JP2002336223A - Image processor, image processing system, image processing method, program, and record medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an image processor, etc., capable of precisely extracting a pixel value of a boundary between a prescribed region and the other regions in a target picture. SOLUTION: An extraction means 120 extracts the pixel value of the boundary pixel value between the prescribed region and the other regions of the target picture by approximating the prescribed region (region of a direct line part, or a part where a radiation is directly incident on an image sensor without intervening a subject) in a histogram of the target picture (radiation picture, etc.), by a polygonal line comprising two straight lines.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a medical image, for example, which is used to obtain a radiation image of a subject (an image of the amount of radiation transmitted through the subject) obtained by radiography such as X-ray photography. The present invention relates to an image processing apparatus, an image processing system, an image processing method, a computer-readable storage medium storing a program, and a program used in an apparatus or system to be processed.

[0002]

2. Description of the Related Art Conventionally, for example, when an X-ray is irradiated onto a subject, an X-ray transmission distribution through the subject is observed to observe the inside of the subject, particularly the inside of a human body. Have been. In recent years, a large-format image sensor using a solid-state image sensor called a “flat panel X-ray sensor” has been used to distribute an X-ray transmitted through a subject (hereinafter, an image of this is referred to as an “X-ray image”). Acquiring is becoming commonplace.

One of the advantages of the solid-state imaging device is that a plurality of image receiving pixels existing on a sensor plane can directly spatially sample a plane energy distribution (X-ray transmission distribution) and signal it. Can be

On the other hand, a drawback of the solid-state imaging device is that a plurality of pixel elements for spatially sampling an X-ray transmission distribution are basically independent elements, and have different characteristics. . For this reason, in order to obtain an appropriate image (X-ray image) as if it were output from an ideal solid-state imaging device composed of pixels having uniform characteristics, it is necessary to correct variations in characteristics of each pixel. Image processing is required.

[0005] For example, when the pixel element is an energy conversion element having linear characteristics, the main characteristic variations include variations in conversion efficiency (gain) and offset. Therefore, when an X-ray image of a subject is obtained by an image sensor using a solid-state imaging device, it is first necessary to correct a gain and an offset.

As a method of correcting offset variation (hereinafter, also simply referred to as “offset correction”), a unique offset value for each pixel obtained without irradiating the image sensor with X-ray energy is obtained, and the offset value is obtained. A method of subtracting a value from image information including a subject image obtained by making an X-ray passing through a subject incident on an image sensor is exemplified. In this method, it is also possible to set the offset value acquisition time (energy accumulation time in the image sensor, etc.) in accordance with the actual X-ray image acquisition time.

As a method of correcting the gain variation (hereinafter, also simply referred to as “gain correction”), a so-called “white image” is obtained by irradiating X-ray energy to an image sensor without a subject. After performing the offset variation correction, the gain correction for each pixel using the white image is performed by division. In practice, the gain correction is often realized by the difference between the pixel values between the log-transformed images.

[0008]

The feature of the above-described X-ray image acquisition by the image sensor is that the X-ray dynamic range received by the image sensor is very wide. This is to adjust the X-ray dose applied to the subject so that the dynamic range of the X-ray transmission amount distribution through the subject matches the dynamic range of the image sensor in order to render information inside the subject more accurately. Then, the amount of X-rays in a portion that does not pass through the subject (hereinafter, also referred to as “absent portion”) becomes extremely large and reaches the image sensor directly.

In an image sensor, that is, an image sensor using a solid-state image pickup device, when an extremely strong X-ray is directly incident, its output value is saturated. Such saturation occurs when the incident X-ray dose exceeds the allowable charge amount of the pixel on the X-ray image sensor, when the subsequent electric system amplifier is saturated, and in the A / D conversion system that digitizes the electric amount. In some cases, however, the output value obtained is fixed to a substantially constant value.

When the offset correction or the gain correction is performed in the above-described saturated state, the correction data obtained in the non-saturated state is used as the data used in each correction, so that the correction for each pixel to be corrected is performed. Conversely, the fluctuation is superimposed on the image (for example, a blank portion), and as a result, noise occurs.

In order to solve the above problem, for example, in Japanese Patent Application Laid-Open No. 2000-244824, a gain control amplifier is used for any X-ray dose.
A configuration has been proposed in which the maximum value is adjusted to be a signal value that can always be handled.

However, in the above-described configuration, the adjustment is performed so that even a direct line portion having no information originally (a portion where a direct X-ray is directly incident) is accurately imaged. The contrast resolution of the part is greatly reduced. In particular, in the case of an X-ray image, for example, in order to improve the diagnostic ability of the image, the dynamic range of the image information may be compressed for display output or print output in order to clarify the details of the X-ray image. In such a case, the direct line portion with much noise is emphasized and output.

In order to solve the above problem, for example, Japanese Patent Application Laid-Open No. Hei 6-292013 proposes a configuration in which the degree of enhancement of a direct line portion is reduced when the dynamic range of image information is compressed. . However, in this proposal, means for accurately extracting a direct line portion is not clarified, and erroneous recognition of a pixel value forming the direct line portion causes pressure on even the contrast in an X-ray image for image diagnosis. Maybe.

For example, Japanese Patent Application Laid-Open No. 5-328357 proposes a configuration in which the dynamic range of an X-ray image is determined by analyzing the peak value of a histogram obtained from a target X-ray image. Have been. However, in this configuration, since only the peak value of the histogram of the X-ray image is analyzed, a large error may occur in accurately grasping the pixel value of the direct line portion.

The present invention has been made in order to eliminate the above-mentioned drawbacks, and an image processing apparatus and an image processing method capable of accurately extracting a pixel value at a boundary between a predetermined area and another area in a target image. It is an object to provide a processing system, an image processing method, a computer-readable storage medium storing a program, and the program.

Alternatively, according to the present invention, a pixel value of a predetermined region such as a direct line portion (a blank portion) in a target radiation image is accurately recognized, and the pixel value of the predetermined region is substantially fixed to a predetermined value. Therefore, an object of the present invention is to provide an image processing apparatus, an image processing system, an image processing method, a computer-readable storage medium storing a program, and a program capable of providing a good radiation image.

[0017]

For such a purpose,
An image processing apparatus according to the present invention includes a histogram creating unit that creates a pixel value histogram of a target image, and approximating a predetermined range of the histogram created by the histogram creating unit with a polygonal line formed by two straight lines. Extracting means for extracting a boundary pixel value between a predetermined area of the target image and another area.

Further, in the image processing method according to the present invention, the predetermined range of the histogram created in the step of creating the pixel value histogram of the target image and the step of creating the histogram is defined by a polygonal line composed of two straight lines. And extracting a boundary pixel value between the predetermined area of the target image and other areas by approximating the target image.

Further, a program according to the present invention is a program for causing a computer to function as a predetermined means, wherein the predetermined means comprises: a histogram creation means for creating a pixel value histogram of a target image; Extracting means for extracting a boundary pixel value between a predetermined area of the target image and other areas by approximating a predetermined range of the histogram created by the means with a polygonal line composed of two straight lines. Features.

Further, a computer readable storage medium according to the present invention is characterized by recording the program according to claim 21.

[0021]

Embodiments of the present invention will be described below with reference to the drawings.

[First Embodiment] The present invention is applied to, for example, an X-ray imaging apparatus 100 as shown in FIG. The X-ray imaging apparatus 100 according to the present embodiment particularly calculates a start pixel value (minimum pixel value) of a direct line portion (a transparent portion where the X-ray directly enters the image sensor without passing through the subject) in the X-ray image. It is configured to accurately and stably extract from the histogram information of the X-ray image. Further, the X-ray imaging apparatus 100 reduces the pixel value of the extracted direct line portion to a substantially constant value by a pixel value conversion process in order to suppress pixel value fluctuation (noise due to image correction, etc.) in the direct line portion. It is configured to be fixed to. Hereinafter, the X-ray imaging apparatus 100 of the present embodiment will be specifically described.

<Overall Configuration and Operation of X-Ray Imaging Apparatus 100> As shown in FIG. 1, the X-ray imaging apparatus 100 includes an X-ray generator 101, a bed 103, and an X-ray image sensor 10.
5, A / D converter 106, memory (MEM) 107, switch 108, memories 109 and 110, difference block (SUB) 111, memory (MEM) 112, reference table (look-up table: LookUpTable)
e, hereinafter also referred to as “LUT”) 113, switch 11
4, memories 115 and 116, difference block (SUB) 1
17, a memory (MEM) 118, a histogram creation unit (HST) 119, an operation block (ANAL) 120,
And an LUT 121.

The X-ray generator 101 emits X-rays to the subject 102 on the bed 103 under the control of a control device (not shown) that generates a high voltage (see the arrow in the figure). As the subject 102, a human body is used as an example, and the subject 102 is lying on a bed 103.

The X-ray image sensor 105 detects the object 10
It converts the intensity distribution of the X-ray dose transmitted through 2 into an electric signal, and includes a large-sized solid-state imaging device.
Specifically, for example, an X-ray image sensor (hereinafter, also referred to as a “flat panel sensor”) 105 uses a plurality of pixels arranged in a matrix in a plane to spatially express the X-ray transmission distribution on a two-dimensional plane. Sampling. When the subject 102 is a human body, the sampling pitch is usually set to about 100 μm to 200 μm.

The flat panel sensor 105 as described above
Converts a charge value, which is present in each pixel and is proportional to the X-ray dose, into an electric quantity (image signal), which is a voltage or a current, under the control of a controller (not shown). The image signals obtained by the flat panel sensor 105 are sequentially scanned and output to an A / D converter 106 at the subsequent stage.

The A / D converter 106 digitizes the image signal output from the flat panel sensor 105 and outputs it. The memory 107 temporarily stores the output of the A / D converter 106 as image data.

The switch 108 reads image data from the memory 107 and selectively stores the read image data in one of the two memories 109 and 110.
As a result, the A / D converter 106 is supplied from the flat panel sensor 105 to the memory 109 in a state where X-rays are not exposed.
And image data output via the memory 107 are stored as offset fixed pattern image data. In addition, the memory 110 stores an A / A signal from the flat panel sensor 105 in a state where X-rays are actually irradiated (actual imaging state).
Image data output via the D converter 106 and the memory 107 is stored as captured image data.

Specifically, for example, by using an X-ray dose measuring device (not shown) for monitoring the amount of X-ray transmitted through the subject 102 as an X-ray emission control sensor called a “photo timer”, The X-ray irradiation in the X-ray generation unit 101 is stopped at the moment when the integrated value (charge amount) of the X-ray dose emitted from the X-ray 101 reaches a predetermined value. At the same time,
By scanning the flat panel sensor 105, image data of the subject 102 is stored in the memory 107. Then, the switch 108 switches the output destination to the terminal A side, so that the image data in the memory 107 is output to the memory 110 as captured image data.

Immediately after that, the X-ray generator 101 does not perform X-ray irradiation, and the flat panel sensor 1 does not emit X-rays until the same accumulation time as the irradiation time obtained by the photo timer is reached.
05 is driven. When the charge accumulation in the flat panel sensor 105 is completed, the flat panel sensor 105 is scanned, and the image data without X-ray exposure is stored in the memory 107. Then, the switch 108 switches the output destination to the terminal B side so that the image data in the memory 107 is stored in the memory 10 as fixed offset pattern image data.
9 is output.

The difference block 111 exists at a corresponding position in the pixel data constituting the fixed offset pattern image data in the memory 109 for each of the pixel data constituting the photographed image data in the memory 110. A process of subtracting pixel data is performed, and the result is stored in the memory 1.
12 is stored.

The LUT 113 is a lookup table for logarithmic value conversion used to perform gain correction by subtraction (described later). The memory 112 obtained as described above
Are output via the LUT 113 (image data including the subject image). The switch 114 is
By switching the output destination to the terminal C side, the image data output from the LUT 113 is output to the memory 115.

The memory 116 stores image data obtained when an operation called "calibration" is performed in the X-ray imaging apparatus 100.

In the above-described calibration, first, image data is obtained by the above-described photographing operation in a state where the subject 102 is not present. At switch 114,
By switching the output destination to the terminal D side, the image data output from the LUT 113 is output to the memory 116. Therefore, the image data in the memory 116 is
It consists of the X-ray dose distribution itself without the subject 102 and data of only the gain variation for each pixel.

Normally, the calibration operation is performed at the start of operation about once a day, and the data (image data in the memory 116) obtained by this operation makes it possible to obtain the gain variation that the flat panel sensor 105 has for each pixel. Can be corrected.

That is, the difference block 117 converts the image data (original image data) in the memory 115 into the memory 1
By subtracting the image data (gain image data) in 16, the variation in gain for each pixel in the original image is corrected, and the corrected image data is stored in the memory 118.

The histogram creating section 119 stores the data in the memory 11
The image data (corrected image data) stored in 8 is read out, and a histogram of the image is created. The operation block 120 analyzes the histogram obtained by the histogram creation unit 119 and determines the heel point (HEELPOIN).
T) is detected.

For example, the operation block 120 executes processing in accordance with the flowcharts of FIGS. 2 and 3 whose details will be described later. In particular, in the processing shown in FIG. 3, the starting pixel value (heel value) of the direct line portion is obtained from the histogram of the processing target image, that is, the histogram including the direct line portion (the portion where the X-rays are directly incident on the flat panel sensor 105). When selecting a base point k1 to be used for extracting the point (step S302), the highest value of the image data of the subject 102 is determined based on a ratio empirically obtained from the highest pixel value or a specific pixel value. Is selected as the base point k1. Also, the end point k2 with respect to the base point k1
Is selected as the highest value of the image data. Since the selection of the base point k1 and the end point k2 does not have a significant effect on the heel point search (extraction) in the subsequent processing step, an appropriate heel point can be obtained.

The LUT 121 is, for example, a lookup table as shown in FIG. 4, which is generated based on the heel points (the histogram heel points P of the direct line portion) obtained in the operation block 120. In FIG. 4, the horizontal axis indicates the input pixel value, and the vertical axis indicates the output pixel value. "40" in FIG.
1 ”, in the LUT 121, when the input pixel value is smaller than the pixel value p (the value of the heel point P),
The input pixel value itself is output as an output pixel value. If the input pixel value is greater than the output pixel value, a fixed value p is output as an output pixel value. As a result, the density fluctuation of the direct line portion having a large fluctuation is suppressed, and is fixed at the fixed value p. The image data output from the LUT 121 is displayed on a display unit (not shown) for image diagnosis, for example.

<Characteristic Configuration of X-Ray Imaging Apparatus 100>
The most characteristic configuration of the present embodiment is the operation block 1
20, the direct line portion is accurately and stably extracted from the target image, the pixel value fluctuation (noise due to image correction, etc.) of the direct line portion is suppressed, and the pixel value of the direct line portion The value is fixed at a substantially constant value. In particular, when extracting the direct line portion, the operation block 120 accurately and stably extracts a pixel value (start pixel value) at which the direct line portion starts. Hereinafter, the above configuration will be specifically described.

First, FIG. 5 shows an image (target image) 500 as an example of an image to be processed (a captured image of the subject 102 stored in the memory 118), and a target image 500 created by the histogram creating section 119. 3 shows a histogram 510 of FIG.

In FIG. 5, "501" indicates a portion not irradiated with X-rays and a histogram portion thereof (hereinafter, also referred to as "histogram 501H").
02 ”indicates a subject portion and a histogram portion thereof (hereinafter, also referred to as“ histogram 502H ”), and“ 5.
03 ″ is a direct line portion outside the subject and its histogram portion (hereinafter, also referred to as “histogram 503H”)
Is shown.

The general feature of the direct ray portion 503 is that the X-ray intensity is extremely strong because the subject (here, “human body”) does not significantly reduce the X-ray intensity.
The histogram 5 is located at a position far apart from the pixel value
03H exists. Therefore, although the direct line portion 503 generally seems to be easily separated,
It is very difficult to separate pixel information of a boundary portion between the subject portion (human body portion) 502 and the direct line portion 503, that is, the pixel information of a portion such as skin or ear.

The pixel value of the upward arrow indicated by “520” in FIG. 5 is the pixel value at the boundary between the object portion 502 and the direct line portion 503 (hereinafter, also referred to as “boundary value 520”).
If the pixel values above the boundary value 520 are fixed to a constant value,
Pixel value fluctuations of the direct line portion 503, that is, noise are eliminated, and operations such as dynamic range compression can be tolerated.

However, if the boundary value 520 is slightly changed, the parts such as the skin and the ear in the subject part 502 will not be drawn, or the noise of the direct line part 503 will remain. Also, the boundary value 52
0 indicates that if the histogram 510 is graphed and displayed appropriately, it is relatively stable heuristically by human judgment. However, the boundary value 5
It is relatively difficult to automatically and stably determine 20 by a computer.

Therefore, in the present embodiment, the operation block 120 calculates the boundary values 520,
That is, the start pixel value of the direct line portion 503 is accurately and stably extracted by a simple method.

FIG. 6 shows the histogram 5 shown in FIG.
10, the histogram 502H of the subject portion 502
And the vicinity of the boundary between the direct line portion 503 and the histogram 503H. Here, in the histogram 502H of the subject portion 502, the frequency of a small amount of pixel values such as skin decreases linearly, and the direct line portion 503
In the histogram 503H, modeling is performed such that the frequency increases linearly. The arrow indicated by “530” in FIG. 6 indicates a point at which the two are optimally separated.

In order to obtain an optimal separation point (optimal separation point) 530 based on the above modeling, as shown in FIG. 7, the boundary between the object portion 502 and the direct line portion 503 is optimally approximated by a polygonal line. This break point P is extracted as the optimum separation point 530, that is, the heel point (HEELPOINT). Thereby, an optimal and stable heel point can be extracted.

The analysis of the histogram is a method that has been performed for a long time, but the analysis of the peak and the concentration is mainly performed.
With the configuration of the present embodiment as described above, there is no method for extracting a heel point.

FIG. 8 is a diagram for explaining the approximation using the broken line shown in FIG. In FIG. 8, points indicated by black circles are 0 to n of the target image (digital image).
It shows n histogram points up to -1, and is bent at the position of the break point P as shown in FIG. This broken line, that is, a straight line 530L of 0 to P, and P to (n−
The two straight lines with the straight line 530R in 1) are represented by “x” on the horizontal axis and “y” on the vertical axis.

[0051]

(Equation 1)

This is expressed by the following equation (1). In the above equation (1), the value of “P” represents the index of the break point P.

In the above equation (1), the parameters a0,
a1 and b0 are parameters obtained linearly algebraically in accordance with the least squares rule when the break point P is fixed. At this time, the parameter a0 indicates the slope of the straight line 530L (left straight line), and the parameter a1 indicates the slope of the straight line 530R (right straight line).

Assuming that the n histogram points shown in FIG. 8 are [(xi, yi); i = 0 to n−1], the sum of square errors ε is

[0055]

(Equation 2)

This is expressed by the following equation (2).

The sum of the square errors ε expressed by the above equation (2)
In order to minimize the partial differential of each of the parameters a0, a1, and b0 to "0".

[0058]

(Equation 3)

A simultaneous equation represented by the following equation (3) is solved.

Here, in the above-described arithmetic processing, the break point P is treated as known, but actually, the break point P is also an unknown number. Is changed, and a break point P at which the calculation result (sum of square errors ε) is minimized is obtained.

FIG. 2 shows the above arithmetic processing. In the above equation (1), P, a0, a1, b0
To p0, A0, A1, B0, which are the results of the processing of FIG.
Represents the broken line to be obtained.

First, an initial value "1" is set for each of the polygonal line indexes P and p0 (step S2).
01), the arithmetic processing of the above equation (3) is executed with this value (step S202). The parameters a0 and a1 thus obtained are set as parameters A0 and A1 (step S203).

Next, using the current index P and the parameters a0, a1, b0, the arithmetic processing of the above equation (2) is executed (step S204). Next, "2" is set for the index P (step S205), and it is determined whether or not the value of the index P exceeds (n-1) (step S206).

As a result of the determination in step S206, "P> n
−1 ”, that is, when the processing for all of the n histogram points is completed, the present processing is terminated.

As a result of the determination in step S206, "P> n
If it is not −1 ”, the arithmetic processing of the above equation (3) is executed using the current index P and the parameters a0, a1, b0 (step S207), and thereafter, the above equation (2)
(Step S208).

Then, the result (sum of the square errors) of the arithmetic processing of the above equation (2) is defined as “E”, and the resulting value E
Is smaller than the previous processing result E0 (step S209).

As a result of the determination in step S209, "E <E
If "0", the respective values of the current E, P, a0, a1, and b0 are set to E0, p0, A0, A1, and B0 (step S210). Thereafter, the index P is counted up (step S211), and again the step S206.
Execute the process from.

On the other hand, as a result of the determination in step S209,
If “E <E0” is not satisfied, the process of step S211 is executed as it is, and then the process from step S206 is executed again.

Further, in the present embodiment, in the histogram (histogram data series) including the direct line portion obtained by the histogram creating unit 119, the above-described polygonal line approximation (fitting) is performed while selecting the partial data. And the selected partial data is

[0070]

(Equation 4)

When the partial data satisfying the conditional expressions (1) and (2), the heel point P is extracted. Thereby, the heel point P can be stably extracted.

In the conditional expressions (1) and (2), for example, the value of “T” is a numerical value of about 50%,
As the value of “δ ₀ ”, a numerical value of about 10% can be applied. In this case, for example, when the rate of change of the inclination of the two straight lines forming the polygonal line is 50% or more and the ratio of the error resulting from the polygonal line fitting is 10% or less, a heel point is extracted.

The reason for extracting the heel point P as described above is that if the histogram data sequence used for extracting the heel point P is not properly selected, it is the point at which the accurate direct line portion starts. This is because the heel point cannot be determined.

A specific description will be made with reference to FIG. FIG. 9
, “553” indicates a histogram including a direct line portion. First, in the histogram 553, a point 545 that is clearly not a direct line portion is obtained as a base point.

The target data range is defined from the base point 545 to the point 5
Assuming that the range extends from point 52, point 550 and point 548 to point 546, the optimal polygonal line as shown in FIG.
This is indicated by "547" in FIG. However, the polygonal line 547 does not satisfy any of the conditions of the conditional expressions (1) and (2).

Further, base points 545, 552, 550,
Using the data series of points 548 and 548, “5” in FIG.
49 "is obtained. However, this broken line 549 may satisfy the above conditional expression (1).
It cannot be said that satisfactory fitting represented by the above condition (2) has been performed.

When the data series of the base point 545 and the point 552 is used, the fitting is sufficient, so that the above-mentioned conditional expression (2) is satisfied, but the above-mentioned condition (1) of the inverted L-shape is not satisfied. .

On the other hand, base points 545, 552, and 55
When the data series of 0 is used, a polygonal line indicated by “551” in FIG. 9 is obtained. This polygonal line 551 satisfies both the conditional expressions (1) and (2). Therefore, the broken point of the broken line 551 obtained here can be obtained as the heel point of the histogram.

FIG. 3 shows a flowchart of the heel point extraction processing as described above.

First, the histogram creating section 119 creates a histogram of the target image (step S301).
The operation block 120 determines the next step S302 based on the histogram obtained by the histogram creation unit 119.
, The heel point P is extracted.

That is, first, the operation block 120
In the histogram obtained by the histogram creating unit 119, a point that sufficiently includes a direct line portion is set as a base point k1 (step S302). This base point k1 is obtained by a method such as empirical or simple analysis of a histogram.

Next, the operation block 120 selects the end point k2 in the histogram (step S30).
3). As the end point k2, the point of the maximum pixel value in the histogram can be used.

Next, the arithmetic block 120 performs the fitting of the polygonal line using the data series in the range from the base point k1 to the end point k2 by the processing shown in FIG. 2 to obtain the polygonal point P (step S304).

Then, the operation block 120 determines whether or not the data series satisfies the conditional expressions (1) and (2) at the same time based on the fitting result in step S304 (step S305). ).

If the result of the determination in step S305 is that the above conditional expressions (1) and (2) are not satisfied, the operation block 1
20 updates the end point k2, that is, updates the data sequence in order to perform fitting with a smaller data sequence (step S306). Here, the data series is reduced by subtracting the value indicated by “α” from the value (index) of the end point k2. At this time, the value of “α” may be, for example, “1”, but an appropriate numerical value of “10” is sufficient to increase the computation efficiency.

Thereafter, the operation block 120 determines whether or not the value of the end point k2 after the update in step S306 is smaller than the base point k1 (step S307).

As a result of the determination in step S307, “k2 <
If “k1”, that is, if the value of the end point k2 is not a valid value, the operation block 120 determines that there is no heel point, that is, because there is no direct line portion having a sufficiently large area in the current data sequence, the heel point As a result, this processing ends.

As a result of the determination in step S307,
If “k2 <k1” is not satisfied, the operation block 120 executes the processing from step S304 again.

On the other hand, if the result of the determination in step S305 is that the above conditional expressions (1) and (2) are satisfied, the operation block 120 sets the current break point P as the heel point P (step S309), and ends this processing. And

[Second Embodiment] In this embodiment, in the X-ray imaging apparatus 100 of FIG.
Is, for example, a reference table as shown in FIG. That is, in the present embodiment, as shown in FIG. 4, instead of bending at an angle at the heel point P, an output pixel value to be clipped (fixed value p) and an output pixel value not to be clipped (input value In order to continuously connect the pixel value (pixel value equal to the pixel value), as shown in FIG.
LUT using the following function (y = ax ³ + bx ² + cx + d)
121 are configured.

In FIG. 10, “p0” is the heel point P, but may be a value that is a predetermined value around the heel point P. “Q” is a value at which the output is clipped, and an arbitrary value can be applied, but the value is the same as the heel point P or a value larger by a predetermined value. Also, “δ” is
This defines the range in which the LUT 121 is curved.

The above cubic function in the present embodiment is obtained under the condition that the value and the differential value at each node are the same.

[0093]

(Equation 5)

The solution of the simultaneous equations of equation (4) is obtained.

The solution of the simultaneous equation of the above equation (4) can be specifically calculated, and each of the parameters a to d is

[0096]

(Equation 6)

This is expressed by the following equation (5).

As described above, in the present embodiment, the LUT
Since 121 is constituted by a continuous and differentiable curve, a stable processed image can be provided.

[Third Embodiment] In the second embodiment, the LUT 121 has a continuous and differentiable curve. However, in the LUT 121 in this case, noise information of the subject portion and the direct line portion may be mixed in the curved (non-linear) portion. In the present embodiment, the above-described noise information is reduced, and more natural clipping is realized.

For this reason, as shown in FIG. 11, the X-ray imaging apparatus 100 according to the present embodiment is configured by further providing an operation block 122 at the subsequent stage of the LUT 121 in addition to the configuration shown in FIG. You. Arithmetic block 122
Executes, for example, the processing shown in the flowchart of FIG.

That is, first, the operation block 122
The address A indicating the pixel position on the target image is initialized (step S601). Next, the operation block 122 calculates
The value of the pixel indicated by the address A is obtained from the output of the LUT 121 (Step S602).

Next, the operation block 122 executes step S
The pixel value (target pixel value) acquired in 602 is the same as that in FIG.
It is determined whether or not there is a curve portion (non-linear portion) of the LUT 121 shown in (2), that is, within the range of (p0−δ) to p0 (step S603).

If the result of determination in step S603 is that the target pixel value exists in the curved portion, the operation block 122
Replaces the output value with the average value of nine pixels including the target pixel and its surrounding pixels (step S604),
The process proceeds to the next step S605.

If the result of determination in step S603 is that the target pixel value does not exist in the curved portion, the operation block 122
Proceeds to the next step S605.

In the step S605, the operation block 12
2 determines whether or not the processing of steps S602 to S604 has been completed for all pixels constituting the target image (step S605). As a result of this determination,
This process is terminated only when all the processes are completed.

If the result of the determination in step S605 is that the processing has not been completed, the operation block 122 increments the address A in order to execute the processing for the next pixel (step S606), and repeats the processing from step S602 again. Execute.

As described above, in the present embodiment, the LUT
When the output pixel value exists in the curve portion 121, the average of the pixel values around the target pixel is configured, so that the noise information of the remaining direct line portion can be reduced, and more natural clipping can be performed. realizable.

[Fourth Embodiment] In this embodiment, the configuration of the X-ray imaging apparatus 100 shown in FIG. 11 is, for example, a configuration further provided with an LUT 123 as shown in FIG. You.

The LUT 123 is configured to receive the output of the difference block 117 and output it to the memory 118, and is a reference table for obtaining a display image particularly suitable for image diagnosis. Therefore, the above-described process of removing noise in the direct line portion is performed after the LUT 123 performs image conversion for display on the image data output from the difference block 117.

The pixel value output from the LUT 123 depends on the display unit (not shown). However, the configuration according to the present embodiment is effective when the output image from the X-ray imaging apparatus 100 is already converted for display.

In the first to fourth embodiments, the direct line portion, that is, the portion where X-rays are directly incident on the image sensor (the transparent portion) is set as the target region, but the target region is limited to this. For example, it may be a background portion of a general radiation image (a region outside a radiation irradiation field).
In this case, the result of detecting the pixel value range of the background portion by the above-described heel point detection is effective not only for the creation of the LUT but also for the image processing for extracting or deleting only the background portion.

Further, after fixing (clipping) the direct line portion as described above, image processing including enhancement processing such as dynamic range compression may be executed. This makes it possible to provide a stable processed image without emphasizing correction noise.

Further, an object of the present invention is to supply a storage medium storing program codes of software for realizing the functions of the first to fourth embodiments to a system or an apparatus, and to supply the storage medium to the system or the apparatus. Computer (CP
U or MPU) can also be achieved by reading and executing the program code stored in the storage medium. In this case, the program code itself read from the storage medium realizes the functions of the first to fourth embodiments, and the storage medium storing the program code constitutes the present invention. As a storage medium for supplying the program code, RO
M, a flexible disk, a hard disk, an optical disk, a magneto-optical disk, a CD-ROM, a CD-R, a magnetic tape, a nonvolatile memory card, and the like can be used. The functions of the first to fourth embodiments are realized by executing the program code read by the computer, and the OS running on the computer based on the instruction of the program code. Perform part or all of the actual processing,
It goes without saying that the present invention is also configured when the functions of the fourth embodiment are realized. Further, after the program code read from the storage medium is written into a memory provided in an extended function board inserted into the computer or an extended function unit connected to the computer, the function of the program is performed based on the instruction of the program code. The present invention may also be configured when a CPU or the like provided in an expansion board, a function expansion unit, or the like performs part or all of actual processing, and the processing realizes the functions of the first to fourth embodiments. Needless to say.

FIG. 14 shows the functions 70 of the computer.
0 shows an example of the configuration. The computer function 700 includes, as shown in FIG.
ROM 702, RAM 703, and keyboard (KB)
709 keyboard controller (KBC) 705;
A CRT controller (CRTC) 706 of a CRT display (CRT) 710 as a display unit, a hard disk (HD) 711, and a flexible disk (FD)
A disk controller (DKC) 712 and a network interface card (NIC) 708 connected to an arbitrary network 720 are communicably connected to each other via a system bus 704.

The CPU 701 has the ROM 702 or the H
By executing the software stored in the D711 or the software supplied from the FD712, the respective components connected to the system bus 704 are comprehensively controlled.
That is, the CPU 701 stores a processing program according to a predetermined processing sequence in the ROM 702 or the HD 71.
1 or by reading from the FD 712 and executing the control, control for realizing the operations in the first to fourth embodiments is performed.

The RAM 703 functions as a main memory or a work area for the CPU 701. KBC705
Controls an instruction input from the KB 709 or a pointing device (not shown). CRTC 706 is
The display of the RT 710 is controlled. The DKC 707 controls access to the HD 711 or the FD 712 that stores a boot program, various application software, an editing file, a user file, a network management program, and a processing program according to the first to fourth embodiments. The NIC 708 is connected to the network 720
Data is exchanged bidirectionally with the above devices or systems.

As described above, in the present embodiment, a histogram of an object image (radiation image or the like) is created, and a predetermined area (a part where the radiation directly enters the image sensor without passing through the subject) in the histogram is created. The starting point (boundary point) of the area of the line portion) is determined by fitting with a polygonal line or the like.

Thus, for example, the predetermined area can be accurately extracted based on the start point of the predetermined area.
In addition, since a predetermined region in the target image can be accurately extracted, noise information due to image correction processing or the like relating to fluctuations in characteristics of each pixel of the image sensor can be appropriately suppressed, and therefore, good processing can be achieved. A post-image can be provided. For example, when the pixel value of a predetermined area such as a direct ray portion or a background portion (outside irradiation field area or the like) in a radiographic image is substantially fixed to a predetermined value, noise in the predetermined area is removed or suppressed. can do.

[0119]

As described above, according to the present invention, an image processing apparatus, an image processing system, an image processing method, and an image processing method capable of accurately extracting a pixel value at a boundary between a predetermined area and another area in a target image, A computer-readable storage medium storing a program, and the program can be provided.

Alternatively, according to the present invention, the pixel value of a predetermined area such as a direct line portion (absent portion) in a target radiation image is accurately recognized, and the pixel value of the predetermined area is substantially fixed to a predetermined value. By doing so, an image processing apparatus, an image processing system, an image processing method,
A computer-readable storage medium storing a program, and the program can be provided.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of an X-ray imaging apparatus to which the present invention is applied in a first embodiment.

FIG. 2 is a flowchart for explaining a polygonal-line fitting process executed in a calculation block 120 of the X-ray imaging apparatus.

FIG. 3 is a flowchart for explaining a heel point detection process including the above-described broken line fitting process.

FIG. 4 is a diagram illustrating a configuration of an LUT used in the X-ray imaging apparatus.

FIG. 5 is a diagram for explaining an example of an image to be processed in the X-ray imaging apparatus and a histogram thereof;

FIG. 6 is a diagram for describing a process of modeling a histogram of a direct line portion in the histogram.

FIG. 7 is a diagram for specifically explaining the polygonal line fitting.

FIG. 8 is a diagram for specifically explaining a calculation process in the polygonal line fitting.

FIG. 9 is a diagram for specifically explaining the heel point detection process.

FIG. 10 is a diagram illustrating a configuration of an LUT used in an X-ray imaging apparatus according to a second embodiment.

FIG. 11 is a block diagram illustrating a configuration of an X-ray imaging apparatus according to a third embodiment.

FIG. 12 is a flowchart illustrating a process of a calculation block 122 of the X-ray imaging apparatus.

FIG. 13 is a block diagram illustrating a configuration of an X-ray imaging apparatus according to a fourth embodiment.

FIG. 14 is a block diagram illustrating a configuration of a computer that can execute a program related to a function or a process of the first to fourth embodiments.

[Explanation of symbols]

 119 Histogram creation unit (HST) 120 Operation block (ANAL)

Claims (22)

[Claims] A histogram creating means for creating a pixel value histogram of the target image; and a predetermined range of the histogram created by the histogram creating means being approximated by a polygonal line composed of two straight lines. An image processing apparatus comprising: an extraction unit that extracts a boundary pixel value between a predetermined area and another area. 2. The image processing apparatus according to claim 1, wherein the target image includes a radiation image of a subject, and the predetermined area includes a radiation missing area or an irradiation field area. 3. The image processing apparatus according to claim 1, wherein the extraction unit extracts a broken point of the broken line obtained by the approximation as the boundary pixel value. 4. An image processing apparatus according to claim 1, wherein said extraction means includes a setting means for setting said predetermined range. 5. The image processing apparatus according to claim 1, wherein said extracting means includes a changing means for changing said predetermined range. 6. The image processing apparatus according to claim 1, wherein the extraction unit extracts the boundary pixel value when the polygonal line obtained by the approximation satisfies a predetermined condition. 7. The predetermined condition is that a ratio of an error as a result of fitting by a polygonal line composed of two straight lines is equal to or less than a predetermined value, and a change rate of a slope of the two straight lines is equal to or more than a predetermined value. 7. The image processing apparatus according to claim 6, wherein the condition is included. 8. The image processing apparatus according to claim 6, wherein the predetermined condition includes a condition regarding at least one of the degree of approximation of the polygonal line and the inclination of the two straight lines. 9. The predetermined condition regarding the degree of approximation is as follows:
9. The method according to claim 8, further comprising a condition relating to a degree of variation of the data of the histogram with respect to the polygonal line.
The image processing apparatus according to any one of the preceding claims. 10. The predetermined condition regarding the inclination is as follows:
9. The image processing apparatus according to claim 8, further comprising a condition relating to a change in the inclination of the two straight lines. 11. The image processing apparatus according to claim 1, further comprising a conversion unit configured to convert a pixel value of the predetermined area into a predetermined value. 12. The image processing apparatus according to claim 11, wherein said conversion means converts the pixel value according to a continuous and differentiable conversion function. 13. The conversion function converts a pixel value of a predetermined area to a substantially predetermined value and a pixel value of another area to a substantially linear value, and further includes a boundary pixel between the predetermined area and the other area. 13. The image processing apparatus according to claim 12, further comprising a function for nonlinearly converting a pixel value near the value. 14. The apparatus according to claim 13, further comprising: a smoothing unit configured to smooth a value of a pixel nonlinearly converted by the conversion function in accordance with a value of the pixel and pixels around the pixel. Image processing device. 15. The image processing apparatus according to claim 1, further comprising a display conversion unit that performs a pixel value conversion on the target image depending on a display unit before the processing by the extraction unit. 16. An image processing apparatus according to claim 11, further comprising an emphasis processing means for performing a predetermined emphasis processing on the image after the processing by said conversion means. 17. The image processing apparatus according to claim 16, wherein said predetermined enhancement processing includes processing for changing a dynamic range of an image. 18. The image processing apparatus according to claim 1, further comprising a recognition unit that recognizes the predetermined area based on the boundary pixel value. 19. An image processing system in which a plurality of devices are communicably connected to each other, wherein at least one of the plurality of devices is the image processing device according to claim 1. An image processing system having the function of: 20. A step of creating a pixel value histogram of the target image, and approximating a predetermined range of the histogram created in the step of creating the histogram with a polygonal line composed of two straight lines. Extracting a boundary pixel value between a predetermined region and another region. 21. A program for causing a computer to function as predetermined means, said predetermined means comprising: a histogram creating means for creating a pixel value histogram of a target image; and said histogram created by said histogram creating means. A program extracting means for extracting a boundary pixel value between a predetermined area of the target image and another area by approximating the predetermined range by a polygonal line formed by two straight lines. 22. A computer-readable storage medium on which the program according to claim 21 is recorded.