JPH09322020A - Image processor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To reduce the archifact by locally adjusting an amount of after-image to be added based on a motion of an object in the image processor where noise is reduced and an image is displayed by adding after-image to the image collected by a measurement section such as an image diagnostic device or the like intentionally. SOLUTION: A spatial filter circuit 9 to apply spatial frequency processing to a difference image from a subtractor 2 is provided to adjust an attenuation by a lookup memory 3 to a pre-stage of the lookup memory 3 as a motion correction means to set a prescribed attenuation in response to the difference as to each picture element of the difference image from the subtractor 2. Thus, the motion of the object and separation accuracy of noise are improved and archifact due to erroneous addition of after-image based on a motion detection error of the object with a low contrast is reduced.

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 section of an image diagnostic apparatus or the like, and more particularly, for an object. The present invention relates to an image processing apparatus that locally adjusts the amount of afterimage added based on motion.

[0002]

2. Description of the Related Art A conventional image processing apparatus of this type includes an image storage means for storing a processed image, and a processed image of a previous frame read from the image storage means and an image of a newly input current frame. , A difference extraction means for obtaining a difference for each pixel, a motion correction means for setting a predetermined attenuation rate according to the difference value for each pixel of the difference image obtained by the difference extraction means, and the motion correction The afterimage adding means for multiplying the difference image by the attenuation rate set by the means and the addition result and the processed image of the previous frame from the image storing means, and the image output from the afterimage adding means are displayed. And an image display means.

When such an image processing apparatus is used to display, for example, an X-ray fluoroscopic image on an X-ray fluoroscopic imaging apparatus, if there are many afterimages, X-ray quantum noise is reduced on the displayed image. On the other hand, a moving subject will be blurred. Therefore, it is necessary to prevent the blurring of the subject by detecting the movement of the subject and reducing the afterimage of the moving portion. Under normal usage conditions, the X-ray quantum noise can be suppressed to a certain level or lower, so conventionally, the difference between the previous frame image and the current frame image is taken for each pixel, and the absolute value of the difference is larger than a certain fixed value. At this time, this was regarded as motion and the afterimage added to the pixel was reduced. As a device related to such an image processing device, for example, there is a device described in Japanese Patent Application No. 4-345137.

[0004]

However, in such a conventional image processing apparatus, since a portion where the contrast is less than the predetermined level such as an edge portion of a subject is not regarded as a motion, an afterimage is added. There was a problem such as becoming an artifact. Such a problem will be described with reference to FIG. In the following description, for simplicity, afterimages are not added at all when the difference is at or above a certain level, and when the difference is below a certain level, afterimages are added so that the current frame image and the afterimage have the same ratio. Further, noise is omitted in FIG.

First, FIG. 8 (a) shows a case in which the position of a subject to be measured, for example, a blood vessel, moves from A to B between the previous frame and the current frame, and FIG.
8E shows a profile curve in which the position of the XX ′ cross section of the subject shown in FIG. 8A is the horizontal axis, the pixel value (density value) is the vertical axis, and the pixel value is plotted on the vertical axis. There is. When the subject moves as shown by the arrow A → B in FIG. 8A, the profile curves are shown in FIGS. 8B and 8C.
It changes as shown in. Here, FIG. 8B shows the processed image of the previous frame (hereinafter referred to as “previous image”), and FIG. 8C shows the image of the current frame (hereinafter referred to as “current image”). Then, when the difference between the current image of FIG. 8C and the previous image of FIG. 8B is taken, the profile curve shown in FIG. 8D is obtained. The difference value is a constant level ± T
If (T indicates a predetermined threshold value) or more, the current image is left as it is (range of M ₁ and M ₂ in FIG. 8D), and ± T
If the following is the case, the half of the current image and the half of the previous image (afterimage) are added and the processing result is shown in FIG.
(E). This is displayed on the screen of the image display means.

In this case, as can be seen from FIG. 8 (e),
The image is correctly displayed in the part with sufficient contrast in the center of the subject, but the afterimage c is added to the part with insufficient contrast in the subject edge, and the subject is displayed in the previous image in FIG. 8B. The artifact d is generated at the position where it was erased. Further, as described above, the afterimage c is added to the edge portion of the subject that is correctly displayed in the image this time, so that the contrast of the subject that is correctly displayed becomes lower than the original contrast, and the contrast of the image of the subject is reduced. There was a step and it was an artifact. In reality, noise is added to this, so that artifacts are frequently generated in the vicinity of the subject edge portion.

In order to reduce the above-mentioned artifacts, FIG.
The residual image c to be added may be reduced by lowering the constant level ± T shown in (d). FIG. 9A shows FIG.
Noise is superimposed on the difference image in (d). The subject to be measured has a linear or planar spread, and generally has a lower spatial frequency than noise. In the conventional example, the low-pass filter process is not performed on the difference image, and the noise component occupying the difference in FIG. 9A is large, so that the constant level ± T cannot be lowered. Therefore, the afterimage cannot be reduced,
I could not reduce the artifacts.

SUMMARY OF THE INVENTION Therefore, an object of the present invention is to provide an image processing apparatus which can cope with such a problem and locally adjust the amount of an afterimage to be added on the basis of the movement of an object to reduce the artifact. And

[0009]

In order to achieve the above object, an image processing apparatus according to the first aspect of the present invention includes an image storage unit for storing a processed image and a processed image of a previous frame read from the image storage unit. A difference extracting means for obtaining a difference for each pixel from the newly input image of the current frame, and a predetermined attenuation according to the difference value for each pixel of the difference image obtained by the difference extracting means. Motion compensation means for setting a rate, and afterimage adding means for multiplying the difference image by the attenuation rate set by the motion compensation means and then adding the multiplication result and the processed image of the previous frame from the image storage means, In an image processing apparatus comprising image display means for displaying an image output from the afterimage adding means, an image processing device is provided in front of the motion correcting means in order to adjust the attenuation rate by the motion correcting means. A difference image from the difference extraction means is provided with a spatial frequency processing means for applying a spatial frequency processing.

The image processing apparatus according to the second invention is
An image storage unit that stores the processed image, a difference extraction unit that obtains a difference for each pixel between the processed image of the previous frame read from the image storage unit and the image of the newly input current frame, and this difference The motion compensation means sets a predetermined attenuation rate according to the difference value for each pixel of the difference image obtained by the extraction means, and the difference image is multiplied by the attenuation rate set by the motion compensation means. An image processing apparatus comprising an afterimage adding means for adding the multiplication result and the processed image of the previous frame from the image storage means, and an image display means for displaying the image output from the afterimage adding means, Before the motion correction means, in order to adjust the attenuation rate by the motion correction means, the difference image from the difference extraction means is referred to not only the difference value of the corresponding pixel but also that of its peripheral pixels. The largest difference value of the absolute value is provided with a maximum value reference unit for the difference value of the corresponding pixel.

[0011]

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 first invention. This image processing apparatus displays an image by reducing noise by intentionally adding an afterimage to an image collected by a measurement unit such as an image diagnostic apparatus such as an X-ray fluoroscopic apparatus. As shown, the frame memory 1 and the subtractor 2
, Lookup memory 3, multiplier 4, adder 5
, Delay circuit 6, D / A converter 7, and television monitor 8
And a spatial filter circuit 9.

The frame memory 1 serves as an image storing means for storing an image processed by inputting an image from an external device, and sequentially stores the processed image in each frame. The external device is, for example, an X-ray fluoroscopic imaging device, and a subject (not shown) is irradiated with X-rays by the X-ray device, the fluoroscopic image is captured by the TV camera 10, and the obtained video signal is A / D converter. At 11, it is converted into digital luminance data and output. The subtractor 2 also extracts a difference for each pixel between the processed image of the previous frame read from the frame memory 1 and the image of the current frame newly input from the A / D converter 11. It is a means.

The look-up memory 3 serves as a motion correction means for setting a predetermined attenuation rate in accordance with the difference value for each pixel of the difference image obtained by the subtracter 2 and has a built-in inside thereof. As shown in FIG. 3, a table for determining the multiplication constant k in the multiplier 4 at the next stage according to the difference value is written in advance. In FIG. 3, the multiplication constant k is, for example, 0.
It is a table that changes in the range of 5 to 1.0. Then, this table is related to setting the above-mentioned attenuation rate and is related to the threshold value T shown in FIG.

The multiplier 4 multiplies the difference image from the subtractor 2 by the attenuation rate set in the look-up memory 3. The adder 5 adds the multiplication result of the multiplier 4 and the processed image of the previous frame read from the frame memory 1. Then, the multiplier 4 and the adder 5 multiply the difference image by the attenuation rate set in the lookup memory 3, and then add the multiplication result and the processed image of the previous frame from the frame memory 1. And an afterimage adding means for doing so. In addition,
The delay circuit 6 is for adjusting the timing of both data when adding the multiplication result of the multiplier 4 and the processed image of the previous frame from the frame memory 1 in the adder 5.

The D / A converter 7 receives the image data output from the adder 5 and converts it into a video signal. Furthermore, the TV monitor 8 is
The video signal output from the converter 7 is input and displayed as an image. The D / A converter 7 and the television monitor 8 constitute image display means for displaying the image output from the afterimage adding means. The processed image data output from the adder 5 is
It is input to the D / A converter 7 and stored in the frame memory 1 for the next frame processing.

Here, in the present invention, a spatial filter circuit 9 is provided in the preceding stage of the lookup memory 3. The spatial filter circuit 9 serves as spatial frequency processing means for performing spatial frequency processing on the difference image from the subtractor 2 in order to adjust the attenuation rate by the look-up memory 3 as the motion correction means. As shown in FIG. 2, it is composed of two line memories 12a and 12b, a pre-stage adder 13a, two latches 14a and 14b, and a post-stage adder 13b. In the spatial frequency processing by the spatial filter circuit 9, for example, a total of 9 pixels of vertical 3 pixels × horizontal 3 pixels centering on the pixel of interest are added and averaged for the input difference image.

Next, the operation of the image processing apparatus thus constructed will be described. First, in FIG. 1, a subject (not shown) is irradiated with X-rays by an X-ray device, a perspective image thereof is photographed by a television camera 10, and the obtained video signal is A
The / D converter 11 converts the digital brightness data. The image data output from the A / D converter 11 becomes the image of the current frame (current image f). The current image f is input to one port of the subtractor 2, and the processed image of the previous frame (previous image g) read from the frame memory 1 is input to the other port of the subtractor 2. Thereby, the subtractor 2 subtracts the previous image g from the current image f to generate a difference image (f−g).

The difference image (f-
The signal g) is input to the next spatial filter circuit 9 and is subjected to low-pass filtering. That is, in FIG. 2, the difference image (f−g) from the subtractor 2 is input to the first line memory 12a and the adder 13a at the preceding stage. Then, the adder 13a delays the difference image (f-g) by one line in the first line memory 12a and the difference image (f-g).
The value obtained by delaying one line at b and delaying a total of two lines is added. As a result, vertical 3 pixels are added. Next, the result of this addition is the result of the first latch 14a and the adder 13 in the subsequent stage.
Input to b. Then, this adder 13b is
The addition result obtained by adding the pixels, the addition result obtained by delaying the addition result by one pixel by the first latch 14a, and the addition result obtained by delaying the addition result by one pixel by the second latch 14b and delaying the total of two pixels are added. As a result, horizontal 3 pixels are added. As a result, a total of 9 pixels of vertical 3 pixels × horizontal 3 pixels are added.
Then, in order to obtain this as the addition average, the addition result must be divided by "9", but the horizontal axis of the table of the lookup memory 3 shown in FIG. If this is done, this spatial filter circuit 9
Then, only addition is required.

In this way, the spatial filter circuit 9 performs low-pass filter processing on the difference image (fg). As a result, as shown in FIG. 9A, the noise component occupying the difference was large when the conventional low-pass filter processing was not performed. However, according to the present invention, the low-pass filter processing is performed by the spatial filter circuit 9. By doing so, as shown in FIG. 9B, the noise component in the difference can be reduced. Therefore, the constant level ± T '(T' represents a predetermined threshold value) with respect to the difference value can be lowered. Since the spatial filter circuit 9 has a delay of one line as described above, the first line memory 1
2a, the image (f
-G) 'is created and output to the next multiplier 4.

The output of the spatial filter circuit 9 is input to the lookup memory 3. This lookup memory 3
In FIG. 3, a table for determining the multiplication coefficient k in the next multiplier 4 according to the difference value of the difference image (f−g) is written in advance as shown in FIG. 3, and the coefficient determined by this table is written in advance. k is input to one port of the next multiplier 4. As shown by the solid line L ', the table of the present invention has a steeper slope than the table of the conventional example shown by the broken line L, and is created so that the threshold value T'determined by this table becomes small. ing. At the same time, the difference image (f−g) obtained by the subtractor 2 is delayed by the first line memory 12a in the spatial filter circuit 9 to obtain the image (f−).
g) ′ is input to the other port of the multiplier 4. This delay is for compensating the delay of the calculation in the spatial filter circuit 9.

Next, the multiplier 4 multiplies the multiplication coefficient k determined by the lookup memory 3 by the image (f-g) 'and outputs the result. The output of this multiplier 4 is the next adder 5
Is input to one of the ports. At the same time, the previous image g read from the frame memory 1 is input to the other port of the adder 5 as an image g ′ delayed by the delay circuit 6. As a result, the adder 5 adds the above two inputs and generates the processed image g ₁ of the current frame. At this time, the delay circuit 6 delays the previous image g to obtain the image g '.
The reason is to compensate for the delay of the calculation in the spatial filter circuit 9.

Now, the processed image g ₁ of the current frame is expressed by the following formula (1). g ₁ = g ′ + k (f−g) ′ (1) where g ′ and (f−g) ′ are delayed in order to match the phase with g and (f−g), In the formula (1), “′” may be removed and the following may be expressed. g ₁ = g + k (f−g) (2) Then, a table as shown in FIG. 3 is written in the lookup memory 3 shown in FIG. 1, and in the area of the multiplication coefficient k = 1, In the equation (2), g ₁ = g + (f−g) = f, and the current image f becomes the processed image g ₁ of the current frame as it is. Further, if k = 0.5, the above equation (2)
Is g ₁ = g + 0.5 (f−g) = 0.5f + 0.5g, and the image obtained by adding 1/2 of the current image f and the processed image g of the previous frame, which is the afterimage component, is It becomes the processed image g ₁ . That is, when the difference between the current image f and the processed image g of the previous frame is large (when the subject seems to have moved), the multiplication coefficient k is 1, no afterimage is attached, and when the difference is small, , The afterimage increases as the multiplication coefficient k becomes 1 or less and the difference decreases.

FIG. 4 shows such a processing state as a profile curve similar to that shown in FIG. 4A to 4E are exactly the same as the conventional example shown in FIGS. 8A to 8E. In the present invention, by the low-pass filter processing by the spatial filter circuit 9 shown in FIG. 1, the noise component occupying the difference can be reduced as shown in FIG.
T '(T' is a predetermined threshold value) can be lowered.
The lowering of the threshold value T'is realized by the table as shown in FIG. 3 written in the look-up memory 3 shown in FIG. As a result, the constant level ± T 'for the difference value becomes smaller than the constant level ± T in the conventional example shown in FIG. 4D as shown in FIG. 4F, and the component added as an afterimage becomes smaller. . Therefore, FIG. 4 (g)
In the processing result shown in FIG. 4, the afterimage c ′ added to the portion where the contrast of the subject edge portion is not sufficient is small, and further the artifact d ′ generated at the position where the subject is displayed in the previous image of FIG. 4B is small. Become.

The processed image g ₁ of the current frame processed in this way is converted into a video signal by the D / A converter 7 shown in FIG. 1, and this video signal is sent to the television monitor 8 for image display. Together, it is stored in the frame memory 1 for processing the next frame. Thereafter, this operation is repeated.

FIG. 5 is a block diagram showing an embodiment of the image processing apparatus according to the second invention. This image processing device
In FIG. 1, a maximum value reference circuit 15 is provided in front of the lookup memory 3 instead of the spatial filter circuit 9. This maximum value reference circuit 15 refers not only to the difference image from the subtracter 2 but also to the difference value of the peripheral pixel thereof in order to adjust the attenuation rate by the look-up memory 3 as the motion correction means. Among them, the difference value having the largest absolute value is used as the difference value of the corresponding pixel. As shown in FIG. 6, the maximum value reference circuit 15 includes, for example, four line memories 16a, 16b, 16c, 16
d, the maximum value selector 17a in the previous stage, and the four latches 18
a, 18b, 18c, 18d and the maximum value selector 1 in the subsequent stage
7b and. Other configurations and operations are exactly the same as in the case of FIG.

Next, the operation of the maximum value reference circuit 15 will be described with reference to FIG. In the example of FIG. 6, the largest absolute difference value is selected from a total of 25 pixels, which are 5 pixels vertically × 5 pixels horizontally with the corresponding pixel as the center. The difference image (f−g) obtained by the subtracter 2 is input to the first line memory 16a and the maximum value selector 17a in the preceding stage. Then, the maximum value selector 17a delays the difference image (f-g) by one line in the first line memory 16a and the difference image (f-g) and the second line memory 16b delayed by 1 line and a total of 2 lines delayed, third line memory 16c delayed by 1 line and a total of 3 lines delayed, and further delayed by 1 line by a fourth line memory 16d. And a total of 4
The line-delayed one is input, and the difference value having the maximum absolute value is selected from the vertical 5 pixels.

Next, the selected maximum value is input to the first latch 18a and the maximum value selector 17b in the subsequent stage. Then, the maximum value selector 17b selects the maximum value selected from the vertical 5 pixels, the maximum value delayed by one pixel by the first latch 18a, and the maximum value selected by the second latch 18b by 1 pixel.
A pixel delayed by a total of 2 pixels, a third latch 18c delayed by 1 pixel and a total of 3 pixels, and a fourth latch 18d delayed by 1 pixel and a total of 4 pixels. The delayed value is input, and the difference value having the maximum absolute value is selected from the horizontal 5 pixels. As a result,
The largest absolute difference value is selected from a total of 25 pixels, which are 5 pixels vertically × 5 pixels horizontally. Since the maximum value reference circuit 15 causes a delay of 2 lines due to the calculation, the image (fg) 'obtained by delaying the difference image (fg) by 2 lines in the second line memory 16b is multiplied by the next multiplier. It outputs to 4.

As described above, the maximum value reference circuit 15 refers not only to the difference image (f-g) but also to the difference between the corresponding pixel and its peripheral pixels, and finds the maximum difference value of the absolute values of the difference pixel (fg) of the corresponding pixel. By performing the processing as the difference, if it is determined that the surrounding pixels have moved even if it is not determined that the corresponding pixel has moved, the corresponding pixel is also processed as moving. That is, since the same processing is performed on the peripheral portion of the moving portion as on the moving portion itself, it is possible to prevent an artifact generated at the edge portion of the subject. The profile curve shown in FIG. 7 represents this state only in the one-dimensional direction.
FIGS. 7A and 7B correspond to FIG. 4D.
Now, in FIG. 7A, the difference value of the pixel P ₁ itself is below a certain level ± T. However, by replacing the maximum difference value within the range around L of the pixel P ₁ (difference value of the pixel P ₂₎ as a difference value of the pixel P _1, the difference value becomes constant level ± T or more, of the Pixel P ₁ is treated as moving. By doing this, FIG. 7 (b)
Ranges M ₁ and M ₂ determined to move in the conventional example shown in FIG.
Since the processing is performed as if it moved with respect to the ranges M ₁ ′ and M ₂ ′ including the peripheral portion of
As a result of the processing, it is possible to obtain a processed image with almost no artifacts.

[0029]

Since the image processing apparatus according to the first aspect of the present invention is configured as described above, a predetermined attenuation rate is set for each pixel of the difference image obtained by the difference extraction means in accordance with the difference value. By providing the spatial frequency processing means for adding spatial frequency processing to the difference image from the difference extracting means in order to adjust the attenuation rate by the motion correcting means, the movement and noise of the subject are provided. It is possible to improve the accuracy of separation of the image and the image, and it is possible to reduce the artifact due to the erroneous afterimage addition based on the motion detection error with respect to the subject having low contrast.

Further, since the image processing apparatus according to the second aspect of the invention is configured as described above, a predetermined attenuation rate is set for each pixel of the difference image obtained by the difference extracting means in accordance with the difference value. In order to adjust the attenuation rate by the motion compensation means, the difference image of the difference image from the difference extraction means is referred to in the preceding stage of the motion compensation means to refer to the difference value of not only the corresponding pixel but also its peripheral pixels, and the absolute value thereof is included therein. By providing the maximum value reference means for setting the largest difference value of the difference value of the corresponding pixel, it is possible to reduce the afterimage also in the peripheral portion of the portion in which the movement of the subject is detected, as in the actually moved portion. Can be processed to eliminate artifacts due to erroneous afterimage addition due to motion detection error for a subject with low contrast.

[Brief description of drawings]

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

FIG. 2 is a block diagram showing an internal configuration of a spatial filter circuit.

FIG. 3 is a graph showing an example of a table written in a lookup memory.

FIG. 4 is an explanatory diagram showing a state in which artifacts are reduced by the operation of the first invention.

FIG. 5 is a block diagram showing an embodiment of an image processing apparatus according to the second invention.

FIG. 6 is a block diagram showing an internal configuration of a maximum value reference circuit.

FIG. 7 is an explanatory diagram showing a state in which artifacts are removed by the operation of the second invention.

FIG. 8 is an explanatory diagram showing a state in which an artifact is generated in the conventional example.

FIG. 9 is an explanatory diagram showing a state in which noise is superimposed on a difference image in the conventional example and the present invention.

[Explanation of symbols]

 1 frame memory 2 subtractor 3 lookup memory 4 multiplier 5 adder 6 delay circuit 7 D / A converter 8 TV monitor 9 spatial filter circuit 10 TV camera 11 A / D converter 15 maximum value reference circuit.

Claims (2)

[Claims] Claim: What is claimed is: 1. An image storage means for storing a processed image, and a difference extraction for obtaining a difference for each pixel between the processed image of the previous frame read from the image storage means and the newly input image of the current frame. Means, a motion correction means for setting a predetermined attenuation rate according to the difference value for each pixel of the difference image obtained by the difference extraction means, and the difference image for the attenuation rate set by the motion correction means. An image having an afterimage adding means for multiplying the result of the multiplication with the processed image of the previous frame from the image storage means, and an image display means for displaying the image output from the afterimage adding means. In the processing device, a spatial frequency processing means for adding spatial frequency processing to the difference image from the difference extracting means in order to adjust the attenuation rate by the motion correcting means is provided before the motion correcting means. The image processing apparatus characterized by a. 2. An image storage means for storing a processed image, and a difference extraction for obtaining a difference for each pixel between the processed image of the previous frame read from the image storage means and the newly input image of the current frame. Means, a motion correction means for setting a predetermined attenuation rate according to the difference value for each pixel of the difference image obtained by the difference extraction means, and the difference image for the attenuation rate set by the motion correction means. An image having an afterimage adding means for multiplying the result of the multiplication with the processed image of the previous frame from the image storage means, and an image display means for displaying the image output from the afterimage adding means. In the processing device, in front of the motion correction means, in order to adjust the attenuation rate by the motion correction means, not only the corresponding pixel but also the difference value of the peripheral pixel thereof is added to the difference image from the difference extraction means. The image processing apparatus characterized in that a maximum value reference unit for the difference value of the corresponding pixel greatest difference value of the absolute value in the irradiation perilla.