JP4595294B2 - Image processing device - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to an image processing apparatus used for a medical or industrial X-ray imaging apparatus, and more particularly to an image processing apparatus used for imaging an angiography or an organ such as a stomach.
[0002]
[Prior art]
For example, when the heart is to be imaged in a state where the movement of blood vessels is clearly understood, the lungs that become bright images because they hardly absorb X-rays and the backbones that become dark images because they absorb X-rays overlap. , The heart and blood vessels there will be blurred.
[0003]
In view of this, it has been proposed that the difference between the blurred image subjected to frequency processing and the original image is taken and the dynamic range of the image is compressed and processed so that the entire image is not darkened while suppressing the high luminance portion.
[0004]
As such a method, conventionally, the applied voltage of the cathode electrode of the image intensifier is switched to switch between the focus state and the defocus state of the image intensifier to obtain the original image and the blurred image, and the blurred image is thereby obtained. Create a blurred image by converting the image obtained by an imaging device such as an image intensifier or the like (see Patent Document 1) into a digital image and then performing template processing or fast Fourier transform. There was a way to do it.
[0005]
[Patent Document 1]
JP-A-2000-341589 [0006]
[Problems to be solved by the invention]
However, when image processing is realized on a moving image, a large amount of computation is required for frequency processing. Therefore, when the processing speed is increased, the range of frequency characteristics obtained becomes narrower. Conversely, the range of frequency characteristics obtained is reduced. There is a drawback that a trade-off problem occurs in which the processing speed and the range of the obtained frequency characteristics are limited such that the processing speed is slowed down when the width is increased.
[0007]
In the method of switching the voltage applied to the cathode electrode of the image intensifier, the original image and the blurred image cannot be collected at the same time. In addition, the frequency change range indicating the degree of blurring is also limited by the physical characteristics of being limited to the adjustment range of the applied voltage.
[0008]
In the case of a method for performing digital processing, an enormous number of arithmetic circuits are required to increase the processing speed, and in order to increase the processing speed while avoiding such an enormous number of arithmetic circuits, the frequency characteristics obtained can be reduced. A simple calculation method with a limited range has to be taken, and there is a drawback that the quality of the image is lowered.
[0009]
The present invention has been made in view of the above points, and the invention according to claim 1 can obtain a high-quality image with a small processing speed delay while widening the range of the obtained frequency characteristics. An object of the present invention is to make it possible to obtain an image with less influence of motion artifacts.
[0010]
[Means for Solving the Problems]
In order to achieve the above object, the present invention has the following configuration.
In other words, the image processing apparatus according to the first aspect of the present invention outputs a blurred image generation unit that generates a blurred image from an image of a past frame in which a subject is captured, an original image in which the subject is captured, and the blurred image generation unit. An image processing apparatus including a subtracting unit that subtracts the blurred image, wherein the blurred image generation unit includes a low frequency component discriminating unit that discriminates a low frequency component of the input image using a filter , A blurred image is generated from a plurality of past images obtained by discriminating low frequency components for the number of frames passed by the low frequency component discrimination means .
[0012]
According to the configuration of the image processing apparatus of the first aspect of the present invention, a low-frequency component can be repeatedly extracted to create a blurred image, so that a past image delayed in the time direction can be used. For example, template processing When processing a large area that requires a large amount of computation, it is possible to obtain a substantially larger area, that is, a low frequency by repeating processing with a small template size filter that can obtain a processing result quickly. The processing with a further extracted component can be performed, so that a high-quality image with a uniform degree of blur can be obtained with a small processing speed delay while widening the range of the obtained frequency characteristics.
[0013]
According to a third aspect of the present invention, in the image processing apparatus according to the first or second aspect , according to the amount of motion detected by the amount of motion detected by the motion amount detecting means for detecting the amount of motion of the subject, a coefficient converting means for extracting the impact reduction factor so that the influence of the image of the past frame as the amount of motion is large is reduced, configured to impart the impact reduction factor Bo Ke images.
[0014]
(Action / Effect)
According to the configuration of the image processing apparatus of the invention according to claim 3, when the motion amount is large, the impact reduction factor given to the ball Ke images with less influence of the past frame image, conversely, when the motion amount is small, it is possible to influence reduction factor given to the ball Ke image to those large influence of the past frame image. Therefore, for example, when it is desired to observe only a change in the image of a higher frequency component such as a catheter, the influence of the image of the past frame is reduced and the influence of the high frequency component such as a minute movement of another organ is reduced. A blurred image can be obtained satisfactorily and an image with little influence of motion artifacts can be obtained. In addition, when an image having periodic motion due to heartbeat or the like is targeted, it is possible to obtain a uniform blurred image in which low-frequency components are extracted by increasing the influence of the image of the past frame.
[0015]
DETAILED DESCRIPTION OF THE INVENTION
Next, an embodiment of the present invention will be described with reference to the drawings.
FIG. 1 is an overall configuration diagram showing an embodiment of an X-ray imaging apparatus using an image processing apparatus according to the present invention. X-rays radiate X-rays with an X-ray fluoroscopic table 2 on which a subject 1 is placed. A ray tube 3 and an image intensifier 4 for taking an X-ray image from the X-ray tube 3 that has passed through the subject 1 are provided, and an image processing device 5 is connected to the image intensifier 4 and an image An X-ray imaging apparatus is configured by connecting a monitor 6 to the processing apparatus 5.
[0016]
The image processing device 5 includes an A / D converter 7, a blurred image generation unit 8, a subtraction device 9, and a D / A converter 10, and converts an image obtained from the image intensifier 4 into a digital image. The original image is input to the blurred image generating unit 8 and the subtracting device 9, and the blurred image generated by the blurred image generating unit 8 and the original image are subtracted to balance the overall brightness without losing the contrast. It is configured to obtain an image from which the image is removed and display the image on the monitor 6.
[0017]
FIG. 2 is a circuit diagram showing a first embodiment of the image processing apparatus according to the present invention. The blurred image generation unit 8 includes first and second multipliers 11 and 12, a coefficient calculator 13, an adder 14, A template filter 15 as low frequency component discrimination means and a frame memory 16 as blur image storage means are provided.
[0018]
The first multiplier 11 multiplies the original image input from the A / D converter 7 by the addition coefficient (1 / K), and outputs an image obtained by the multiplication to the adder 14. ing. On the other hand, the second multiplier 12 multiplies the coefficient (1-1 / K) calculated by the coefficient calculator 13 by the blurred image of the past frame stored in the frame memory 16 and obtains the result of multiplication. The resulting image is output to the adder 14.
The adder 14 adds the images input from the first and second multipliers 11 and 12 to create a blurred image obtained by extracting the low frequency component, outputs the blurred image to the template filter 15, and outputs the blurred image. After the components are discriminated, they are output to the subtracting device 9.
The first and second multipliers 11 and 12, the coefficient calculator 13 and the adder 14 are referred to as a blurred image creating means.
[0019]
The template filter 15 uses the fact that the number of times of filtering differs as it becomes a past frame when integrating in the time direction, and superimposes an image with a higher number of times of filtering on the blurred image, thereby lowering the frequency. A frequency component region is extracted.
[0020]
That is, as shown in the graph of the relationship between the elapsed frame and the addition ratio in FIG. 3, by repeating the filtering by the template filter 15, the template size A (the number of pixels for which the average value is calculated for one pixel is simply calculated). (This is also true for the graph of FIG. 5.) Unlike the case where the number of pixels is increased, the number of pixels does not increase, so that the processing speed delay can be reduced and the influence of the image of the past frame is reduced. The template size can be substantially increased.
[0021]
FIG. 4 is a circuit diagram showing a second embodiment of the image processing apparatus according to the present invention, in which the blurred image generation unit 8 uses first, second, third and fourth template filters as low frequency component discriminating means. 21, 22, 23, 24, first, second and third frame memories 25, 26, 27 as blur image storage means, first, second and third adders 28, 29, 30, It is configured to use the principle of moving addition for processing in the time direction.
[0022]
According to the configuration of the second embodiment, as in the first embodiment, images stored and delayed in the first, second, and third frame memories 25, 26, and 27 are subjected to template processing with a larger template size. Therefore, as the delay period becomes longer, an image with a uniform blur degree in which lower frequency components are extracted is obtained.
[0023]
In the first, second, third, and fourth template filters 21, 22, 23, and 24, as shown in the graph of the relationship between the elapsed frame and the addition ratio in FIG. The first, second, and third frame memories 25, 26, and 27 are configured to be added to a past frame image stored and delayed at a uniform rate and a low rate component at a large rate. Yes. The degree of influence of the past image is only the number of the three first, second, and third frame memories 25, 26, and 27 provided, and does not affect after that.
The first, second, and third adders 28, 29, and 30 are used as a blurred image creating means.
[0024]
FIG. 6 is a circuit diagram showing a third embodiment of the image processing apparatus according to the present invention, in which the blurred image generation unit 8 uses first, second, third and fourth template filters as low frequency component discriminating means. 31, 32, 33, 34, first, second and third frame memories 35, 36, 37 as blur image storage means, first, second, third and fourth multipliers 38, 39, 40. , 41 and first, second and third adders 42, 43, 44.
[0025]
Addition coefficients a, b, c, and d are input to the first, second, third, and fourth multipliers 38, 39, 40, and 41, respectively, and the first, second, and third frame memories 35 are input. , 36, and 37, weighting is performed when the images of the past frames stored and delayed are added, and the influence degree of the images of the past frames can be controlled.
A blurred image is created with the first, second, third and fourth multipliers 38, 39, 40, 41 and the first, second and third adders 42, 43, 44. Called means.
[0026]
FIG. 7 is a circuit diagram showing a fourth embodiment of the image processing apparatus according to the present invention. The coefficient calculator 13 of the first embodiment and the input line of the original image from the A / D converter 7 are added. Connected to each other through a motion amount detecting means 53 and a coefficient converting means 54, which are composed of a device 51 and a frame memory 52.
[0027]
In the motion amount detection means 53, the adder 51 calculates the difference between the original image and the previous frame image stored and stored in the frame memory 52, and detects the motion amount of the subject based on the difference. It is configured.
[0028]
The coefficient conversion means 54 includes a motion detection value conversion section 55, a coefficient calculation section 56, and a motion amount-coefficient table 57, and inputs the motion amount detected by the motion amount detection means 53 to the coefficient calculation section 56. An addition coefficient corresponding to the motion amount is derived from the motion amount-coefficient table 57, and the derived addition coefficient (K) is assigned by the motion detection value conversion unit 55.
[0029]
In the motion amount-coefficient table 57, an influence reduction coefficient is set in advance so that the influence of the image of the past frame decreases as the motion amount increases. Other configurations are the same as those of the first embodiment, and the description thereof is omitted by giving the same reference numerals.
[0030]
According to the fourth embodiment, in accordance with the amount of movement of the subject, the influence reduction coefficient is extracted so that the influence of the image of the past frame decreases as the amount of movement increases. The effect reduction coefficient to be applied to the blurred image is made to have less influence. Conversely, when the amount of motion is small, the effect reduction coefficient to be applied to the original image and the blurred image may be made to have much influence. It can be done.
[0031]
As a result, for example, when it is desired to observe only a change in the image of a higher frequency component such as a catheter, the influence of the image of the past frame is reduced and the influence of the high frequency component such as a minute movement of another organ is reduced. A few blurred images can be obtained satisfactorily, and an image with little influence of motion artifacts can be obtained. In addition, when an image having periodic motion due to heartbeat or the like is targeted, it is possible to obtain a uniform blurred image in which low-frequency components are extracted by increasing the influence of the image of the past frame.
[0032]
FIG. 8 is a circuit diagram showing a fifth embodiment of the image processing apparatus according to the present invention. Each of the first, second, third and fourth multipliers 38, 39, 40 and 41 of the third embodiment is shown. The input line of the original image from the A / D converter 7 is connected to the motion amount detection means 53 comprising the adder 51 and the frame memory 52 and the coefficient conversion means 54 as in the fourth embodiment. It is connected.
[0033]
The coefficient conversion means 54 includes a motion detection value conversion section 55, a coefficient calculation section 56, and a motion amount-coefficient table 57, and inputs the motion amount detected by the motion amount detection means 53 to the coefficient calculation section 56. An addition coefficient corresponding to the motion amount is derived from the motion amount-coefficient table 57, and the derived addition coefficient (a, b, c, d) is provided by the motion detection value conversion unit 55.
[0034]
In the motion amount-coefficient table 57, an influence reduction coefficient is set in advance so that the influence of the image of the past frame decreases as the motion amount increases. In other words, when the amount of motion is large, even when the amount of motion is large, even when the addition coefficient one frame before is set to be larger than the sum of the previous addition coefficients, such as a>b>c> d, when the amount of motion is small. On the contrary, the addition coefficient of one frame before is set to be smaller than the previous addition coefficient such that a <b <c <d, and each addition coefficient (a , B, c, d) are set so that the influence of the image of the past frame decreases as the amount of motion increases, as described above. Other configurations are the same as those of the third embodiment, and the description thereof is omitted by giving the same reference numerals.
[0035]
In the above embodiment, the template filter 15 is used as the low frequency component discriminating means, but it may be replaced with the following filter.
(1) A moving average filter that calculates an average value of several points around each pixel and sets the average value as the value of the corresponding pixel is used.
[0036]
As shown in the explanatory diagram of the moving average filter in FIG. 9, for example, an average value of nine points of 3 × 3 around the corresponding pixel (indicated by ●) is used as a pixel value [see FIG. 9A. ].
When calculating this moving average, the total value of three points before moving [A + B + C, see FIG. 9 (b) for each column without always adding the values of 9 points around 3 × 3. ] Is subtracted from the value A of the pixel excluded after the movement, and the value D of the pixel added after the movement is added [A + B + C−A + D = B + C + D, see FIG. can do.
[0037]
(2) The values around each pixel are expanded in a two-dimensional direction, and the values of each pixel are calculated by filtering in the horizontal direction and the vertical direction with increasing weight as it approaches the corresponding pixel, An average value is calculated by adding the calculated values multiplied by each other, and a pyramid filter that uses the average value as the value of the corresponding pixel is used.
[0038]
As shown in the explanatory diagram of the pyramid filter in FIG. 10, for example, it is expanded in a two-dimensional direction of 5 × 5 25 points around the corresponding pixel (indicated by ●), and a larger coefficient is assigned to the apex side. The average value is set as the pixel value [see (a), (b), and (c) of FIG. 10].
When calculating the next pixel value, it is not necessary to always add the values of 25 points around 5 × 5. Depending on the direction of filtering, the total value of 5 points before movement in the variable part for each column [A + 2B + 3C + 2D + E] is subtracted from the pixel values [A + B + C, see FIGS. 10 (d) and (e)] that are excluded after movement, and the pixel values [D + E + F] that are added after movement are added. The total value [B + 2C + 3D + 2E + F] of the column can be obtained and easily calculated.
[0039]
In the above embodiment, the present invention is applied to a medical X-ray imaging apparatus. However, the image processing apparatus of the present invention can be applied to industrial and general imaging apparatuses.
[0040]
【The invention's effect】
As described above, according to the image processing apparatus of the first aspect of the present invention, a blurred image is obtained by performing low frequency component discrimination processing on the original image by the low frequency component discriminating means, and the original image is used. Using the blurred image and the original image, it is possible to create a blurred image from which a low frequency component is extracted.
Therefore, since an original image and a blurred image using the original image are simultaneously used and a low-frequency component is repeatedly extracted to create a blurred image, a past image delayed in the time direction can be used. When processing, it is possible to extract a substantially larger area, that is, a low-frequency component by repeating processing with a filter with a small template size without performing template processing for a large area that requires a large amount of computation. Thus, it is possible to obtain a high-quality image with a small degree of blurring and a uniform degree of blur while widening the range of frequency characteristics to be obtained.
[Brief description of the drawings]
FIG. 1 is an overall configuration diagram showing an embodiment of an X-ray imaging apparatus using an image processing apparatus according to the present invention.
FIG. 2 is a circuit diagram showing a first embodiment of the image processing apparatus according to the present invention;
FIG. 3 is a graph showing the relationship between the elapsed frame and the addition ratio in the first embodiment.
FIG. 4 is a circuit diagram showing a second embodiment of the image processing apparatus according to the present invention;
FIG. 5 is a graph showing the relationship between the elapsed frame and the addition ratio in the second embodiment.
FIG. 6 is a circuit diagram showing a third embodiment of the image processing apparatus according to the present invention;
FIG. 7 is a circuit diagram showing a fourth embodiment of the image processing apparatus according to the present invention;
FIG. 8 is a circuit diagram showing a fifth embodiment of the image processing apparatus according to the present invention;
FIG. 9 is an explanatory diagram of a moving average filter.
FIG. 10 is an explanatory diagram of a pyramid filter.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 1 ... Subject 5 ... Image processing apparatus 11, 12 ... 1st, 2nd multiplier (blurred image creation means)
13. Coefficient calculator (blurred image creation means)
14 ... Adder (Bokeh image creation means)
15 ... Template filter (low frequency component discrimination means)
16: Frame memory (blurred image storage means)
21, 22, 23, 24 ... 1st, 2nd, 3rd, 4th template filter (low frequency component discrimination means)
25, 26, 27... First, second and third frame memories (blurred image storage means)
28, 29, 30... First, second and third multipliers (blurred image creating means)
31, 32, 33, 34... First, second, third and fourth template filters (low frequency component discriminating means)
35, 36, 37 ... first, second and third frame memories (blurred image storage means)
38, 39, 40, 41... First, second, third and fourth multipliers (blurred image creating means)
42, 43, 44... First, second and third adders (blurred image creating means)
53. Movement amount detecting means 54 ... Coefficient converting means

Claims (3)

A blur image generating unit that generates a blurred image from an image of a past frame in which the subject is captured; and a subtracting unit that performs a subtraction process on the original image in which the subject is captured and the blurred image output from the blur image generating unit. In the image processing apparatus, the blurred image generation unit includes low-frequency component discrimination means for discriminating low-frequency components of an input image by using a filter , and the low-frequency component is reduced by the number of frames passed by the low-frequency component discrimination means. An image processing apparatus that generates a blurred image from a plurality of past images obtained by discriminating components .   The image processing apparatus according to claim 1, wherein the filter is a template filter. 3. The image processing apparatus according to claim 1, wherein a motion amount detection unit that detects a motion amount of a subject and a motion amount detected by the motion amount detection unit increase the amount of motion of a past frame as the motion amount increases. effect and a coefficient converting means for extracting the impact reduction coefficient so that less is intended to impart impact reduction factor Bo Ke image image processing apparatus.

Images