JP2001103376A - Image processing unit - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an image processing unit that can adjust an after-image to be added in response to a motion of an object and a luminance of a received image. SOLUTION: The image processing unit consists of a frame memory 3 that stores an image from a measurement section that receives, processes and collects the image, a sbutractor 4 that takes a difference by each pixel between a received image this time from the measurement section and a preceding output image read from the frame memory 3, an LUTs 5 consisting of a plurality of lookup tables that set a filter coefficient depending on the obtained difference image and the image received from the measurement this time and multiply the set filter coefficient with the difference image, an adder 6 that sums that result of multiplication and the preceding output image from the frame memory 3, and image display means (7, 8) that display the image outputted from the adder 6. Thus, the after-image to be added can be adjusted depending on the motion of the object (difference image) and the received image.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing apparatus for displaying an image by reducing noise by intentionally adding an afterimage to an image collected by a measuring unit such as an image diagnostic apparatus, and more particularly, to an image processing apparatus. The present invention relates to an image processing apparatus capable of adjusting an afterimage to be added according to a motion (difference image) and an input image.

[0002]

2. Description of the Related Art A conventional image processing apparatus of this type includes an image storage means for inputting and processing an image collected by a measuring unit, storing an image output from the measuring unit, and a current input image from the measuring unit. A motion detecting means for obtaining a difference for each pixel from the previous output image read from the storage means, a predetermined filter coefficient is set according to the obtained difference image, and the set filter coefficient is A motion correcting unit that multiplies the difference image, an afterimage adding unit that adds the result of the multiplication to the previous output image from the image storage unit, and an image display unit that displays an image output from the afterimage adding unit. It was made up.

When such an image processing apparatus is used to display an X-ray fluoroscopic image in an X-ray fluoroscopic apparatus, for example, if there are many afterimages, X-ray quantum noise and the like on the displayed image are reduced. Then, a moving subject (for example, a guide wire) is blurred. For this reason, it is necessary to prevent the motion blur of the subject by detecting the motion of the subject and reducing the afterimage of the moving part. In this case, under normal use conditions, the time change of the X-ray quantum noise and the like is approximately within a certain range. Therefore, conventionally, the difference between the current input image and the previous output image is calculated for each pixel. If the absolute value of the difference image is larger than a certain threshold value, it is determined that the subject at that position is moving, and the afterimage added to the pixel is reduced. A device related to such an image processing device is, for example, a device described in JP-A-6-169920.

[0004] The addition of an afterimage when an image is displayed in a conventional image processing apparatus will be described using mathematical expressions. First,
The mathematical expression for adding an afterimage to the current input image is f (t) for the current input image, g (t) for the current output image, g (t-1) for the previous output image, and f (t-1) for the current input image. t) and the previous output image g (t-
If the difference image from 1) is d, the filter coefficient is w, the initial value of the input image is f (0), and the initial value of the output image is g (0),
The following equations (1) to (3) are obtained. g (t) = w · d + g (t−1) (t: positive integer) (1) d = f (t) −g (t−1) (2) g (0) = f (0) … (3)

[0005] Next, filter coefficient characteristics in this case, the filter coefficient when the difference image d is 0 and w _0, the threshold value of the difference image d is T, when the residual image coefficient is a, the following equation (4 ) To (6). w = (1−w ₀ ) · | sin (πd / 2T) | ^a + w ₀ | d | <T (4) w = 1 | d | ≧ T (5) 0 <w ₀ <1 (6) )

In the above equations (1) to (6), the filter coefficient w increases as the difference image d increases, and becomes “1” when the value exceeds the threshold value T. Here, when the filter coefficient w is 1, no afterimage is added, and the current input image f (t) is used as the current output image g (t) as it is. The above w ₀ is the minimum value of the filter coefficient w, and the smaller the difference image d, the more the afterimage to be added.
The afterimage coefficient a is a parameter that changes the filter coefficient characteristic, and the afterimage coefficient a is large (for example, a = 1.
2) When the difference image d is small, the filter coefficient w is small. On the other hand, when the residual image coefficient a is small (for example, a = 0.45) and the difference image d is small, the filter coefficient w is large.

FIG. 8 shows the relationship between the afterimage coefficient a and the afterimage characteristic in this case. In FIG. 8, when the input image f (t) is a time series of luminance that changes stepwise in time, if the afterimage coefficient a is large (for example, a = 1.2), the output is performed as shown by a triangular dot line graph. The afterimage of the image g (t) increases. Also, the afterimage coefficient a is small (for example, a = 0.
45), the afterimage of the output image g (t) decreases as indicated by the line graph of the circular dots. Therefore, when the afterimage coefficient a is large, the motion blur of the output image increases, but the noise reduction effect increases. Conversely, when the afterimage coefficient a is small, the motion blur of the output image is reduced, but the noise reduction effect is reduced.

[0008]

However, in such a conventional image processing apparatus, the afterimage to be added is set separately from the input image. Is determined only by the difference image of the input image, while suppressing the motion blur of the guide wire or the like in the high-brightness part (for example, the non-bone part) of the input image, the low-brightness part (for example, the overlapping part of the spine and the guide wire) ), It was difficult to reduce the noise. In addition, in the conventional apparatus, the afterimage to be added cannot be changed according to the input image. Therefore, in the input image, the afterimage is added in a low luminance portion having a low SN ratio to improve the S / N ratio. In a high-brightness portion having a high ratio, motion blur could not be reduced by adding a small amount of afterimage.

Accordingly, an object of the present invention is to provide an image processing apparatus capable of adjusting such a problem and adjusting the motion of a subject (difference image) and an afterimage to be added according to an input image. I do.

[0010]

In order to achieve the above object, an image processing apparatus according to the present invention comprises: an image storage means for inputting image data collected by a measurement unit, storing an image output after processing, and Motion detecting means for taking a difference for each pixel between the current input image data from the measuring section and the previous output image data read from the image storage means; and the obtained difference image data and the measuring section A motion correction means for setting a filter coefficient according to the current input image data from the input image data and multiplying the difference image data by the set filter coefficient; and a result of the multiplication and the previous output image data from the image storage means. And an image display means for displaying the image output from the afterimage adding means.

In addition, the motion correction means may set a filter coefficient such that an afterimage to be added to a low luminance portion of an input image is larger than that of a high luminance portion.

[0012]

Embodiments of the present invention will be described below in detail with reference to the accompanying drawings. FIG. 1 is a block diagram showing an embodiment of an image processing apparatus according to the present invention. This image processing apparatus is for inputting an image collected by a measurement unit such as an image diagnostic apparatus such as an X-ray fluoroscope, and intentionally adding an afterimage to reduce noise and display an image. It comprises a frame memory 3, a subtractor 4, a look-up table LUTs 5, an adder 6, a D / A converter 7, and a television monitor 8.

First, a measuring unit such as an image diagnostic apparatus as an external device has a television camera 1 and an A / D converter 2. The television camera 1 irradiates a subject (not shown) with X-rays using an X-ray device to capture a perspective image of the subject. The A / D converter 2 receives the video signal obtained by the television camera 1, converts the video signal into digital image data, and outputs the digital image data.

The frame memory 3 serves as image storage means for storing an image collected by the measuring section, and storing an image output after processing. The frame memory 3 takes in an output image from an adder 6 described later and sequentially stores the image. It is designed to remember. The subtracter 4 is provided for each pixel between the current input image f (t) from the A / D converter 2 of the measuring unit and the previous output image g (t-1) read from the frame memory 3. It serves as a motion detecting means for obtaining a difference, and outputs a difference image represented by d = f (t) -g (t-1).

The LUTs 5 set a filter coefficient w in accordance with the difference image d obtained by the subtractor 4 and the current input image f (t) from the A / D converter 2 of the measuring section. This is a motion compensating means for multiplying the difference image d by the set filter coefficient w, and is a feature of the present invention. As shown in FIG. 2, it is determined by the input image f (t) and the difference image d. .., 9n in which a value w · d obtained by multiplying the obtained filter coefficient w by the difference image d is stored in advance.

The adder 6 serves as an afterimage adding means for adding the multiplication result w · d from the LUTs 5 and the previous output image g (t−1) from the frame memory 3 to add this afterimage. The result is the current output image g (t).
Note that the output image g (t) from the adder 6 is represented by D
/ A converter 7 and the next input image f (t)
Is stored in the frame memory 3 as a previous output image g (t-1) for adding an afterimage to the image.

The D / A converter 7 converts the digital image data of the output image g (t) from the adder 6 into a video signal. Further, the television monitor 8 receives the video signal output from the D / A converter 7 and displays it as an image. The D / A converter 7 and the television monitor 8 constitute image display means for displaying the image output from the adder 6.

Next, the operation of the thus configured image processing apparatus of the present invention will be described. In FIG. 1, first, an X-ray device irradiates a subject (not shown) with X-rays, a fluoroscopic image thereof is photographed by a television camera 1 of a measuring unit, and an obtained video signal is converted into a digital image by an A / D converter 2. Convert to data. The image data output from the A / D converter 2 becomes the current input image f (t).

Next, the current input image f (t) is input to one input terminal of the subtractor 4 and, at the same time, the other input terminal of the subtractor 4 Output image g (t-1) is input. Then, by the operation of the subtracter 4, the previous output image g (t-1) is subtracted from the current input image f (t), and a difference image d = f (t) -g (t
-1) is generated.

The difference image d output from the subtractor 4
Is input to one input terminal of the LUTs5, and the input image f (t) of the present time from the A / D converter 2 of the measurement unit is input to the other input terminal of the LUTs5. Then, by the operation of the LUTs 5, the filter coefficient w is set according to the difference image of each pixel of the difference image d and the luminance value of each pixel of the current input image f (t). The set filter coefficient w is multiplied by the difference image of the difference image d, and the value w · d is the value of L in the LUTs5 shown in FIG.
Stored as UTs 9a, 9b,..., 9n.

Thereafter, the LUTs 5 are obtained by multiplying the filter image w set in advance and the difference image d by referring to the corresponding LUT from the internal LUTs 9a to 9n in accordance with the input image f (t) of this time. Output the value w · d.

The value w · d output from the above LUTs5
Is input to one input terminal of the adder 6, and the other input terminal of the adder 6 is connected to the frame memory 3
From the previous output image g (t-1). By the operation of the adder 6, a current output image g (t) in which an afterimage is added to the current input image f (t) is generated.

The current output image g (t) output from the adder 6 is input to a D / A converter 7 to be converted into a video signal, and the converted video signal is transmitted to a television monitor 8. Input and image display. At the same time, the current output image g (t) output from the adder 6 is sent to the frame memory 3 for the next frame processing, and stored as the previous output image g (t-1). . Thereafter, this operation is repeated.

As another example of the LUTs 5, when setting the filter coefficient w, an afterimage to be added to the low luminance portion of the input image f (t) may be made larger than that of the high luminance portion. That is, instead of making the afterimage coefficient a in the above equations (4) to (6) constant with respect to the input image f (t),
For example, the afterimage coefficient a is increased (maximum value a _max ) in a low-luminance part (for example, an overlapping part of a spine and a guide wire) of the input image f (t), and the high-luminance part (for example, a non-bone region) Reduces the afterimage coefficient a (minimum value a _min ) and changes the afterimage coefficient a according to the input image f (t).

When this is represented by a mathematical expression, the following expressions (7) to (7)
It becomes like (9). a = a _max − (a _max −a _min ) · f (t) / f (t) _max (7) 0 ≦ f (t) ≦ f (t) _max (8) a _max = 1.2, a _min = 0.45 (9) When this is represented by a graph, the afterimage coefficient characteristic as shown in FIG. 3 is obtained.

Due to the after-image coefficient characteristic as shown in FIG. 3, the filter coefficient w in the area where the difference image d corresponding to the time change of the X-ray quantum noise or the like is small in the low luminance portion of the input image f (t) is reduced. (See FIG. 4), while on the other hand, in the high-luminance portion of the input image f (t), the filter coefficient w in the region where the difference image d is small can be increased (see FIG. 5).

4 and 5, the afterimage is increased in the low luminance part of the input image f (t) as shown by the line graph of the black dot in FIG. 6, while the input image f (t) is increased. In the high-luminance portion of (t), the afterimage can be reduced as shown by the line graph of the black dot in FIG. As described above, by using the afterimage coefficient characteristic that changes according to the input image f (t), it is possible to reduce the noise in the low luminance portion while suppressing the motion blur in the high luminance portion of the input image.

[0028]

The present invention has been configured as described above.
The motion correction means sets a filter coefficient according to the difference image obtained by the motion detection means and the current input image from the image measurement unit, and multiplies the difference image by the set filter coefficient. With this configuration, the afterimage to be added can be adjusted according to the movement of the subject and the input image. That is, since the difference image between the current input image and the previous output image and the residual image added according to the input image can be adjusted, the SN ratio of the input image can be controlled while suppressing motion blur in a high-brightness area. , The noise in the low-luminance part can be reduced.

[Brief description of the drawings]

FIG. 1 is a block diagram showing an embodiment of an image processing apparatus according to the present invention.

FIG. 2 is a block diagram illustrating an internal configuration of LUTs in the image processing apparatus.

FIG. 3 is a graph showing an afterimage coefficient characteristic capable of reducing noise in a low luminance portion while suppressing motion blur in a high luminance portion of an input image.

FIG. 4 is a graph showing filter coefficient characteristics in a low-luminance portion of an input image.

FIG. 5 is a graph showing filter coefficient characteristics in a high-luminance portion of an input image.

FIG. 6 is a graph showing an afterimage characteristic in a low luminance portion of an input image.

FIG. 7 is a graph showing an afterimage characteristic in a high luminance portion of an input image.

FIG. 8 is a graph showing a relationship between an afterimage coefficient and an afterimage characteristic in a conventional apparatus.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 ... TV camera 2 ... A / D converter (ADC) 3 ... Frame memory 4 ... Subtractor 5 ... LUTs (Look Up Tables) 6 ... Adder 7 ... D / A converter (DAC) 8 ... TV monitor 9a- 9n LUT (Look Up Table)

Claims (1)

[Claims] An image storage means for inputting and processing image data collected by a measurement unit and storing an image output therefrom, a current input image data from the measurement unit and a previous image data read from the image storage means. A motion detection unit that calculates a difference between each pixel with the output image data, and sets a filter coefficient according to the obtained difference image data and the current input image data from the measurement unit, and sets the filter coefficient. Motion correction means for multiplying the difference image data by a filter coefficient,
An image sticking means for adding the multiplication result and the previous output image data from the image storage means, and an image display means for displaying an image output from the image sticking means. Image processing device.