WO2013035255A1 - 画像処理装置およびそれを備えた放射線撮影装置 - Google Patents
画像処理装置およびそれを備えた放射線撮影装置 Download PDFInfo
- Publication number
- WO2013035255A1 WO2013035255A1 PCT/JP2012/005237 JP2012005237W WO2013035255A1 WO 2013035255 A1 WO2013035255 A1 WO 2013035255A1 JP 2012005237 W JP2012005237 W JP 2012005237W WO 2013035255 A1 WO2013035255 A1 WO 2013035255A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- image
- band
- blur
- pixel
- anisotropic
- Prior art date
Links
- 238000012545 processing Methods 0.000 title claims abstract description 134
- 238000003384 imaging method Methods 0.000 title claims description 35
- 230000005855 radiation Effects 0.000 title claims description 34
- 238000000034 method Methods 0.000 claims abstract description 28
- 230000008569 process Effects 0.000 claims abstract description 18
- 238000001514 detection method Methods 0.000 claims description 22
- 238000009499 grossing Methods 0.000 claims description 16
- 238000002601 radiography Methods 0.000 claims 1
- 210000003205 muscle Anatomy 0.000 abstract description 6
- 239000013598 vector Substances 0.000 description 37
- 238000010586 diagram Methods 0.000 description 20
- 239000011159 matrix material Substances 0.000 description 18
- 230000000694 effects Effects 0.000 description 11
- 230000008859 change Effects 0.000 description 6
- 239000000725 suspension Substances 0.000 description 6
- 238000004458 analytical method Methods 0.000 description 4
- 239000003550 marker Substances 0.000 description 4
- 230000007246 mechanism Effects 0.000 description 4
- 230000014509 gene expression Effects 0.000 description 3
- 238000007689 inspection Methods 0.000 description 3
- 230000002194 synthesizing effect Effects 0.000 description 3
- 230000009471 action Effects 0.000 description 2
- 210000004204 blood vessel Anatomy 0.000 description 2
- 210000000988 bone and bone Anatomy 0.000 description 2
- 238000003745 diagnosis Methods 0.000 description 2
- 230000002093 peripheral effect Effects 0.000 description 2
- 230000009467 reduction Effects 0.000 description 2
- QBFXBDUCRNGHSA-UHFFFAOYSA-N 1-(4-fluorophenyl)-2-(methylamino)pentan-1-one Chemical compound FC1=CC=C(C=C1)C(C(CCC)NC)=O QBFXBDUCRNGHSA-UHFFFAOYSA-N 0.000 description 1
- 238000012935 Averaging Methods 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 238000000354 decomposition reaction Methods 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000001678 irradiating effect Effects 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000035515 penetration Effects 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 238000001028 reflection method Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
- 230000007704 transition Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/10—Image enhancement or restoration by non-spatial domain filtering
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B6/00—Apparatus for radiation diagnosis, e.g. combined with radiation therapy equipment
- A61B6/52—Devices using data or image processing specially adapted for radiation diagnosis
- A61B6/5258—Devices using data or image processing specially adapted for radiation diagnosis involving detection or reduction of artifacts or noise
-
- G06T5/70—
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B6/00—Apparatus for radiation diagnosis, e.g. combined with radiation therapy equipment
- A61B6/42—Apparatus for radiation diagnosis, e.g. combined with radiation therapy equipment with arrangements for detecting radiation specially adapted for radiation diagnosis
- A61B6/4208—Apparatus for radiation diagnosis, e.g. combined with radiation therapy equipment with arrangements for detecting radiation specially adapted for radiation diagnosis characterised by using a particular type of detector
- A61B6/4233—Apparatus for radiation diagnosis, e.g. combined with radiation therapy equipment with arrangements for detecting radiation specially adapted for radiation diagnosis characterised by using a particular type of detector using matrix detectors
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B6/00—Apparatus for radiation diagnosis, e.g. combined with radiation therapy equipment
- A61B6/44—Constructional features of apparatus for radiation diagnosis
- A61B6/4429—Constructional features of apparatus for radiation diagnosis related to the mounting of source units and detector units
- A61B6/4452—Constructional features of apparatus for radiation diagnosis related to the mounting of source units and detector units the source unit and the detector unit being able to move relative to each other
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B6/00—Apparatus for radiation diagnosis, e.g. combined with radiation therapy equipment
- A61B6/50—Clinical applications
- A61B6/505—Clinical applications involving diagnosis of bone
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B6/00—Apparatus for radiation diagnosis, e.g. combined with radiation therapy equipment
- A61B6/54—Control of apparatus or devices for radiation diagnosis
- A61B6/542—Control of apparatus or devices for radiation diagnosis involving control of exposure
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10116—X-ray image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10116—X-ray image
- G06T2207/10121—Fluoroscopy
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20016—Hierarchical, coarse-to-fine, multiscale or multiresolution image processing; Pyramid transform
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20172—Image enhancement details
- G06T2207/20192—Edge enhancement; Edge preservation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2210/00—Indexing scheme for image generation or computer graphics
- G06T2210/41—Medical
Definitions
- the present invention relates to an image processing apparatus capable of removing statistical noise that appears in an image in radiation imaging, and a radiation imaging apparatus including the image processing apparatus.
- Radiologists are equipped with radiation imaging devices that acquire images of subjects with radiation.
- the radiation dose to be irradiated during imaging is suppressed as much as possible. This is because it is necessary to prevent unnecessary radiation exposure to the subject.
- This statistical noise is derived, for example, from variations in radiation detection of each of the detection elements of the radiation detector, and tends to appear more prominently as the radiation dose is smaller. Therefore, this false image appears remarkably in a portion such as a bone portion in which the radiation dose tends to be insufficient in the image.
- the edge indicated by the edge reliability is a streak image derived from the structure of the subject. Since the false image derived from statistical noise is granular, it does not appear streak in the image. Therefore, the streak image in the image shows not the statistical noise but the structure of the subject. Therefore, a streak-like structure derived from the subject is reflected in a portion where the edge reliability is high in the image. Therefore, an anisotropic filter is applied to the streak-like structure in image processing.
- the anisotropic filter is applied so as to be smoothed along the streaky structure.
- the orientation of the anisotropic filter that is, the orientation of this streak structure, reduces the effects of noise by isotropically averaging nearby gradient vectors and correcting them using low resolution gradient vectors. And calculated accurately. Thereby, although the statistical noise superimposed on the streak-like structure is erased, the streak-like structure itself is not blurred.
- the anisotropic filter used is changed according to the edge reliability even in the same direction.
- the higher the edge reliability the more the anisotropic filter with higher direction dependency can be applied.
- the conventional configuration has the following problems. That is, according to the conventional radiographic apparatus, there is a problem that a clear image cannot be obtained even if image processing is performed.
- FIG. 20 shows the guide wire reflected in the image.
- the guide wire is a wire that is introduced into the subject during the operation and is not a structure derived from the subject, but is an object that appears in a streak shape in the image and that the operator needs to visually recognize.
- the upper part of FIG. 20 shows a state before image processing is performed.
- the guide wire is shown to be partially thick, but this portion is a balloon marker attached to the guide wire indicating the position of the balloon that dilates the blood vessel.
- the middle part of FIG. 20 shows the state of image processing applied to the upper part of FIG.
- the portion of the image that is not the guide wire has a low edge reliability, so that the weight of the band image increases.
- the portion of the guide wire in the image has high edge reliability, so that the weight of the anisotropic filter becomes large.
- the portion of the guide wire in the image is blurred in the direction of the arrow as shown in the middle part of FIG.
- the balloon marker of the guide wire is also blurred in the direction of the arrow, the balloon marker is stretched in the same direction as shown in the lower part of FIG. That is, when image processing is performed, the shape of a point-like structure such as a balloon marker in the image becomes unnatural. This problem cannot be solved even if the direction of the guide wire is accurately calculated.
- FIG. 21 represents a stage before image processing.
- the image is in a state where two dark and thin stripes intersect.
- the streak portion in the image is subjected to an anisotropic filter because the edge reliability is high.
- the streak portion in the image is blurred in the direction of the arrow as shown in the middle of FIG.
- the present invention has been made in view of such circumstances, and its purpose is not to deteriorate the visibility in a point-like structure or an overlapping portion of two streaks when removing a false image related to statistical noise. It is an object of the present invention to provide a high-speed image processing apparatus and a radiation imaging apparatus including the same.
- an image processing apparatus is an image processing apparatus that processes an image obtained by fluoroscopic imaging of a subject, and (A) a frequency component in an original image in which an image of the subject is reflected.
- Band image generation means for extracting a part and generating a band image
- Gradient calculation means for calculating the gradient magnitude and direction of the pixel value for each pixel of the band image
- C Equal direction to the band image
- An isotropic blur means for generating an isotropic blur image by applying a smoothing filter of the characteristic
- the image processing apparatus in which the visibility at the overlapping portion of two streaks is not deteriorated when a false image related to statistical noise is removed. That is, the image processing apparatus according to the present invention generates an isotropic blur image and an anisotropic blur image from the original image.
- An isometric blur image is an image in which an appropriate false image is removed from a portion of the original image where the subject is not reflected, and an anisotropic blur image is conversely appropriate for the portion where the subject is reflected. It is an image from which a false image has been removed.
- the processed image generation means superimposes these two images on the band image while weighting each pixel, generates a noise band image in which the false image component is extracted from the band image, and the noise band image is the original image. It is used to remove false images that appear in the image.
- the most characteristic feature of the present invention is that the processed image generation means performs image processing by superimposing the band image, the equal direction blurred image, and the anisotropic blurred image on each pixel while changing the weighting based on the edge reliability. That is.
- the edge reliability is an index for determining whether a variation in pixel value seen between a pixel to be processed and surrounding pixels is due to an image of a subject or a false image. For example, when the edge reliability is an intermediate value, the pixel to be processed is located at the intersection of a dot-like image or two linear images that appear in the band image.
- the noise band image is used without using an anisotropic blur image at the point-like image or at the intersection (by using it at a weak intensity compared with the isometric blur image). Is generated.
- the portion corresponding to the point-like image in the anisotropic blur image is stretched in a specific direction and the image is unnatural, and the portion corresponding to the intersection is more intense in the dark line-like image.
- the image is unnatural.
- since such a portion is not used when generating the noise band image, the point-like structure of the subject in the finally obtained processed image or the intersection where the outline of the structure overlaps each other. This part retains visibility without being disturbed. In this way, a processed image with high visibility can be output according to the shape of the structure of the subject in the original image.
- the present invention it is not necessary to hold a plurality of anisotropic filters in each direction as in the conventional configuration, and a storage device for storing these can be small. Further, since it is only necessary to determine which anisotropic filter is used on the condition of only the direction indicated by the gradient according to the image portion, an image processing apparatus with improved processing speed can be provided.
- the isotropic blur means of the image processing apparatus according to the present invention is configured to generate an isotropic blur image by applying an isotropic smoothing filter that does not depend on the original image to (c) the original image
- the anisotropic blur means may be configured to (d) generate an anisotropic blur image by applying an anisotropic smoothing filter depending on the gradient direction to the original image.
- the processed image generation means has a high edge reliability corresponding to the pixel to be processed, and indicates that this pixel is located in a linear image reflected in the band image, etc. It is more desirable to perform the image processing so that the anisotropic blur image is largely subtracted from the band image than the direction blur image.
- the edge reliability corresponding to the pixel to be processed is low and this pixel is not located in the linear image reflected in the band image, the isotropic blur image and anisotropic It is more desirable to perform the processed image so that the characteristic blurred image is not further subtracted from the band image.
- the isotropic blur image is subtracted from the band image more than the anisotropic blur image. It is more desirable to perform image processing.
- the above-described configuration shows a more specific configuration of the image processing apparatus of the present invention. If the image processing is performed so that the isotropic blur image is largely subtracted from the band image rather than the anisotropic blur image in the portion where the edge reliability shows an intermediate value, the point-like shape of the subject appearing in the band image is obtained. It is possible to output a processed image with high visibility without being disturbed in the portion of the intersection where the outline of the structure or the structure overlaps.
- the processed image generation means change the weighting format based on the frequency component of the band image.
- the processed image generation means change the weighting format based on information indicating the exposure amount at the time of capturing the original image.
- the isotropic blur means changes the shape and size of the isotropic smoothing filter based on the information indicating the exposure amount at the time of capturing the original image, or the anisotropic blur means It is more desirable to change the shape and size of the anisotropic smoothing filter based on the information indicating the exposure amount at the time of capturing the original image.
- the above-described configuration shows a more specific configuration of the image processing apparatus of the present invention.
- the isometric direction blur means changes the shape and size of the isotropic smoothing filter based on the information indicating the exposure amount
- the anisotropic smoothing means changes the shape and size of the anisotropic smoothing filter based on the information indicating the exposure amount. If the shape and size are changed, an appropriate false image removal process can be performed according to the exposure amount.
- a radiation source that irradiates radiation
- a radiation source control unit that controls the radiation source
- a detection unit that detects the emitted radiation and outputs a detection signal
- the above-described configuration shows a configuration in which the image processing apparatus of the present invention is incorporated in an actual radiation imaging apparatus. If it is attempted to suppress exposure of the subject in fluoroscopic imaging, a false image derived from statistical noise is easily reflected in the obtained image. Since the false image is erased by the image processing apparatus according to the present invention, an image with excellent visibility can be obtained without re-taking or taking a high dose for the purpose of preventing the false image.
- a radiographic apparatus capable of outputting can be provided.
- the image processing apparatus capable of outputting a processed image with high visibility according to the shape of the structure of the subject that is reflected in the original image when removing the false image related to statistical noise. That is, the image processing apparatus according to the present invention generates an isotropic blur image and an anisotropic blur image from the original image.
- the processed image generation means of the present invention performs image processing by superimposing the band image, the isotropic blur image, and the anisotropic blur image while changing the weighting of each pixel based on the edge reliability. For example, the portion corresponding to the point image in the anisotropic blur image is stretched in a specific direction and the image is unnatural, and the portion corresponding to the intersection point is the dark line image.
- the image is unnatural and the image is unnatural. According to the present invention, since such a part is not used in the false image removal process of the original image, the point-like structure of the subject or the outline of the structure overlapped in the finally obtained processed image. The intersection part maintains visibility without being disturbed. In this way, a processed image with high visibility can be output according to the shape of the structure of the subject in the original image.
- the present invention it is only necessary to hold one anisotropic filter in each direction, and a storage device for storing these is small. Further, since it is only necessary to determine which anisotropic filter is used on the condition of only the direction indicated by the gradient according to the image portion, an image processing apparatus with improved processing speed can be provided.
- FIG. 1 is a functional block diagram illustrating a configuration of an image processing apparatus according to a first embodiment.
- 6 is a schematic diagram illustrating a band image according to Embodiment 1.
- FIG. 6 is a schematic diagram illustrating a band image according to Embodiment 1.
- FIG. 6 is a schematic diagram illustrating a band image according to Embodiment 1.
- FIG. 6 is a schematic diagram illustrating a band image according to Embodiment 1.
- FIG. 3 is a schematic diagram illustrating a configuration of a gradient according to the first embodiment.
- FIG. 1 is a functional block diagram illustrating a configuration of an image processing apparatus according to a first embodiment.
- FIG. 6 is a schematic diagram illustrating a band image according to Embodiment 1.
- FIG. 6 is a schematic diagram illustrating a
- FIG. 3 is a schematic diagram for explaining the operation of the image processing apparatus according to the first embodiment.
- FIG. 3 is a schematic diagram for explaining the operation of the image processing apparatus according to the first embodiment.
- FIG. 3 is a schematic diagram for explaining the operation of the image processing apparatus according to the first embodiment.
- FIG. 3 is a schematic diagram for explaining the operation of the image processing apparatus according to the first embodiment.
- FIG. 3 is a schematic diagram for explaining the operation of the image processing apparatus according to the first embodiment.
- FIG. 3 is a schematic diagram for explaining the operation of the image processing apparatus according to the first embodiment.
- FIG. 3 is a schematic diagram for explaining the operation of the image processing apparatus according to the first embodiment.
- FIG. 3 is a schematic diagram for explaining the operation of the image processing apparatus according to the first embodiment.
- FIG. 3 is a schematic diagram for explaining the operation of the image processing apparatus according to the first embodiment.
- FIG. 3 is a schematic diagram for explaining the operation of the image processing apparatus according to the first embodiment.
- FIG. 3 is
- FIG. 3 is a schematic diagram for explaining the operation of the image processing apparatus according to the first embodiment.
- 6 is a functional block diagram illustrating a configuration of an X-ray imaging apparatus according to Embodiment 2.
- FIG. It is a functional block diagram explaining the structure which concerns on 1 modification of this invention. It is a schematic diagram explaining the subject of a conventional structure. It is a schematic diagram explaining the subject of a conventional structure.
- X-rays in the examples correspond to the radiation of the present invention.
- FPD is an abbreviation for flat panel detector.
- the image processing apparatus 1 when the image processing apparatus 1 according to the first embodiment inputs an image (referred to as an original image P0) acquired by fluoroscopically imaging a subject with X-rays, the entire original image P0 is input.
- the processed image P4 from which the granular false image derived from the statistical noise reflected in is removed is output.
- Statistical noise is noise derived from variations in intensity when a detection pixel included in an FPD that detects X-rays during fluoroscopic imaging detects X-rays, and is associated with detection characteristics of the detection elements. Therefore, the granular false image derived from statistical noise does not disappear even if the FPD is uniformly irradiated with X-rays.
- the image processing apparatus 1 includes (A) a band image generation unit 12 that generates band images ⁇ , ⁇ , ⁇ ... Obtained by extracting frequency components of each band from the original image P0, and (B) bands.
- a gradient calculation unit 13 that calculates a gradient m ( ⁇ , ⁇ , ⁇ %) For each of the images ⁇ , ⁇ , ⁇ ..., And (C) an equal direction blur image for each of the band images ⁇ , ⁇ , ⁇ .
- the equal direction blur part 14 that generates P1 ( ⁇ , ⁇ , ⁇ ...)
- D gradient m ( ⁇ , ⁇ , ⁇ .
- the edge reliability E is acquired based on the anisotropic blur part 15 that generates the anisotropic blur image P2 ( ⁇ , ⁇ , ⁇ %) And (E) the gradient m ( ⁇ , ⁇ , ⁇ ).
- the gradient calculation unit 13 corresponds to the gradient calculation unit of the present invention.
- the band image generation unit 12 corresponds to the band image generation unit of the present invention
- the gradient calculation unit 13 corresponds to the gradient calculation unit of the present invention.
- the equal direction blur part 14 corresponds to the equal direction blur means of the present invention
- the anisotropic blur part 15 corresponds to the anisotropic blur means of the present invention.
- the edge reliability acquisition unit 16 corresponds to the edge reliability acquisition unit of the present invention
- the noise band image generation unit 17 and the processed image generation unit 22 correspond to the processed image generation unit of the present invention.
- the image processing apparatus 1 superimposes the noise band image P3 ( ⁇ , ⁇ , ⁇ %) To generate the noise image N, and the pixel value of the noise image N with reference to the original image P0.
- a processed image generation unit 22 that generates a processed image P4 by superimposing the noise image N after pixel value adjustment on the original image P0.
- FIG. 2 shows the result of frequency analysis of the original image P0.
- the original image P0 has a wide frequency component from high frequency to low frequency. For convenience of explanation, it is assumed that the response of each frequency is 1.
- FIG. 3 shows the result of frequency analysis of the band image ⁇ . As shown in FIG. 3, the band image ⁇ is obtained by extracting frequency components existing in the frequency region on the highest frequency side of the original image P0.
- FIG. 4 shows the result of frequency analysis of the band image ⁇ . As shown in FIG. 4, the band image ⁇ is obtained by extracting a frequency component existing in the frequency region on the second high frequency side of the original image P0.
- the band image ⁇ is obtained by extracting a frequency component existing in the third frequency region on the high frequency side of the original image P0.
- the band images ⁇ , ⁇ , ⁇ have frequency components derived from the high-frequency original image P0 in this order.
- the original image P0 is assumed to be separated into three band images ⁇ , ⁇ , and ⁇ for simplicity of description. However, actually, three or more band images are generated from the original image P0.
- the gradient m represents a change in the pixel value on the band image, and is a data set in which data having characteristic values corresponding to the pixels of the band image are two-dimensionally arranged.
- the data constituting the gradient m is in a vector format, and the length is a value representing how much the pixel value of the pixel on the band image is different from the pixel values of the surrounding pixels. .
- the direction of this vector is the one having the most different pixel value for the central pixel among the peripheral pixels when comparing the peripheral pixels around a certain pixel on the band image. It indicates in which direction from the center pixel.
- the gradient m for the band image filled with the same pixel values is filled with data representing 0 because the pixel values of adjacent pixels are all the same. Each of these data has no direction or length.
- the gradient m corresponding to the band image having a portion where the pixel values are different is shown in the lower part of FIG. That is, as can be seen from the lower part of FIG. 6, the gradient m has a longer data vector length at the transition of pixel values appearing in a ring shape in the band image in the upper part of FIG. 6. This length means the maximum difference among the pixel value differences between the pixel on the band image and the adjacent pixel.
- the direction of the vector represents the characteristics of the pixel on the band image corresponding to the location, and specifically means the direction in which the pixel value changes most steeply when moving from one pixel to the next. ing.
- a vector appears along the contour of the subject image in the band image.
- the direction in which the vector extends and the direction in which the contour of the subject image extends are orthogonal to each other.
- the band image shown in the upper part of FIG. 7 includes a granular false image. This false image is derived from statistical noise.
- the actual gradient m is in a state where a vector representing the contour of the image of the subject and a vector representing the granular false image are mixed.
- the edge reliability E acquired by the edge reliability acquisition unit 16 will be described.
- the edge reliability E is a value for determining whether each vector in the gradient m represents the contour of an image of the subject or a granular false image. Therefore, the edge reliability exists for each vector of the gradient m, and is a feature value corresponding to each pixel of the band image. The higher the edge reliability E, the higher the possibility that the vector represents the contour of the image of the subject.
- the image processing apparatus 1 includes a CPU as hardware resources, and the CPU 11 executes various programs to realize the units 11, 12, 13, 14, 15, 16, 17, 18, 19, 21, and 22. Yes. Further, each of the above-described units may be divided and executed by an arithmetic device that takes charge of them.
- the storage unit 28 stores various parameters, tables, and mathematical expressions used for image processing. The user's designation of image processing can also be stored in advance.
- a band image is first generated (band image generation step S1), and the gradient m is set for each band image. It is calculated (gradient calculation step S2).
- each band image is blurred in the same direction to generate an equal direction blur image P1 (equal direction blur step S3), and each band image is blurred by a filter having anisotropy based on the gradient m.
- an anisotropic blur image P2 is generated (anisotropic blur step S4).
- the edge reliability E is acquired based on the gradient m (edge reliability acquisition step S5), and the noise band image P3 based on the band image, the equal direction blur image P1, the anisotropic blur image P2, and the edge reliability E. Is generated (noise band image generation step S6).
- the noise band image P3 is synthesized to generate a noise image N (noise image generation step S7), and the original image P0 is referred to and the noise image N is subjected to pixel value adjustment (pixel value adjustment step S8).
- the noise image N is superimposed on the original image P0 to generate the processed image P4 (processed image generation step S9).
- ⁇ Band Image Generation Step S1> The operation of the band image generation unit 12 will be described. As shown in FIG. 9, the band image generation unit 12 acquires the band image ⁇ , the band image ⁇ , and the band image ⁇ in this order. Each of these operations will be described in order.
- the following generation method of the band images ⁇ , ⁇ , ⁇ is an improvement of the conventional Laplacian pyramid decomposition.
- the original image P0 is sent to the band image generation unit 12.
- the band image generation unit 12 applies a matrix functioning as a high-pass filter to each of the pixels constituting the original image P0.
- FIG. 10 shows a state when the high-pass filter process is performed on the pixel s constituting the original image P0.
- the band image generation unit 12 reads, for example, a 5 ⁇ 5 high-pass filter matrix from the storage unit 28 and applies this matrix to the pixel s. Then, as shown in FIG. 10, the matrix acts on the pixel region R having a size of 5 rows and 5 columns with the pixel s as the center.
- the band image generation unit 12 arranges the pixel data obtained by applying the matrix at a position corresponding to the pixel s in the band image ⁇ (the same position as the pixel s).
- the band image generation unit 12 performs the same operation for all the pixels other than the pixel s constituting the original image P0, and maps the acquired pixel data to the band image ⁇ in association with the original image P0 each time.
- the band image ⁇ is a rough image in which pixel data changes finely. This high-pass filter process is represented by the symbol HPF in FIG.
- the band image generation unit 12 generates a reduced image p1 obtained by reducing the original image P0 by, for example, halving both vertically and horizontally.
- this image reduction processing is represented by Mag ( ⁇ ).
- the band image generation unit 12 applies a low-pass filter to the reduced image p1. That is, the band image generation unit 12 reads a 5 ⁇ 5 low-pass filter matrix having the same size as the high-pass filter matrix from the storage unit 28, and applies this matrix to each of the pixels constituting the reduced image p1. Act. Pixel data obtained by the action of the matrix is mapped to the low-pass image L1 in correspondence with the reduced image p1. This situation is the same as described with reference to FIG. The difference is that the matrix used is different and the size of the image is reduced. As described above, once the original image P0 is reduced and the low-pass filter is applied, the frequency component can be extracted without enlarging the matrix that defines the low-pass filter, thereby greatly reducing the calculation cost. it can. This low-pass filter processing is represented by the symbol LPF in FIG.
- the band image generation unit 12 generates an enlarged low-pass image M1 obtained by enlarging the low-pass image L1 twice, for example, both vertically and horizontally.
- this image reduction processing is represented by Mag (+). That is, the enlarged low-pass image M1 and the original image P0 have the same size.
- the band image generation unit 12 subtracts the band image ⁇ and the enlarged low-pass image M1 from the original image P0 to generate a band image ⁇ .
- FIG. 11 schematically shows a range of frequency components included in each image.
- the original image P0 has all frequency components as shown in FIG.
- the band image ⁇ is composed of only the highest frequency component
- the enlarged low-pass image M1 is composed of only the low-frequency component of the reduced image p1.
- the band image ⁇ obtained by subtracting the band image ⁇ and the enlarged low-pass image M1 from the original image P0 is an enlarged low-pass image from the lowest frequency of the band image ⁇ among all the frequency components of the original image P0. It has a frequency component in the section sandwiched by the highest frequency that M1 has.
- the band image generation unit 12 reads out a 9 ⁇ 9 band-pass filter matrix, which is approximately twice the size of the high-pass filter matrix, from the storage unit 28, and applies each pixel constituting the reduced image p1. Use the lever matrix. Pixel data obtained by the action of the matrix is mapped to the band image ⁇ in correspondence with the reduced image p1. This situation is the same as described with reference to FIG. The differences are that the type of matrix used is different, the size of the matrix is doubled both vertically and horizontally, and the area of the reduced image p1 to be processed is 1/4 of the original image P0. .
- This band pass filter process is represented by the symbol BPF in FIG.
- the band image ⁇ generated in this manner is obtained by extracting the frequency component of the original image P0 for the band on the lower frequency side than the band image ⁇ .
- the band image generation unit 12 In addition to the reduced image p1, the band image generation unit 12 also generates a reduced image p2 obtained by reducing the reduced image p1 by 1/2 in the vertical and horizontal directions. This reduced image p2 is also subjected to a band pass filter, and a band image ⁇ is generated. The band image ⁇ generated in this way is obtained by extracting the frequency component of the original image P0 for the band on the lower frequency side than the band image ⁇ . As described above, the band image generation unit 12 may generate a band image on the lower frequency side than the band image ⁇ . These band images may also be used for subsequent image processing. However, in the description of the first embodiment, it is assumed that image processing is performed using only the band images ⁇ , ⁇ , and ⁇ for the purpose of simple description.
- the band images ⁇ , ⁇ , ⁇ are sent to the gradient calculation unit 13.
- the gradient calculation unit 13 creates a gradient m ( ⁇ , ⁇ , ⁇ ) by performing a predetermined operation on each of the band images ⁇ , ⁇ , ⁇ . 12 and 13 show how the gradient calculation unit 13 creates a gradient m ⁇ based on the band image ⁇ .
- the gradient calculation unit 13 reads the pixel value (target pixel value) of the target pixel a constituting the band image ⁇ . This pixel value is 115 in FIG. Next, the pixel values (surrounding pixel values) of the eight surrounding pixels b surrounding the target pixel a are read out. This pixel value has various values as shown in FIG.
- the gradient calculation unit 13 compares the eight surrounding pixel values and selects the surrounding pixel b having the largest difference (pixel value difference) between the target pixel value and the surrounding pixel value among the eight surrounding pixel values.
- the selected surrounding pixel b is shown surrounded by a broken line in FIG.
- the gradient calculating unit 13 arranges a vector at a position corresponding to the target pixel a on the gradient m ⁇ (the same position as the target pixel a).
- the length of this vector represents the pixel value difference for the selected pixel
- the direction of the vector represents the direction in which the selected pixel exists when viewed from the target pixel a.
- the gradient calculation unit 13 calculates the gradient m ⁇ for the entire band image ⁇ while changing the target pixel a.
- the operation of the gradient calculation unit 13 for other band images is the same as this operation.
- the gradient calculation unit 13 calculates the magnitude and direction of the gradient of the pixel value for each pixel of the band images ⁇ , ⁇ , and ⁇ .
- the band images ⁇ , ⁇ , ⁇ are also sent to the equal direction blur unit 14.
- the equal direction blur unit 14 applies a Gaussian filter to each of the band images ⁇ , ⁇ , and ⁇ to generate an equal direction blur image P1 ( ⁇ , ⁇ , ⁇ ).
- the equal direction blur unit 14 operates a matrix that defines a Gaussian filter while changing the target pixel, and generates an image by two-dimensionally arranging each of the values obtained at this time.
- the isotropic blur part 14 blurs both the structure of the subject and the granular false image derived from the statistical noise.
- the operation of the equal direction blur unit 14 for other band images is the same as this operation.
- the shape and size of the Gaussian filter may be changed based on information indicating the exposure amount at the time of shooting. That is, the larger the irradiation dose, the smaller the shape and size of the Gaussian filter in advance, and the higher the irradiation dose, the less noise is reflected, and the less blurred blur isotropic blur image P1 ( ⁇ , ⁇ , ⁇ ). Will be acquired.
- ⁇ Anisotropic blur step S4> The gradient m ( ⁇ , ⁇ , ⁇ ) is sent to the anisotropic blur unit 15.
- the anisotropic blur unit 15 applies an anisotropic blur filter as shown in FIG. 14 to the band images ⁇ , ⁇ , ⁇ to generate an anisotropic blur image P2 ( ⁇ , ⁇ , ⁇ ).
- this anisotropic blur filter is applied to one of the pixels of the image, a blurring effect is applied so that the pixel spreads on both sides in a certain direction.
- an anisotropic blur image P2 ⁇ is generated.
- an anisotropic blur filter will be described.
- an anisotropic blur filter is applied in the vertical direction to an image in which an image is reflected at the center as shown in FIG. 15, the circular image is blurred and stretched in the vertical direction as shown on the left side of FIG. It is.
- the right side of FIG. 16 shows a case where an isotropic filter is applied to the image of FIG.
- an isotropic filter is applied to the image, the image reflected in the image is blurred so as to spread as it is.
- an anisotropic blur filter is applied to each pixel.
- the direction in which the anisotropic blur filter is applied is different for each pixel. That is, the anisotropic blur unit 15 determines the anisotropic blur filter with reference to the direction indicated by each data of the gradient m ⁇ . If the anisotropic blur unit 15 applies an anisotropic blur filter to a pixel in the band image ⁇ , the anisotropic blur filter is applied in a direction orthogonal to the direction of vector data on the gradient m ⁇ corresponding to the pixel. The operation of the anisotropic blur unit 15 for other band images is the same as this operation.
- the anisotropic blur filter is on the left side of FIG. As indicated by the solid line arrow in FIG. 5, the band is applied in the direction along the contour of the subject image in the band image ⁇ . Then, even if the anisotropic blur filter is applied, the image of the subject is not blurred. On the other hand, the granular false image superimposed on the contour of the image of the subject is blurred and cannot be visually recognized.
- the shape and size of the anisotropic blur filter may be changed based on information indicating the exposure amount at the time of photographing. That is, the larger the irradiation dose, the smaller the shape and size of the anisotropic blur filter in advance, which reflects the property of less noise as the irradiation dose increases, and the anisotropic blur image P2 ( ⁇ , less blur). ⁇ , ⁇ ) is acquired.
- the anisotropic blur image P2 ⁇ seems to have improved visibility.
- the anisotropic blur filter also acts on a portion that is not the contour of the image of the subject. Therefore, in this portion, the slight directionality of the pixel values of the band image ⁇ is emphasized, so that the image becomes fluffy as shown on the right side of FIG.
- ⁇ Edge reliability acquisition step S5> The gradient m ( ⁇ , ⁇ , ⁇ ) is also sent to the edge reliability acquisition unit 16.
- the edge reliability acquisition unit 16 whether a difference in pixel value indicated by data from a certain data on the gradient m ( ⁇ , ⁇ , ⁇ ) and data of a band image adjacent thereto is derived from noise.
- Edge reliability E which is an index indicating whether or not, is obtained for each piece of data. That is, the edge reliability acquisition unit 16 reads the length of the vector of the target data constituting the gradient m (target vector length Vt).
- the edge reliability acquisition unit 16 acquires the edge reliability E from each value based on the following expression.
- the edge reliability E corresponds to an index of the present invention.
- E Average of Vt / Vn
- the edge reliability E is an index indicating how long the target vector is with respect to its surrounding vectors. Assuming that a part of the band image corresponding to the target vector only includes a non-directional granular false image, since the surrounding vector is long even if the target vector is long, only low edge reliability E is obtained. I can't get it.
- the contour of the subject is reflected in a part of the band image corresponding to the target vector, this part has directionality. That is, the target vector located at the contour of the subject is significantly longer than the surrounding vector. Therefore, a high edge reliability E is obtained for such a target vector.
- the intersection point where the contours of the subject intersect in the band image is determined in a direction perpendicular to the dark linear image, and the length of the vector is also suppressed. Therefore, the edge reliability E at this intersection is an intermediate value.
- the anisotropic blur image P2 ( ⁇ , ⁇ , ⁇ )
- the dark linear image is more emphasized, so the visibility is poor.
- the edge reliability E in this point-like structure is an intermediate value.
- This portion of the anisotropic blur image P2 ( ⁇ , ⁇ , ⁇ ) has poor visibility because the point image is extended in a specific direction.
- the edge reliability acquisition unit 16 separately calculates the edge reliability E for the gradient m ( ⁇ , ⁇ , ⁇ ). Assuming that the gradient m ⁇ has pixels arranged in the vertical and horizontal directions of 1,000 ⁇ 1,000, the edge reliability E is obtained 250,000 times for three gradients m ( ⁇ , ⁇ , ⁇ ). Will be.
- the edge reliability E is sent to the noise band image generation unit 17. Based on the edge reliability E, the noise band image generation unit 17 uses the band images ⁇ , ⁇ , ⁇ , the equal direction blur image P1 ( ⁇ , ⁇ , ⁇ ), and the anisotropic blur image P2 ( ⁇ , ⁇ , ⁇ ). ) Are superimposed while changing the weighting for each pixel to generate a noise band image P3 ( ⁇ , ⁇ , ⁇ ) in which only noise components are extracted from the band images ⁇ , ⁇ , ⁇ .
- the noise band image generation unit 17 extracts noise components from the band images ⁇ , ⁇ , ⁇ by subtracting components that are not noise components from the band images ⁇ , ⁇ , ⁇ . At that time, the noise band image generation unit 17 is the same on each of the band images ⁇ , ⁇ , ⁇ , the equal direction blur image P1 ( ⁇ , ⁇ , ⁇ ), and the anisotropic blur image P2 ( ⁇ , ⁇ , ⁇ ). A new pixel value is obtained by superimposing the pixel values of the three pixels at the positions while weighting them. That is, the noise band image generation unit 17 acquires the pixel value while changing the position of the pixel on the band image ⁇ , ⁇ , ⁇ .
- the acquired pixel values are arranged two-dimensionally following the band images ⁇ , ⁇ , ⁇ , and a noise band image P3 ( ⁇ , ⁇ , ⁇ ) is generated. That is, the band images ⁇ , ⁇ , ⁇ , ⁇ , the equal direction blur image P1 ( ⁇ , ⁇ , ⁇ ), and the anisotropic blur image P2 ( ⁇ , ⁇ , ⁇ ) are combined into one image.
- the noise band image P3 ( ⁇ , ⁇ , ⁇ ) generated at this time is just the noise extracted from the band images ⁇ , ⁇ , ⁇ .
- a method for generating the noise band image P3 ⁇ for the band image ⁇ will be described.
- the noise band images P3 ( ⁇ , ⁇ ) are generated by the same operation.
- a specific configuration when the noise band image generation unit 17 performs image processing on a processing target pixel (target pixel) will be described.
- a new pixel based on the pixel values of the three target pixels in the noise band image generation unit 17 at the same position and in each of three different types of images (band image, equidirectional blur image, and anisotropic blur image).
- the operation for acquiring the value varies depending on the value of the edge reliability E corresponding to the pixel. That is, when the edge reliability E corresponding to the target pixel is high and indicates that this pixel is located in a linear image reflected in the band image ⁇ , the noise band image generation unit 17 uses the isotropic direction blur image P1 ⁇ .
- the anisotropic blur image P2 ⁇ is greatly subtracted from the band image ⁇ . Since this portion is a portion in which the contour of the subject is reflected, the contour of the subject from which noise has been removed is clearly reflected in the anisotropic blur image P2 ⁇ . Therefore, if the noise band image generation unit 17 preferentially subtracts the anisotropic blur image P2 ⁇ from the band image ⁇ , the subject component of the band image ⁇ is removed and the noise component remains.
- the noise band The image generation unit 17 performs image processing so that the isotropic direction blur image P1 ⁇ is more subtracted from the band image ⁇ than the anisotropic blur image P2 ⁇ . Since this portion is a portion where the point-like structure of the subject and the intersection of the contour are reflected, the anisotropic blur image P2 ⁇ is not suitable for visibility.
- the noise component existing in the band image ⁇ is blurred in the equal direction blurred image P1 ⁇ , and the noise component is reduced. Therefore, if the noise band image generation unit 17 preferentially subtracts the equal direction blurred image P1 ⁇ from the band image ⁇ , the component of the subject of the band image ⁇ is removed and the noise component remains.
- the noise band image generation unit 17 has a high edge reliability E.
- the processed image is performed so that the equal direction blurred image P1 ⁇ and the anisotropic blurred image P2 ⁇ are not further subtracted from the band image ⁇ .
- This portion is a portion where the image of the subject is not reflected, and the band image ⁇ contains a lot of noise components. Therefore, in this portion, a noise component can be extracted without performing much subtraction processing from the band image ⁇ .
- the noise band image P3 ( ⁇ , ⁇ , ⁇ ) generated by the noise band image generation unit 17 in this way is obtained by extracting only noise components from the band images ⁇ , ⁇ , ⁇ .
- a method for determining the edge reliability E will be described. Whether or not the edge reliability E is high is determined by the noise band image generation unit 17 based on a table or a mathematical formula stored in the storage unit 28.
- the table is a data set in which the edge reliability E, a coefficient indicating how much the equal direction blurred image P1 ⁇ is subtracted, and a coefficient indicating how much the anisotropic blurred image P2 ⁇ is subtracted are associated with each other.
- the noise band image generation unit 17 reads the edge reliability E corresponding to the target pixel, and acquires the coefficient corresponding to the edge reliability E from the table. Then, the noise band image generation unit 17 performs a subtraction process on the equal direction blur image P1 ⁇ and the anisotropic blur image P2 ⁇ with respect to the band image ⁇ based on the acquired coefficient.
- This table may be changed according to the bands of the band images ⁇ , ⁇ , ⁇ , or the weighting mode may be changed based on information indicating the exposure amount at the time of photographing the original image P0.
- the noise band image generation unit 17 uses mathematical expressions, the above coefficients can be derived by substituting the edge reliability E.
- the table value is changed so that the band image has a lower frequency.
- the relationship between the pixel value of the original image P0 and the incident dose is measured in advance, and the table value is changed so as to increase as the pixel value of the original image P0 increases.
- the noise band image P3 ( ⁇ , ⁇ , ⁇ ) with less noise is acquired reflecting the property that the lower the frequency and the larger the exposure amount, the less the noise.
- the pixel value of the enlarged low-pass image M1 with less noise may be used as an exposure amount index.
- the noise band image P3 ( ⁇ , ⁇ , ⁇ ) is sent to the band synthesizing unit 19. Then, the band synthesizing unit 19 generates a noise image N by superimposing the noise band images P3 ( ⁇ , ⁇ , ⁇ ) while adding weights to each other.
- the noise image N represents noise for all frequencies of the original image P0.
- the weighting when the band synthesizing unit 19 superimposes the noise band images P3 ( ⁇ , ⁇ , ⁇ ) can be changed according to the inspection purpose.
- the noise image N is sent to the pixel value adjustment unit 21.
- the pixel value adjusting unit 21 adjusts the pixel value for each of the pixels constituting the noise image N while referring to the original image P0. That is, the pixel value adjustment unit 21 refers to the pixel value of the pixel of the original image P0 corresponding to the pixel to be processed (target pixel) on the noise image N, and sets the pixel value of the target pixel according to this pixel value. adjust. By adding this process, the visibility of the finally obtained image is improved.
- the manner of adjusting the pixel value performed by the pixel value adjusting unit 21 can be changed according to the inspection purpose.
- the noise image N subjected to the pixel value adjustment is sent to the processed image generation unit 22.
- the processed image generation unit 22 subtracts the noise image N from the original image P0 while applying a weight. Since the noise image N is obtained by extracting a noise component from the original image P0, the noise component on the original image P0 is erased by the subtraction process of the noise image N. The noise component is suppressed in the processed image P4 generated in this way.
- the edge reliability E is an intermediate value in the original image P0 (more precisely, the band images ⁇ , ⁇ , ⁇ )
- the influence of the anisotropic blur image P2 ⁇ is suppressed and the processed image P4 is generated.
- the image is not disturbed at the intersection point where the outline of the point-like structure or the subject overlaps in the processed image P4. Further, by adjusting the weight of the noise image N, the strength of processing can be easily adjusted.
- the weight of the noise image N in the processed image generation unit 22 can be changed according to the inspection purpose.
- a high-speed image processing device 1 that does not deteriorate the visibility in a point-like structure or an overlapping portion of two streaks when a false image related to statistical noise is removed is provided. It can. That is, the image processing apparatus 1 according to the configuration of the first embodiment generates an isotropic blur image P1 and an anisotropic blur image P2 from the original image P0 (more precisely, the band images ⁇ , ⁇ , ⁇ ).
- the equal direction blurred image P1 is an image in which an appropriate false image is removed from a portion of the original image P0 where the subject is not reflected, and the anisotropic blurred image P2 is a portion where the subject is reflected.
- the noise band image generation unit 17 superimposes these two images on the band images ⁇ , ⁇ , ⁇ while weighting each pixel, and the noise band in which the false image components are extracted from the band images ⁇ , ⁇ , ⁇ .
- An image P3 is generated, and this noise band image P3 is used to remove a false image reflected in the original image P0.
- the most characteristic feature of the configuration of the first embodiment is that when the edge reliability E corresponding to the pixel to be processed by the noise band image generation unit 17 is intermediate, the isotropic direction blur image P1 is a band rather than the anisotropic blur image P2.
- the image processing is performed so that the image is greatly subtracted from the images ⁇ , ⁇ , and ⁇ .
- the edge reliability E is an index for determining whether the variation of the pixel value seen between the pixel to be processed and the surrounding pixels is due to the subject image or the false image.
- the edge reliability E is an intermediate value
- the pixel to be processed is located at the intersection of a point-like structure or two linear images that appear in the band images ⁇ , ⁇ , and ⁇ .
- the noise band image is used without using the anisotropic blur image P2 at this intersection (by using it with a weaker intensity than the equal direction blur image P1 even if it is used).
- P3 is generated.
- a slight deviation in pixel values is extremely emphasized, and the image is unnatural.
- the configuration of the first embodiment since such a portion is not used when the noise band image P3 is generated, the intersection of the contours of the structure of the subject in the finally obtained processed image P4 is overlapped. The part maintains visibility without being disturbed.
- the configuration of the first embodiment it is not necessary to hold a plurality of anisotropic filters in each direction as in the conventional configuration, and a storage device for storing these can be small. Further, since it is only necessary to determine which anisotropic filter is used on the condition of only the direction indicated by the gradient according to the image portion, an image processing apparatus with improved processing speed can be provided.
- the image processing is performed so that the anisotropic blur image P2 is largely subtracted from the band images ⁇ , ⁇ , and ⁇ than the equal direction blur image P1 in the portion where the edge reliability E is high, the band images ⁇ , ⁇ , It is possible to output a processed image P4 with high visibility without blurring the linear image reflected on ⁇ .
- the noise band image generation unit 17 changes the weighting mode based on the frequency component of the band image, an appropriate false image removal process can be performed according to the frequency component.
- the noise band image generation unit 17 changes the weighting format based on the information indicating the exposure amount, an appropriate false image removal process can be performed according to the exposure amount.
- the shape and size of the Gaussian filter is changed based on the information indicating the exposure amount at the time of shooting by the isotropic blur portion 14, or the difference is determined based on the information indicating the exposure amount at the time of shooting by the anisotropic blur portion 15. If the shape and size of the isotropic blur filter are changed, an appropriate false image removal process can be performed according to the exposure amount.
- the X-ray imaging apparatus 20 according to the second embodiment includes an image processing apparatus 1 according to the first embodiment (shown as an image processing unit 32 in FIG. 18) as illustrated in FIG. It is a device. Therefore, in the X-ray imaging apparatus 20 according to the second embodiment, the configuration and operation description of the image processing unit 32 according to the first embodiment will be omitted.
- the X-ray imaging apparatus 20 is configured to image a standing subject M.
- the support 2 extending in the vertical direction v from the floor surface, and X for irradiating X-rays. It has a line tube 3, an FPD 4 supported by the support column 2, and a suspension support 7 that extends in the vertical direction v and is supported by the ceiling.
- the suspension support 7 supports the X-ray tube 3 in a suspended manner.
- the X-ray tube 3 corresponds to the radiation source of the present invention
- the FPD 4 corresponds to the detection means of the present invention.
- the FPD 4 can slide in the vertical direction v with respect to the support column 2. Moreover, the suspension support body 7 is extendable in the vertical direction v, and the position of the X-ray tube 3 in the vertical direction v is changed as the suspension support body 7 expands and contracts.
- the movement of the FPD 4 in the vertical direction v with respect to the support 2 is performed by an FPD moving mechanism 35 provided between the two and the four. This is controlled by the FPD movement control unit 36.
- the movement of the X-ray tube 3 will be described.
- the X-ray tube 3 is performed by an X-ray tube moving mechanism 33 provided on the suspension support 7.
- the X-ray tube movement control unit 34 is provided for the purpose of controlling the X-ray tube movement mechanism 33.
- the X-ray tube 3 is moved by the X-ray tube moving mechanism 33 (1) in the vertical direction v, (2) in the approach / separation direction with respect to the FPD 4, and (3) in the horizontal direction S orthogonal to the direction from the X-ray tube 3 toward the FPD 4 (see FIG. 18 in the paper surface penetration direction and the body side direction of the subject M).
- the suspension support 7 expands and contracts.
- the FPD 4 has a detection surface 4a (see FIG. 18) for detecting X-rays.
- the detection surface 4a is arranged in the X-ray imaging apparatus 20 upright in the vertical direction v. Thereby, the standing subject M can be efficiently imaged.
- the detection surface 4 a is disposed so as to face the X-ray irradiation port of the X-ray tube 3.
- the detection surface 4a is arranged along a plane formed by two directions of the horizontal direction S and the vertical direction v. Further, the detection surface 4a is rectangular, and one side is in the horizontal direction S, and the other side orthogonal to the one side is in the vertical direction v.
- the X-ray tube controller 6 controls the tube voltage, tube current, and X-ray irradiation time of the X-ray tube 3.
- the X-ray tube control unit 6 controls the X-ray tube 3 so as to output radiation with a predetermined tube current, tube voltage, and pulse width. Parameters such as tube current are stored in the storage unit 37.
- the X-ray tube control unit 6 corresponds to the radiation source control means of the present invention.
- the image generation unit 31 assembles the detection data output from the FPD 4 and generates the original image P0 in which the projection image of the subject M is reflected.
- the image processing unit 32 removes the false image derived from statistical noise reflected in the original image P0 and generates a processed image P4.
- the image generation unit 31 corresponds to the image generation unit of the present invention.
- the operation console 38 is provided for the purpose of inputting each instruction of the surgeon, and various instructions for the image processing unit 32 are also performed through the operation console 38.
- the storage unit 37 stores all of various parameters used for X-ray imaging such as control information of the X-ray tube 3, position information of the X-ray tube 3, and position information of the FPD 4 in the vertical direction v.
- the X-ray imaging apparatus 20 includes a main control unit 41 that comprehensively controls the units 6, 34, 36, 31, and 32.
- the main control unit 41 is constituted by a CPU, and realizes each unit by executing various programs. Further, each of the above-described units may be divided and executed by an arithmetic device that takes charge of them.
- the display unit 39 is provided for the purpose of displaying the captured processed image P4.
- the X-ray tube control unit 6 emits pulsed X-rays according to the irradiation time, tube current, and tube voltage stored in the storage unit 37.
- the FPD 4 detects X-rays transmitted through the subject and outputs a detection signal to the image generation unit 31.
- the image generation unit 31 generates an original image P0 in which a fluoroscopic image of the subject M and a false image derived from statistical noise are reflected based on each detection signal.
- the original image P0 is converted into the processed image P4 from which the false image is removed by the image processing unit 32.
- the processed image P4 is displayed on the display unit 39, and the imaging operation by the X-ray imaging apparatus 20 ends.
- the configuration of the second embodiment is a configuration in which the image processing apparatus 1 having the configuration of the first embodiment is incorporated in an actual radiation imaging apparatus. If it is attempted to suppress exposure of the subject in fluoroscopic imaging, a false image derived from statistical noise is easily reflected in the obtained image. Since the false image is erased by the image processing apparatus 1 according to the configuration of the first embodiment, the visibility can be improved without re-shooting or shooting with a strong dose for the purpose of preventing the false image. A radiation imaging apparatus capable of outputting an excellent image can be provided.
- the present invention is not limited to the above-described configuration, and can be modified as follows.
- the band images ⁇ , ⁇ , and ⁇ are input to the equal direction blur part 14 and the anisotropic blur part 15, but instead of this, as shown in FIG.
- the image P0 may be input.
- the units 14 and 15 at this time unlike the embodiment, perform processing for extracting a specific frequency in the original image P0 simultaneously with the blur processing.
- the equal direction blur image P1 ( ⁇ , ⁇ , ⁇ ) and the anisotropic blur image P2 ( ⁇ , ⁇ , ⁇ ) are generated from the original image P0 in one step, so that the configuration is further simplified.
- An image processing apparatus that operates quickly can be provided.
- X-rays referred to in the above-described embodiments are an example of radiation in the present invention. Therefore, the present invention can be applied to radiation other than X-rays.
- the image processing apparatus of the present invention is suitable for the medical field.
Abstract
Description
すなわち、従来の放射線撮影装置によれば、画像処理を行っても鮮明な画像を取得されないという問題点がある。
すなわち、本発明に係る画像処理装置は、被検体を透視撮影することで得られる画像を処理する画像処理装置であって、(A)被検体の像が写り込んでいる元画像における周波数成分の一部を抽出して帯域画像を生成する帯域画像生成手段と、(B)帯域画像の画素各々について画素値の勾配の大きさと方向を算出する勾配算出手段と、(C)帯域画像に等方向性のスムージングフィルタを作用させて、等方向ボカシ画像を生成する等方向ボカシ手段と、(D)帯域画像に勾配の方向に依存した異方性のスムージングフィルタを作用させて、異方性ボカシ画像を生成する異方性ボカシ手段と、(E)画素各々について勾配の大きさと、当該画素の周囲における帯域画像の画素値とに基づいて、指標を算出するエッジ信頼度取得手段と、(F)帯域画像、等方向ボカシ画像、および異方性ボカシ画像とを画素の各々で指標に基づく重み付けにより重ね合わせて、元画像からノイズ成分が除去された処理画像を生成する処理画像生成手段とを備えることを特徴とするものである。
実施例1に係る画像処理装置1は、(A)元画像P0から各帯域の周波数成分が抽出された帯域画像α,β,γ……を生成する帯域画像生成部12と、(B)帯域画像α,β,γ……の各々について勾配m(α,β,γ……)を算出する勾配算出部13と、(C)帯域画像α,β,γ……の各々について等方向ボカシ画像P1(α,β,γ……)を生成する等方向ボカシ部14と、(D)勾配m(α,β,γ……)を参照して帯域画像α,β,γ……の各々について異方性ボカシ画像P2(α,β,γ……)を生成する異方性ボカシ部15と、(E)勾配m(α,β,γ……)を基にエッジ信頼度Eを取得するエッジ信頼度取得部16と、(F)エッジ信頼度Eに基づいて帯域画像α,β,γ……、等方向ボカシ画像P1(α,β,γ……),および異方性ボカシ画像P2(α,β,γ……)を重ね合わせてノイズ帯域画像P3(α,β,γ……)を生成するノイズ帯域画像生成部17とを備えている。勾配算出部13は、本発明の勾配算出手段に相当する。
次に、画像処理装置1の動作について説明する。実施例1に係る画像処理装置1を用いて処理画像P4を生成するには、図8に示す様に、まず帯域画像が生成され(帯域画像生成ステップS1),帯域画像の各々について勾配mが算出される(勾配算出ステップS2)。次に、帯域画像の各々が等方向にボカされて等方向ボカシ画像P1が生成される(等方向ボカシステップS3),勾配mを基に帯域画像の各々が異方性を有するフィルタによりにボカされて異方性ボカシ画像P2が生成される(異方性ボカシステップS4)。そして、勾配mを基にエッジ信頼度Eが取得され(エッジ信頼度取得ステップS5),帯域画像、等方向ボカシ画像P1,異方性ボカシ画像P2,エッジ信頼度Eに基づいてノイズ帯域画像P3が生成される(ノイズ帯域画像生成ステップS6)。そして、ノイズ帯域画像P3が合成されてノイズ画像Nが生成され(ノイズ画像生成ステップS7),元画像P0が参照されてノイズ画像Nに画素値調整が施された後(画素値調整ステップS8),ノイズ画像Nが元画像P0に重ねられて処理画像P4が生成される(処理画像生成ステップS9)。以降、これら各ステップについて順を追って説明する。
帯域画像生成部12の動作について説明する。帯域画像生成部12は、図9に示すように帯域画像α,帯域画像β,帯域画像γをこの順に取得する。これら各動作について順を追って説明する。下記の帯域画像α,β,γの生成方法は、従来のラプラシアンピラミッド分解を改良したものとなっている。
帯域画像α,β,γは、勾配算出部13に送出される。勾配算出部13では、帯域画像α,β,γの各々に所定の操作をすることで、勾配m(α,β,γ)を作成する。図12,図13は、この勾配算出部13が帯域画像αを基に勾配mαを作成する様子を表している。勾配算出部13は、帯域画像αを構成する対象画素aの画素値(対象画素値)を読み出す。この画素値は図12においては、115となっている。次に、対象画素aを囲む8つの周囲画素bの画素値(周囲画素値)を読み出す。この画素値は図12に示す様に様々な値をとっている。勾配算出部13は、8つの周囲画素値とを比較し、8つの周囲画素値の中で対象画素値と周囲画素値との差(画素値差)が最も大きい周囲画素bを選択する。選択された周囲画素b(選択画素)は、図12においては破線で囲んで表されている。
帯域画像α,β,γは、等方向ボカシ部14にも送出されている。等方向ボカシ部14は、帯域画像α,β,γの各々にガウシアンフィルタを施し、等方向ボカシ画像P1(α,β,γ)を生成する。等方向ボカシ部14は、ガウシアンフィルタを規定する行列を作用対象の画素を変更しながら作用させ、このとき得られた値の各々を2次元的に配列することにより画像を生成する。等方向ボカシ部14により、被検体の構造体および統計ノイズに由来する粒状の偽像は、ともにぼかされることになる。他の帯域画像についての等方向ボカシ部14の動作は、この動作と同様である。
勾配m(α,β,γ)は、異方性ボカシ部15に送出される。異方性ボカシ部15は、帯域画像α,β,γに図14のような異方性ボカシフィルタを施し、異方性ボカシ画像P2(α,β,γ)を生成する。この異方性ボカシフィルタを画像の画素の1つにかけると、その画素がある方向の両側に広がるようにぼかしの効果がかかる。異方性ボカシ部15がこのボカシ動作を帯域画像αに施すと異方性ボカシ画像P2αが生成される。
勾配m(α,β,γ)はエッジ信頼度取得部16にも送出される。エッジ信頼度取得部16では、勾配m(α,β,γ)上のあるデータとこれに隣接する帯域画像のデータとから、そのデータの示す画素値の差がノイズに由来するものであるかどうか示す指標であるエッジ信頼度Eをデータの各々について取得する。すなわち、エッジ信頼度取得部16は、勾配mを構成する対象データのベクトルの長さ(対象ベクトル長Vt)を読み出す。次に、帯域画像の対象データを囲む8つの周囲データにおける画素値を読み出し、対象データの画素値と差分を取り、8つの周囲データにおけるベクトルの長さ(周囲ベクトル長Vn)を読み出す。エッジ信頼度取得部16は、それぞれの値からエッジ信頼度Eを次の式に基づいて取得する。エッジ信頼度Eは本発明の指標に相当する。
E=Vt/Vnの平均
エッジ信頼度Eは、ノイズ帯域画像生成部17に送出される。ノイズ帯域画像生成部17では、このエッジ信頼度Eに基づいて帯域画像α,β,γ,等方向ボカシ画像P1(α,β,γ),および異方性ボカシ画像P2(α,β,γ)を画素の各々で重み付けを変更しながら重ね合わせて、帯域画像α,β,γからノイズ成分のみが抽出されたノイズ帯域画像P3(α,β,γ)を生成する。
ノイズ帯域画像P3(α,β,γ)は、帯域合成部19に送出される。そして、帯域合成部19はノイズ帯域画像P3(α,β,γ)を互いに重み付けを加えながら重ね合わせてノイズ画像Nを生成する。このノイズ画像Nは、元画像P0の全周波数についてのノイズが表されている。帯域合成部19がノイズ帯域画像P3(α,β,γ)を重ね合わせるときの重み付けは、検査目的に応じて変更可能となっている。
ノイズ画像Nは画素値調整部21に送出される。画素値調整部21では、ノイズ画像Nを構成する画素の各々について元画像P0を参照しながら画素値の調整を行う。すなわち、画素値調整部21は、ノイズ画像N上の処理対象の画素(対象画素)に対応する元画像P0の画素の画素値を参照して、この画素値に応じて対象画素の画素値を調整する。この処理を加えることで、最終的に得られる画像の視認性が向上する。画素値調整部21が行う画素値を調整の様式は、検査目的に応じて変更可能となっている。
画素値調整が施されたノイズ画像Nは、処理画像生成部22に送出される。処理画像生成部22は、元画像P0からノイズ画像Nに重みをかけて減算する。ノイズ画像Nは、元画像P0からノイズ成分を抽出したものとなっていることからすれば、元画像P0上のノイズ成分はノイズ画像Nの減算処理により消去される。こうして生成された処理画像P4には、ノイズ成分が抑制されている。しかも、元画像P0(正確には帯域画像α,β,γ)におけるエッジ信頼度Eが中間的な値の部分では、異方性ボカシ画像P2αの影響が抑制されて処理画像P4が生成されているので、処理画像P4における点状の構造物や被検体の輪郭が重なった交点の部分において、画像が乱れることがない。また、ノイズ画像Nの重みを調整することで、処理の強さを簡便に調整することができる。処理画像生成部22におけるノイズ画像Nの重みは、検査目的に応じて変更可能となっている。
次に、X線撮影装置20の動作について説明する。撮影に先立って、被検体MがX線管3とFPD4とに挟まれる位置に起立される。これにより、X線撮影装置20に被検体Mが載置されたことになる。術者が操作卓38を通じてX線管3およびFPD4の位置の調整を行うと、X線管3およびFPD4はそれぞれの移動を制御する制御部34,36の制御に従って、被検体Mの撮影領域まで移動する。
4 FPD(検出手段)
6 X線管制御部(放射線源制御手段)
a 対象画素
b 周囲画素
m 勾配
E エッジ信頼度
P0 元画像
P1 等方向ボカシ画像
P2 異方性ボカシ画像
P3 ノイズ帯域画像
P4 処理画像
α,β,γ 帯域画像
12 帯域画像生成部(帯域画像生成手段)
13 勾配算出部(勾配算出手段)
14 等方向ボカシ部(等方向ボカシ手段)
15 異方性ボカシ部(異方性ボカシ手段)
16 エッジ信頼度取得部(エッジ信頼度取得手段)
17 ノイズ帯域画像生成部(処理画像生成手段)
22 処理画像生成部(処理画像生成手段)
31 画像生成部(画像生成手段)
Claims (9)
- 被検体を透視撮影することで得られる画像を処理する画像処理装置であって、
(A)被検体の像が写り込んでいる元画像における周波数成分の一部を抽出して帯域画像を生成する帯域画像生成手段と、
(B)前記帯域画像の画素各々について画素値の勾配の大きさと方向を算出する勾配算出手段と、
(C)前記帯域画像に等方向性のスムージングフィルタを作用させて、等方向ボカシ画像を生成する等方向ボカシ手段と、
(D)前記帯域画像に前記勾配の方向に依存した異方性のスムージングフィルタを作用させて、異方性ボカシ画像を生成する異方性ボカシ手段と、
(E)画素各々について前記勾配の大きさと、当該画素の周囲における前記帯域画像の画素値とに基づいて、指標を算出するエッジ信頼度取得手段と、
(F)前記帯域画像、前記等方向ボカシ画像、および前記異方性ボカシ画像とを画素の各々で前記指標に基づく重み付けにより重ね合わせて、前記元画像からノイズ成分が除去された処理画像を生成する処理画像生成手段とを備えることを特徴とする画像処理装置。 - 被検体を透視撮影することで得られる画像を処理する画像処理装置であって、
(A)被検体の像が写り込んでいる元画像における周波数成分の一部を抽出して帯域画像を生成する帯域画像生成手段と、
(B)前記帯域画像の画素各々について画素値の勾配の大きさと方向を算出する勾配算出手段と、
(c)前記元画像に等方向性のスムージングフィルタを作用させて、等方向ボカシ画像を生成する等方向ボカシ手段と、
(d)前記元画像に前記勾配の方向に依存した異方性のスムージングフィルタを作用させて、異方性ボカシ画像を生成する異方性ボカシ手段と、
(E)画素各々について前記勾配の大きさと、当該画素の周囲における前記帯域画像の画素値とに基づいて、指標を算出するエッジ信頼度取得手段と、
(F)前記帯域画像、前記等方向ボカシ画像、および前記異方性ボカシ画像とを画素の各々で前記指標に基づく重み付けにより重ね合わせて、前記元画像からノイズ成分が除去された処理画像を生成する処理画像生成手段とを備えることを特徴とする画像処理装置。 - 請求項1または請求項2に記載の画像処理装置において、
前記処理画像生成手段は、
処理対象の画素に対応するエッジ信頼度が高いほど、前記等方向ボカシ画像よりも前記異方性ボカシ画像が前記帯域画像から大きく減算されるように画像処理を行うことを特徴とする画像処理装置。 - 請求項1ないし請求項3のいずれかに記載の画像処理装置において、
処理対象の画素に対応するエッジ信頼度が低いほど、前記等方向ボカシ画像および前記異方性ボカシ画像が前記帯域画像からより減算されないように処理画像を行うことを特徴とする画像処理装置。 - 請求項1ないし請求項4のいずれかに記載の画像処理装置において、
処理対象の画素に対応するエッジ信頼度が中間的な値を示すときは、前記異方性ボカシ画像よりも前記等方向ボカシ画像が前記帯域画像から大きく減算されるように画像処理を行うことを特徴とする画像処理装置。 - 請求項1ないし請求項5のいずれかに記載の画像処理装置において、
前記処理画像生成手段は、前記帯域画像の周波数成分を基に重み付けの様式を変更することを特徴とする画像処理装置。 - 請求項1ないし請求項6のいずれかに記載の画像処理装置において、
前記処理画像生成手段は、前記元画像の撮影時の露光量を示す情報を基に重み付けの様式を変更することを特徴とする画像処理装置。 - 請求項1ないし請求項7のいずれかに記載の画像処理装置において、
前記等方向ボカシ手段は、前記元画像の撮影時の露光量を示す情報を基に前記等方向性のスムージングフィルタの形状や大きさを変更すること、または、
前記異方性ボカシ手段は、前記元画像の撮影時の露光量を示す情報を基に前記異方性のスムージングフィルタの形状や大きさを変更することを特徴とする画像処理装置。 - 請求項1ないし請求項8のいずれかに記載の画像処理装置を搭載した放射線撮影装置において、
放射線を照射する放射線源と、
前記放射線源を制御する放射線源制御手段と、
照射された放射線を検出して検出信号を出力する検出手段と、
前記検出手段が出力する検出信号を基に前記元画像を生成する画像生成手段とを備えることを特徴とする放射線撮影装置。
Priority Applications (5)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201280041022.1A CN103747736B (zh) | 2011-09-07 | 2012-08-21 | 图像处理装置以及具备该图像处理装置的放射线摄影装置 |
US14/343,356 US9183621B2 (en) | 2011-09-07 | 2012-08-21 | Image processing apparatus and radiographic apparatus having the same |
EP12830645.3A EP2745780B1 (en) | 2011-09-07 | 2012-08-21 | Image processing device and radiation imaging apparatus comprising same |
JP2013532416A JP5835333B2 (ja) | 2011-09-07 | 2012-08-21 | 画像処理装置およびそれを備えた放射線撮影装置 |
HRP20160199T HRP20160199T1 (hr) | 2011-09-07 | 2016-02-25 | Uređaj za obradu slike i aparat za slikovni prikaz radijacije koji sadrži taj uređaj |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2011-195186 | 2011-09-07 | ||
JP2011195186 | 2011-09-07 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2013035255A1 true WO2013035255A1 (ja) | 2013-03-14 |
Family
ID=47831732
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2012/005237 WO2013035255A1 (ja) | 2011-09-07 | 2012-08-21 | 画像処理装置およびそれを備えた放射線撮影装置 |
Country Status (6)
Country | Link |
---|---|
US (1) | US9183621B2 (ja) |
EP (1) | EP2745780B1 (ja) |
JP (1) | JP5835333B2 (ja) |
CN (1) | CN103747736B (ja) |
HR (1) | HRP20160199T1 (ja) |
WO (1) | WO2013035255A1 (ja) |
Families Citing this family (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2014024691A1 (ja) * | 2012-08-07 | 2014-02-13 | シャープ株式会社 | 画像処理装置、画像処理方法、画像処理プログラム及び画像表示装置 |
JP6169626B2 (ja) * | 2014-03-10 | 2017-07-26 | 富士フイルム株式会社 | 放射線画像処理装置、方法およびプログラム |
TWI524878B (zh) * | 2014-09-16 | 2016-03-11 | 國立臺灣大學 | 疾病檢測方法及穿戴式裝置 |
KR102245745B1 (ko) * | 2014-12-02 | 2021-04-28 | 삼성전자 주식회사 | 이미지 블러링 방법 및 장치 |
JP6594075B2 (ja) * | 2015-07-22 | 2019-10-23 | キヤノン株式会社 | 画像処理装置、撮像システム、画像処理方法 |
US10380084B2 (en) * | 2017-01-26 | 2019-08-13 | Sap Se | Data model view with complex calculation capability |
AU2018225834B2 (en) * | 2017-02-24 | 2022-12-01 | Sunnybrook Research Institute | Systems and methods for noise reduction in imaging |
JP7349345B2 (ja) * | 2019-12-23 | 2023-09-22 | 富士フイルムヘルスケア株式会社 | 画像診断支援装置、画像診断支援プログラム、及び、それを備えた医用画像取得装置 |
JP7436320B2 (ja) * | 2020-07-31 | 2024-02-21 | 富士フイルム株式会社 | 放射線画像処理装置、方法およびプログラム |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2002125153A (ja) * | 2000-10-17 | 2002-04-26 | Fuji Photo Film Co Ltd | ノイズ抑制処理装置並びに記録媒体 |
JP2002133410A (ja) | 2000-10-25 | 2002-05-10 | Fuji Photo Film Co Ltd | ノイズ抑制処理装置並びに記録媒体 |
JP2004242285A (ja) | 2003-01-14 | 2004-08-26 | Fuji Photo Film Co Ltd | ノイズ抑制処理方法および装置並びにプログラム |
JP2007050259A (ja) * | 2005-08-17 | 2007-03-01 | Siemens Ag | ボリュームデータ再構成後の断層撮影3d画像のフィルタリング方法 |
JP2007152106A (ja) * | 2005-12-06 | 2007-06-21 | Siemens Ag | 断層撮影における高コントラスト対象のコンピュータ支援による認識方法およびシステム |
JP4072491B2 (ja) | 2003-10-23 | 2008-04-09 | キヤノン株式会社 | 画像処理装置、画像処理方法、プログラム及びコンピュータ可読媒体 |
WO2009145076A1 (ja) * | 2008-05-28 | 2009-12-03 | 株式会社 日立メディコ | 画像処理装置、画像処理方法、及び画像処理プログラム |
Family Cites Families (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
RU2150146C1 (ru) * | 1998-09-03 | 2000-05-27 | Семенченко Михаил Григорьевич | Способ обработки изображения |
EP1526480A1 (en) * | 2000-10-17 | 2005-04-27 | Fuji Photo Film Co., Ltd | Apparatus for suppressing noise by adapting filter characteristics to input image signal based on characteristics of input image signal |
US7072525B1 (en) * | 2001-02-16 | 2006-07-04 | Yesvideo, Inc. | Adaptive filtering of visual image using auxiliary image information |
US7664326B2 (en) * | 2004-07-09 | 2010-02-16 | Aloka Co., Ltd | Method and apparatus of image processing to detect and enhance edges |
JP4517872B2 (ja) * | 2005-02-10 | 2010-08-04 | ソニー株式会社 | 画像処理装置、画像処理方法、画像処理方法のプログラム及び画像処理方法のプログラムを記録した記録媒体 |
WO2008122056A2 (en) * | 2007-04-02 | 2008-10-09 | Case Western Reserve University | Medical apparatus and method associated therewith |
US20100142790A1 (en) * | 2008-12-04 | 2010-06-10 | New Medical Co., Ltd. | Image processing method capable of enhancing contrast and reducing noise of digital image and image processing device using same |
DE102009019840A1 (de) * | 2009-05-04 | 2011-01-27 | Siemens Aktiengesellschaft | Kontrastverstärkung von CT-Bildern mittels eines Multibandfilters |
-
2012
- 2012-08-21 EP EP12830645.3A patent/EP2745780B1/en not_active Not-in-force
- 2012-08-21 WO PCT/JP2012/005237 patent/WO2013035255A1/ja active Application Filing
- 2012-08-21 US US14/343,356 patent/US9183621B2/en active Active
- 2012-08-21 JP JP2013532416A patent/JP5835333B2/ja active Active
- 2012-08-21 CN CN201280041022.1A patent/CN103747736B/zh active Active
-
2016
- 2016-02-25 HR HRP20160199T patent/HRP20160199T1/hr unknown
Patent Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2002125153A (ja) * | 2000-10-17 | 2002-04-26 | Fuji Photo Film Co Ltd | ノイズ抑制処理装置並びに記録媒体 |
JP4197392B2 (ja) | 2000-10-17 | 2008-12-17 | 富士フイルム株式会社 | ノイズ抑制処理装置並びに記録媒体 |
JP2002133410A (ja) | 2000-10-25 | 2002-05-10 | Fuji Photo Film Co Ltd | ノイズ抑制処理装置並びに記録媒体 |
JP2004242285A (ja) | 2003-01-14 | 2004-08-26 | Fuji Photo Film Co Ltd | ノイズ抑制処理方法および装置並びにプログラム |
JP4072491B2 (ja) | 2003-10-23 | 2008-04-09 | キヤノン株式会社 | 画像処理装置、画像処理方法、プログラム及びコンピュータ可読媒体 |
JP2007050259A (ja) * | 2005-08-17 | 2007-03-01 | Siemens Ag | ボリュームデータ再構成後の断層撮影3d画像のフィルタリング方法 |
JP2007152106A (ja) * | 2005-12-06 | 2007-06-21 | Siemens Ag | 断層撮影における高コントラスト対象のコンピュータ支援による認識方法およびシステム |
WO2009145076A1 (ja) * | 2008-05-28 | 2009-12-03 | 株式会社 日立メディコ | 画像処理装置、画像処理方法、及び画像処理プログラム |
Non-Patent Citations (1)
Title |
---|
See also references of EP2745780A4 |
Also Published As
Publication number | Publication date |
---|---|
JPWO2013035255A1 (ja) | 2015-03-23 |
EP2745780B1 (en) | 2015-12-09 |
US20140314333A1 (en) | 2014-10-23 |
CN103747736A (zh) | 2014-04-23 |
US9183621B2 (en) | 2015-11-10 |
EP2745780A1 (en) | 2014-06-25 |
HRP20160199T1 (hr) | 2016-03-25 |
CN103747736B (zh) | 2016-02-17 |
JP5835333B2 (ja) | 2015-12-24 |
EP2745780A4 (en) | 2014-12-17 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
JP5835333B2 (ja) | 画像処理装置およびそれを備えた放射線撮影装置 | |
US9311695B2 (en) | Image processing method and radiographic apparatus using the same | |
JP5641148B2 (ja) | 画像処理装置およびそれを備えた放射線撮影装置 | |
JP5839710B2 (ja) | 解析点設定装置および方法、並びに体動検出装置および方法 | |
US20110305382A1 (en) | Image processing method and radiographic apparatus using the same | |
WO2011074471A1 (ja) | X線画像診断装置、x線露出制御方法及びプログラム | |
JP5610474B2 (ja) | 画像処理装置およびプログラム並びに画像診断装置 | |
US20080075379A1 (en) | Image processing device and image processing method | |
JP7167564B2 (ja) | X線撮影装置およびx線撮影装置の作動方法 | |
JP2011239804A (ja) | X線検査装置 | |
JP2016131805A (ja) | X線画像診断装置およびx線画像を作成する方法 | |
JP7283325B2 (ja) | 画像処理装置 | |
JP7247719B2 (ja) | 放射線透視撮影装置 | |
JP6743975B2 (ja) | 断層像生成方法および放射線撮影装置 | |
JP2017169904A (ja) | X線撮影装置およびマーク表示方法 | |
JP2010172560A (ja) | 放射線画像撮影装置及び画像処理装置 | |
US20220230286A1 (en) | Imaging with asymmetrical contrast amplification | |
JP7115545B2 (ja) | 医用x線画像処理装置およびx線画像撮影装置 | |
JP6199118B2 (ja) | X線ct装置及び医用画像処理方法 | |
JP2016209121A (ja) | X線撮影装置、x線画像処理装置、x線画像処理方法、および、x線画像処理プログラム | |
JP2016036671A (ja) | 放射線撮影装置およびブレ算出方法 | |
JP2021018115A (ja) | X線位相撮像装置 | |
JP2012090872A (ja) | 放射線撮影装置 | |
JP2005034259A (ja) | X線診断装置 | |
JP2013111365A (ja) | 放射線撮影装置および画像処理方法 |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
WWE | Wipo information: entry into national phase |
Ref document number: 201280041022.1 Country of ref document: CN |
|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 12830645 Country of ref document: EP Kind code of ref document: A1 |
|
ENP | Entry into the national phase |
Ref document number: 2013532416 Country of ref document: JP Kind code of ref document: A |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2012830645 Country of ref document: EP |
|
WWE | Wipo information: entry into national phase |
Ref document number: 14343356 Country of ref document: US |