) pixels) falling within the range of the median filter arranged so as to be. As a result, image data (horizontal median processed image data shown in FIG. 8) from which noise such as line defects along the vertical direction has been removed from the image data of the defect line detection target is obtained.

  Next, a horizontal defect line detection process is performed to detect a horizontal defect line that can be determined to contain a line defect among the lines extending in the horizontal direction on the image data that has undergone the horizontal median process (step 128 in FIG. 8). ). For example, for each line extending in the horizontal direction, the average value of all pixels on the line is calculated, and it is determined whether the calculated average value is equal to or lower than the previously calculated upper limit threshold and lower limit threshold. Can be done. Then, a lateral defect line map (see FIG. 8) is generated in which the calculated average value is larger than the upper threshold value or less than the lower threshold value and the line is clearly indicated as a lateral defect line. As the lateral defect line map, for example, in the lateral median processing image data, all the pixels of the line determined to be the lateral defect line are each added with predetermined information that clearly indicates the pixel on the lateral defect line. Data can be applied.

  Further, in the defect line detection process shown in FIG. 8, one-dimensional median processing along the vertical direction is also performed on the image data to be detected as a defect line (step 130 in FIG. 8). For this median processing in the vertical direction, a median filter with a size of (2 x S_d + 1) pixels x 1 pixel is used, the value of each pixel of the image data for the defective line detection is directed in the vertical direction, and the position of each pixel is the center. This is performed by replacing with the median value of a plurality of pixels ((2 × S_d + 1) pixels) falling within the range of the median filter arranged so as to be. As a result, image data (vertical median processed image data shown in FIG. 8) from which noise such as line defects along the horizontal direction has been removed from the image data of the defect line detection target is obtained.

  Next, a vertical defect line detection process is performed to detect a vertical defect line that can be determined to contain a line defect among the lines extending in the vertical direction on the image data that has undergone the vertical median process (step 132 in FIG. 8). ). For example, for each line extending in the vertical direction, the average value of all the pixel values on the line is calculated, and it is determined whether the calculated average value is equal to or lower than the previously calculated upper limit threshold and lower limit threshold. Can be done. Then, a vertical defect line map (see FIG. 8) is generated in which the calculated average value is greater than the upper limit threshold value or a line determined to be less than the lower limit threshold value is specified as a vertical defect line. As for the vertical defect line map, for example, in the vertical median processed image data, all the pixels of the line determined to be the vertical defect line are each added with predetermined information that clearly indicates the pixel on the vertical defect line. Data can be applied.

  When the vertical defect line map and the horizontal defect line map are generated as described above, the generated vertical defect line map and the horizontal defect line map are combined to perform a map combining process for generating a defect line map (step 134 in FIG. 8). ). This map composition processing is, for example, a pixel on a defective line with respect to a pixel that is specified to be on a vertical defect line or a horizontal defect line in at least one of a vertical defect line map and a horizontal defect line map. This can be done by generating image data (defect line map) to which predetermined information clearly indicating this is added.

  When the defect line detection process described above is performed on the offset correction data and the defect line detection image data to generate defect line maps, a process for synthesizing these defect line maps is performed (step 112 in FIG. 7). Then, the new defect line map obtained by this combining process is further combined with the separately detected point defect detection result, and then stored as defective pixel information (defective pixel map) in the storage unit 74A. Used for the defective pixel correction described above. Note that the composition processing in step 112 in FIG. 7 corresponds to the composition means according to the present invention.

  It should be noted that on the defect line map obtained by the above synthesis process, a plurality of continuous lines may be determined as defect lines, but if the defect lines are continuous for a certain number or more, the radiation image detector 20 Even if defective pixel correction is performed on image data including read image information, a portion corresponding to a defect line that is continuous for a certain number or more in a radiation image that has undergone defective pixel correction is unnatural. It may be finished. For this reason, in the above-described defect line map synthesis process (step 112 in FIG. 7), it is determined whether or not there are n or more defective lines (for example, three lines) on the generated defect line map. If so, it is preferable to issue a warning to the operator. Thus, when the warning is issued, the operator can determine that the radiation image detector 20 is out of order and take measures such as replacing the radiation image detector 20. The above item corresponds to the invention described in claim 7.

As described above, in the defective line extraction process according to the present embodiment, with respect to the offset correction data, it detects the defective line with upper and lower thresholds calculated from preset deviation alpha _1, the defect line detection to use the image data to detect a defect line with upper and lower thresholds calculated from preset deviation alpha _2, since the generated defect line map by combining the detection results of the respective defective line The defect line exists in the radiation image detector 20 and the value of the line adjacent to the defect line is the normal line value (center average value Q_ave) and deviation α ₂ on the defect line detection image data. If the adjacent defect line is different from the above, the adjacent defect line is detected by the defect line detection process using the defect line detection image data. When the defect line exists in the radiation image detector 20 and the line adjacent to the defect line is an adjacent defect line having characteristics as shown in FIGS. In this case, the adjacent defect line is detected as a defect line by defect line processing using offset correction data. Therefore, not only the defect line of the radiation image detector 20 but also the adjacent defect line can be detected with high accuracy.

  In the defect line detection process shown in FIG. 8, a vertical defect line or a horizontal defect line is detected for image data (horizontal median processed image data or vertical median processed image data) that has undergone a one-dimensional median process. However, the present invention is not limited to this, and as shown in FIG. 9 as an example, image data (horizontal median processed image data or vertical median processed image data) that has undergone one-dimensional median processing is used as the original image data. You may make it detect a vertical defect line or a horizontal defect line for the image data which performed the median subtraction which subtracts for every pixel from (defect line detection target image data).

  In addition, as shown in FIG. 10, as the defect line detection processing, a defect line detection threshold value is detected for the defect line detection target image data (this threshold value is also detected when the defect line detection target image data is offset correction data. Binarization is performed to extract defective pixels (step 140 in FIG. 10), and all the vertical and horizontal lines in the binarized image data are included in all pixels of one line. A line having a defective pixel ratio of a certain value or more is determined as a defective line, while a line having a defective pixel ratio of all the pixels in one line is determined as a non-defective line (step 142 in FIG. 10). The process for generating the defect line map may be performed according to the determination result.

  In the above description, the TFT type radiation image detector 20 is described as an example of the image pickup device according to the present invention. However, the present invention is not limited to this, and the image pickup device according to the present invention is an optical reading type. It may be a radiation image detector having another configuration such as a radiation image detector. In the present invention, an image sensor having a characteristic that a line adjacent to a defective line changes to a characteristic different from that of a normal line due to the influence of the defective line is suitable. It is also applicable to.

  In the above description, the image processing program according to the present invention is stored (installed) in the storage unit 74A in advance. However, the image processing program according to the present invention is a recording medium such as a CD-ROM or a DVD-ROM. It is also possible to provide it in the form recorded in

It is a block diagram which shows schematic structure of the radiographic imaging apparatus which concerns on this embodiment. It is the schematic which shows the whole structure of the radiographic image detector of a TFT system. It is the schematic which shows the principal part structure of the radiographic image detector of FIG. It is sectional drawing which shows the structure of 1 pixel unit of a radiographic image detector. FIG. 5 is a plan view of FIG. 4. It is the schematic which shows the peripheral circuit connected to a radiographic image detector. It is the schematic which shows the content of the defect line extraction process. It is the schematic which shows an example of the content of the defect line detection process. It is the schematic which shows the other example of the content of a defect line detection process. It is the schematic which shows the other example of the content of a defect line detection process. It is the schematic which shows an example of the characteristic of a defect line and an adjacent defect line.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 Radiation imaging device 12 Radiation generation part 14 Radiation detection part 20 Radiation image detector 74 Control apparatus 82 Image processing part

Claims (9)

Based on a first image signal read from an image sensor in a state in which photoelectric conversion units are arranged two-dimensionally and the photoelectric conversion unit is sensitive to light in a predetermined wavelength range, the first image is read using a first threshold value. First detection means for detecting a defective line of the image sensor for the entire image represented by the signal;
For image signals read out from the image sensor in a state where the light of the predetermined wavelength range is uniformly irradiated, offset correction for correcting a variation in dark output for each photoelectric conversion unit and for each photoelectric conversion unit A second line that detects a defective line of the image sensor for the entire image represented by the second image signal using a second threshold value based on a second image signal obtained by performing shading correction for correcting sensitivity variations. Detection means;
Combining means for combining the detection result of the defective line by the first detection means and the detection result of the defect line by the second detection means;
An image processing apparatus.   The imaging device is a radiation image detector provided with a photoelectric conversion layer having sensitivity to radiation as the photoelectric conversion units arranged in two dimensions, and an image captured by the radiation image detector as the imaging device is The image processing apparatus according to claim 1, wherein the image processing apparatus is an image representing a spatial distribution of radiation incident on the radiation image detector.   The radiation image detector as the imaging element is provided with a plurality of gate lines and a plurality of data lines, respectively, and corresponding to each of the photoelectric conversion units, any one of the plurality of gate lines and The switching elements respectively connected to any of the plurality of data lines are two-dimensionally arranged, and the switching elements are turned on according to the signal input through the gate line, so that the switching elements turned on The image processing apparatus according to claim 2, wherein a signal of a corresponding photoelectric conversion unit is read out through the data line.   The image processing apparatus according to claim 1, wherein the first threshold is set to a value larger than the second threshold.   The first detection means and the second detection means detect a defective line after performing a median process for removing a noise component on an image signal used for detecting a defective line. The image processing apparatus according to 1.   The first detection means and the second detection means perform a median process in a first direction on an image signal used for detection of a defective line, and use the image signal that has undergone the median process in the first direction to perform the first signal. A defect line extending in a second direction that intersects the direction is detected, the median process in the second direction is performed on the image signal, and the first signal is used by using the image signal that has undergone the median process in the second direction. 6. The image processing apparatus according to claim 5, wherein a defect line extending in the direction is detected.   The synthesizing unit synthesizes the detection result of the defect line by the first detection unit and the detection result of the defect line by the second detection unit. The image processing apparatus according to claim 1, wherein a warning is output when a predetermined number or more is continued. Based on a first image signal read from an image sensor in a state in which photoelectric conversion units are arranged two-dimensionally and the photoelectric conversion unit is sensitive to light in a predetermined wavelength range, the first image is read using a first threshold value. Detecting the defective line of the image sensor for the entire image represented by the signal,
For image signals read out from the image sensor in a state where the light of the predetermined wavelength range is uniformly irradiated, offset correction for correcting a variation in dark output for each photoelectric conversion unit and for each photoelectric conversion unit Based on the second image signal obtained by performing the shading correction for correcting the sensitivity variation, the second threshold is used to detect the defective line of the image sensor for the entire image represented by the second image signal,
An image processing method for combining a defect line detection result based on the first image signal and a defect line detection result based on the second image signal. Computer
Based on a first image signal read from an image sensor in a state in which photoelectric conversion units are arranged two-dimensionally and the photoelectric conversion unit is sensitive to light in a predetermined wavelength range, the first image is read using a first threshold value. First detection means for detecting a defective line of the image sensor for the entire image represented by the signal;
For image signals read out from the image sensor in a state where the light of the predetermined wavelength range is uniformly irradiated, offset correction for correcting a variation in dark output for each photoelectric conversion unit and for each photoelectric conversion unit A second line that detects a defective line of the image sensor for the entire image represented by the second image signal using a second threshold value based on a second image signal obtained by performing shading correction for correcting sensitivity variations. Detection means,
And an image processing program for functioning as a combining unit that combines the detection result of the defective line by the first detection unit and the detection result of the defective line by the second detection unit.

Images