JP3932489B2 - Image processing method and apparatus - Google Patents

Description

  The present invention relates to an image processing method and apparatus, and more particularly to an image processing method and apparatus for selectively enhancing only a specific image portion such as an abnormal shadow in an image.

  2. Description of the Related Art Conventionally, image processing such as gradation processing and frequency processing is performed on image signals representing images obtained by various image acquisition methods to improve image observation and interpretation performance. Particularly in the field of medical images such as radiographic images of human subjects, it is necessary for specialists such as doctors to accurately diagnose the presence or absence of a patient's disease or injury based on the obtained images. Image processing to improve the interpretation performance of the image has become indispensable.

Among these image processing, as so-called frequency enhancement processing, as shown in, for example, Japanese Patent Laid-Open No. 61-169971, an image signal (referred to as an original image signal) Sorg such as a density value of an original image,

What is converted into an image signal Sproc is known. Here, β is a frequency enhancement coefficient, and Sus is an unsharp mask (so-called blur mask) signal. The blur mask signal Sus sets a mask composed of a pixel matrix of N columns × N rows (N is an odd number) having the original image signal Sorg as a central pixel for pixels arranged in two dimensions, that is, a blur mask.

It is an ultra-low spatial frequency component required as

  Since the value (Sorg−Sus) in the second term parenthesis of Expression (13) is obtained by subtracting the blur mask signal, which is an ultra-low spatial frequency component, from the original image signal, the ultra-low space of the original image signal. A frequency component higher than the ultra-low spatial frequency from which the frequency component is removed can be selectively extracted. The relatively high frequency component can be emphasized by multiplying the relatively high frequency component by the frequency enhancement coefficient β and adding the original image signal.

  On the other hand, processing based on a morphology (Morphology) algorithm (hereinafter referred to as morphological operation or morphological processing) that selectively extracts only specific image parts such as abnormal shadows from an image is known. ing. This morphological process has been studied as an effective technique for detecting microcalcifications, which is a characteristic feature especially in breast cancer, but the target image is not limited to microcalcifications in such mammograms. However, what is known in advance for the size and shape of a specific image portion (abnormal shadow or the like) to be detected can be applied to any image.

  Hereinafter, an outline of the morphological process will be described by using an example in which this morphological process is applied to detection of a microcalcification image in a mammogram.

(Basic morphological operations)
The morphological processing is generally developed as a set theory in an N-dimensional space, but will be described for a two-dimensional gray image for intuitive understanding.

  The grayscale image is regarded as a space in which a point at coordinates (x, y) has a height corresponding to the density value f (x, y). Here, the density value f (x, y) is a signal having a high luminance and high signal level that becomes a larger value as the luminance is higher, such as a signal to be displayed on the CRT.

First, for the sake of simplicity, a one-dimensional function f (x) corresponding to the cross section is considered. The structural element g used for the morphological operation is a symmetric function symmetric about the origin, as shown in the following equation (15).

It is assumed that the value is 0 in the domain and the domain G is the following formula (16).

At this time, the basic form of the morphological operation is a very simple operation as shown in the equations (17) to (20).

  That is, the dilation process is a process of searching for the maximum value in the width of ± m (value determined according to the structural element B) with the target pixel as the center (FIG. 12A). On the other hand, the erosion process is a process of searching for the minimum value in the width of ± m centered on the target pixel (see FIG. 12B). The opening (or closing) process corresponds to searching for the maximum value (or minimum value) after searching for the minimum value (or maximum value). That is, the opening process corresponds to smoothing the density curve f (x) from the low luminance side and removing a convex density fluctuation portion (a portion having higher luminance than the surrounding portion) smaller than the mask size 2 m ( (See FIG. 12C). On the other hand, the closing process is equivalent to smoothing the density curve f (x) from the high luminance side and removing a concave density fluctuation portion (a portion whose luminance is lower than the surrounding portion) smaller than the mask size 2 m ( (See FIG. 12D).

  When the structural element g is not symmetric with respect to the origin, the dilation operation shown in the equation (17) is called the Minkowski sum, and the erosion operation shown in the equation (18) is called the Minkowski difference.

  Here, in the case of a signal having a high density and high signal level in which the density value f (x) is a higher value as the density is higher, such as a signal for recording on a negative film, the relationship between luminance and density. Therefore, the dilation processing in the high density high signal level signal coincides with the erosion processing in the high luminance high signal level (FIG. 12B), and the erosion processing in the high density high signal level signal is high. This is consistent with dilation processing (FIG. 12 (A)) at high luminance and high signal levels, and the opening processing at high concentration and high signal levels is consistent with closing processing (FIG. 12 (D)) at high luminance and high signal levels. The closing process for a signal having a high density and a high signal level coincides with the opening process (FIG. 12C) for a high luminance and high signal level.

  In this section, a case of an image signal (brightness value) having a high brightness and a high signal level will be described.

(Application to calcified shadow detection)
For the detection of the calcified shadow, a difference method in which a smoothed image is removed from the original image can be considered. Since it is difficult to distinguish between calcified shadows and elongated non-calcified shadows (such as mammary glands, blood vessels, and mammary support tissues) with a simple smoothing method, Obata et al. At Tokyo University of Agriculture and Technology used multiple structural elements. A morphological filter represented by the following formula (21) based on the opening operation is proposed ("Extraction of microcalcifications by morphological filter using multiple structural elements" IEICE Transactions D-II Vol. J75-D-II No.7 P1170-1176 July 1992 etc.).

  Here, Bi (i = 1, 2,..., M) is, for example, four linear structural elements (in this case, M = 4) shown in FIG. Hereinafter, it is simply referred to as a structural element including the case of i = 1). If the structural element Bi is set to be larger than the calcified shadow to be detected, the calcified shadow, which is a convex signal change portion finer than the structural element Bi, is removed by the processing by the opening operation. On the other hand, if the length of the elongated non-calcified shadow is longer than the structural element Bi and the inclination coincides with any of the four structural elements Bi, the maximum value of the opening process for each structural element Bi (formula Even if the calculation of the second term of (21) is obtained, it remains as it is. Therefore, by removing the smoothed image thus obtained (image from which only the calcified shadow has been removed) from the original image f, an image including only a small calcified shadow can be obtained. This is the idea of equation (21).

As described above, in the case of a signal having a high density and a high signal level, the calcified shadow has a lower density value than the surrounding image portion, and the calcified shadow is a signal change portion that is concave with respect to the surrounding portion. Therefore, the closing process is applied instead of the opening process, and the expression (22) is applied instead of the expression (21).

Thus, morphological processing is
(1) It is effective for extracting the calcified shadow itself, (2) It is not easily influenced by complicated background information, and (3) The extracted calcified shadow is not distorted. That is, this method can perform detection while maintaining geometric information such as the size, shape, and concentration distribution of the calcified shadow better than general differential processing.

  As described above, in order to improve the image interpretation performance, it is indispensable to perform image processing on the target image. However, as disclosed in JP-A-2-1078, In the enhancement processing based on density only, for example, components that obstruct image interpretation, such as radiation noise components in mammograms, are emphasized, so that the interpretation performance is rather lowered.

  Further, as disclosed in JP-B-60-192482, JP-A-2-120985, JP-T-3-502975, etc., the enhancement processing depending on the dispersion value of the image signal has a large density change locally. Since the image portion is strongly emphasized, undershoot and overshoot are relatively conspicuous in the vicinity thereof, and there is a problem that artifacts are likely to occur on the high density side particularly with respect to X-ray images.

  The present invention has been made in view of the above circumstances, and without emphasizing components unnecessary for image interpretation, such as noise components, only a specific image portion of interest is efficiently enhanced to suppress the occurrence of artifacts. An object of the present invention is to provide an image processing method and apparatus.

According to the first image processing method of the present invention, an original image signal Sorg representing an image is subjected to a morphological operation using a structural element Bi and a scale factor λ so that the image signal is spatially converted to the structural element. Extracting a morphological signal Smor indicating that the pixel corresponds to an image portion that fluctuates smaller than Bi and / or an image portion in which the change in the original image signal Sorg is steep,
First emphasis processing according to the morphological signal Smor is performed on the original image signal Sorg so as to enhance the image portion,
The first processed image signal S ′ is emphasized so as to enhance an image portion corresponding to a desired frequency band in the first processed image signal S ′ obtained by performing the first enhancement processing. On the other hand, a second emphasis process according to the first processed image signal S ′ is performed.

  In addition, as the structural element B constituting the structural element Bi, for example, a vertically, horizontally symmetrical element such as a square, a rectangle, a circle, an ellipse, a linear shape, or a rhombus is desirable. The same applies to the following inventions.

  Further, since the frequency band to be enhanced by the first enhancement process and the frequency band to be enhanced by the second enhancement process are different from each other, different frequency bands of the image can be enhanced respectively. Although it is more preferable, it is not necessarily limited to such different frequency bands, and some bands may overlap.

According to the second image processing method of the present invention, a morphological operation using the structural element Bi and the scale factor λ is performed on the original image signal Sorg representing the image so that the image signal is spatially converted to the structural element. Extracting a morphological signal Smor indicating that the pixel corresponds to an image portion that fluctuates smaller than Bi and / or an image portion in which the change in the original image signal Sorg is steep,
Determining an unsharp mask signal Sus corresponding to a first predetermined spatial frequency of the original image signal Sorg;
A first processed image signal S ′ is obtained by performing an enhancement process according to the following equation (1) on the original image signal with an enhancement coefficient αm (Smor) based on the morphological signal Smor.

Determining a non-sharp mask signal S′us corresponding to a second predetermined spatial frequency of the first processed image signal S ′;
The first processed image signal S ′ is subjected to the enhancement processing according to the following equation (2) by the enhancement coefficient β (S ′) based on the first processed image signal S ′, and the second processed image signal S ′. After obtaining the processed image signal Sproc or obtaining the unsharp mask signal S′us, the enhancement factor β (S′us) based on the unsharp mask signal S′us of the first processed image signal S ′ The first processed image signal S ′ is subjected to enhancement processing according to the following equation (3) to obtain a second processed image signal Sproc.

  Here, by using the non-sharp mask signals Sus and S′us based on non-sharp masks having different sizes, different frequency bands can be emphasized. Thus, it is not limited to having different sizes. The same applies to the following inventions.

  The enhancement coefficient αm (Smor) preferably has a function shape as shown in FIG. That is, the function shape as shown in FIG. 2 corresponds to a desired image portion such as a calcified shadow by setting the output to 0 (zero) for a region having a small morphological signal value Smor that is a radiation noise region (granular region). The region where the morphological signal value Smor is extremely large is fixed to the upper limit value of αm (Smor), and the intermediate region is set so as to increase monotonously as the morphological signal value Smor increases.

  Similarly, the enhancement coefficient β (S ′) preferably has a function shape as shown in FIG. 3, for example. The same applies to the enhancement coefficient β (S′us) in the following invention.

Furthermore, as the morphological operation, various types represented by the following formulas (7) to (12) can be applied.

  That is, by applying the morphological operation represented by the equation (7), the value of the original image signal Sorg is larger than the surrounding image portion and spatially smaller than the structural element Bi as the morphological signal Smor. It is possible to extract a signal of a pixel constituting an image portion (for example, a calcified shadow in an image signal having a high luminance and a high signal level), and to effectively enhance the image portion.

  Further, by applying the morphological operation represented by the equation (8), the value of the original image signal Sorg is smaller than the surrounding image part and is smaller than the structural element Bi as the morphological signal Smor. It is possible to extract a signal of a pixel constituting an image portion (for example, a calcified shadow in an image signal having a high density and a high signal level), and to effectively enhance the image portion.

  By applying the morphological operation represented by Equation (9), an image portion in which the value of the original image signal Sorg is larger than the surrounding image portion and spatially smaller than the structural element Bi as the morphological signal Smor. Alternatively, it is possible to extract a signal of a pixel constituting an edge portion where the luminance (density) changes rapidly, and it is possible to effectively enhance such an image portion.

  By applying the morphological operation represented by the equation (10), an image portion in which the value of the original image signal Sorg is smaller than the surrounding image portion and spatially smaller than the structural element Bi as the morphological signal Smor. Alternatively, it is possible to extract a signal of a pixel constituting an edge portion where the luminance (density) changes rapidly, and it is possible to effectively enhance such an image portion.

  By applying the morphological operation represented by the expression (11), the density of the morphological signal Smor is such that the value of the original image signal Sorg is larger than the surrounding image portion and spatially smaller than the structural element Bi. It is possible to extract signals of pixels constituting an image portion (for example, the skeleton portion of the image represented by the original image signal Sorg) having a large (brightness) change, and effectively enhance such an image portion (skeleton portion). can do.

  An example in which Equation (11) is specifically applied is shown in FIG. FIG. 14 shows a difference signal λ (λ = 1, 2) between an erosion-processed image and an erosion-processed image of an erosion-processed image in a structural element B (a circular structure having a radius r) with respect to the original image X. ,..., N) times up to the union shows that the skeleton parts a and b are obtained.

  Further, by applying the morphological operation represented by the equation (12), the value of the original image signal Sorg is smaller than the surrounding image portion and spatially smaller than the structural element Bi as the morphological signal Smor. It is possible to extract signals of pixels constituting an image portion having a large density (brightness) change (for example, a skeleton portion of an image represented by the original image signal Sorg), and to effectively extract such an image portion (skeleton portion). Emphasis processing can be performed. The same applies to the following inventions.

  Note that the morphological operation represented by the equations (11) and (12) is generally referred to as skeleton processing, and according to this skeleton processing, particularly when applied to the image signal of the bone portion, only the skeleton element is selectively selected. Effective enhancement processing can be performed.

The third image processing method of the present invention performs a morphological operation using the structural element Bi and the scale factor λ on the original image signal Sorg representing the image, thereby spatially converting the image signal into the structural element. Extracting a morphological signal Smor indicating that the pixel corresponds to an image portion that fluctuates smaller than Bi and / or an image portion in which the change in the original image signal Sorg is steep,
The divided into frequency components S _n of the original image signal Sorg a different plurality (n pieces)
The original image signal Sorg is subjected to enhancement processing according to the following equation (4) by using a plurality of different enhancement coefficients αm _n (Smor) respectively corresponding to the plurality of frequency components S _n, and first processed Obtain the image signal S ′,

  The enhancement processing according to the above equation (2) is performed on the first processed image signal S ′ by the enhancement coefficient β (S ′) based on the first processed image signal S ′, and the second processing is performed. After obtaining the processed image signal Sproc or obtaining the first processed image signal S ′, the enhancement coefficient β (S′us) based on the unsharp mask signal S′us of the first processed image signal S ′ Thus, the second processed image signal Sproc is obtained by performing the enhancement processing according to the above equation (3) on the first processed image signal S ′.

Note that n described as a subscript of the frequency component S _n and the emphasis coefficient αm _n (Smor) corresponds to the plurality of frequency components.

The fourth image processing method of the present invention, with respect to the original image signal Sorg representing an image, after setting structural elements Bi _n of a plurality of kinds of sizes and / or shapes different from each other, the plurality of types of structures by performing a morphology operation using the elements Bi _n, and scale factor lambda, image portions and / or the original image signal Sorg changes steep image the image signal varies less than the spatially each structural element Bi _n Extracting a plurality of morphological signals Smor _n indicating pixels corresponding to the portion;
The divided original image signal Sorg a different plurality of frequency components S _n together,
The original image signal is subjected to enhancement processing according to the following equation (5) with the enhancement coefficient αm (Smor _n ) based on each morphological signal Smor _n to obtain a first processed image signal S ′.

Determining a non-sharp mask signal S′us corresponding to a predetermined spatial frequency of the first processed image signal S ′;
The enhancement processing according to the above equation (2) is performed on the first processed image signal S ′ by the enhancement coefficient β (S ′) based on the first processed image signal S ′, and the second processing is performed. After obtaining the processed image signal Sproc or obtaining the unsharp mask signal S′us, the enhancement factor β (S′us) based on the unsharp mask signal S′us of the first processed image signal S ′ The first processed image signal S ′ is subjected to enhancement processing according to the above equation (3) to obtain a second processed image signal Sproc.

Note that n described as a subscript of the structural element Bi _n and the morphology signal Smor _n corresponds to the plurality of frequency components.

The fifth image processing method of the present invention, with respect to the original image signal Sorg representing an image, after setting structural elements Bi _n of a plurality of kinds of sizes and / or shapes different from each other, the plurality of types by performing a morphology operation using the structuring element Bi _n and scale factor lambda, the change in image portions and / or the original image signal Sorg said image signal varies less than the spatially each structural element Bi _n steep Extracting a plurality of morphological signals Smor _n indicating pixels corresponding to the image portion;
The divided original image signal Sorg a different plurality of frequency components S _n together,
Corresponding to said plurality of frequency components S _n, wherein the respective morphology signal Smor _n more emphasis mutually different each based coefficients αm _{n (Smor n),} according to the following formula (6) with respect to the original image signal Sorg To obtain a first processed image signal S ′

  The enhancement processing according to the above equation (2) is performed on the first processed image signal S ′ by the enhancement coefficient β (S ′) based on the first processed image signal S ′, and the second processing is performed. After obtaining the processed image signal Sproc or obtaining the first processed image signal S ′, the enhancement coefficient β (S′us) based on the unsharp mask signal S′us of the first processed image signal S ′ Thus, the second processed image signal Sproc is obtained by performing the enhancement processing according to the above equation (3) on the first processed image signal S ′.

The first image processing apparatus of the present invention performs a morphological operation using the structural element Bi and the scale factor λ on the original image signal Sorg representing the image, thereby spatially converting the image signal into the structural element. Morphological signal calculation means for extracting a morphological signal Smor indicating that the pixel corresponds to an image portion that fluctuates smaller than Bi and / or a change in the original image signal Sorg is a steep image portion;
First enhancement means for performing a first enhancement process according to the morphological signal Smor on the original image signal Sorg so as to enhance the image portion;
The first processed image signal S ′ is emphasized so as to enhance an image portion corresponding to a desired frequency band in the first processed image signal S ′ obtained by performing the first enhancement processing. On the other hand, the image processing apparatus includes a second enhancement unit that performs a second enhancement process according to the first processed image signal S ′.

Further, the second image processing apparatus of the present invention performs a morphological operation using the structural element Bi and the scale factor λ on the original image signal Sorg representing the image so that the image signal is spatially Morphological signal calculation means for extracting a morphological signal Smor indicating that the pixel corresponds to an image portion that fluctuates smaller than the structural element Bi and / or a change in the original image signal Sorg is a steep image portion;
First unsharp mask signal computing means for obtaining an unsharp mask signal Sus corresponding to a first predetermined spatial frequency of the original image signal Sorg;
A first conversion table for receiving an input of the morphological signal Smor and outputting an enhancement coefficient αm (Smor) corresponding to the morphological signal Smor;
A first processed image signal S ′ is obtained by performing enhancement processing according to the above equation (1) on the original image signal by using the enhancement coefficient αm (Smor) output by the first conversion table. 1 emphasis means,
Second unsharp mask signal computing means for obtaining an unsharp mask signal S′us corresponding to a second predetermined spatial frequency of the first processed image signal S ′;
A second conversion table for receiving the input of the first processed image signal S ′ and outputting an enhancement coefficient β (S ′) corresponding to the first processed image signal S ′; and the second conversion The second processed image signal Sproc is obtained by applying the enhancement processing according to the above equation (2) to the first processed image signal S ′ by the enhancement coefficient β (S ′) output from the table. Or a second emphasis means, or in place of the second conversion table and the second emphasis means,
A second conversion table that receives the input of the unsharp mask signal S′us and outputs an enhancement coefficient β (S′us) corresponding to the unsharp mask signal S′us, and is output by the second conversion table. The second processed image signal Sproc is obtained by applying the enhancement processing according to the above equation (3) to the first processed image signal S ′ by the enhancement coefficient β (S′us). It is characterized by comprising an emphasizing means.

The third image processing apparatus of the present invention performs a morphological operation using the structural element Bi and the scale factor λ on the original image signal Sorg representing the image, thereby spatially converting the image signal into the structural element. Morphological signal calculation means for extracting a morphological signal Smor indicating that the pixel corresponds to an image portion that fluctuates smaller than Bi and / or a change in the original image signal Sorg is a steep image portion;
A frequency band dividing means for dividing the original image signal Sorg into different frequency components S _n together,
A plurality of first conversion tables different from each other for receiving the input of the morphological signal Smor and outputting enhancement coefficients αm _n (Smor) corresponding to the respective frequency components S _n ;
The original image signal is subjected to enhancement processing according to the above equation (4) by using a plurality of enhancement coefficients αm _n (Smor) output by the plurality of first conversion tables, respectively. First enhancement means for determining the signal S ′;
Unsharp mask signal computing means for obtaining an unsharp mask signal S′us corresponding to a predetermined spatial frequency of the first processed image signal S ′;
A second conversion table for receiving the input of the first processed image signal S ′ and outputting an enhancement coefficient β (S ′) corresponding to the first processed image signal S ′; and the second conversion The second processed image signal Sproc is obtained by applying the enhancement processing according to the above equation (2) to the first processed image signal S ′ by the enhancement coefficient β (S ′) output from the table. Or a second emphasis means, or in place of the second conversion table and the second emphasis means,
A second conversion table that receives the input of the unsharp mask signal S′us and outputs an enhancement coefficient β (S′us) corresponding to the unsharp mask signal S′us, and is output by the second conversion table. The second processed image signal Sproc is obtained by applying the enhancement processing according to the above equation (3) to the first processed image signal S ′ by the enhancement coefficient β (S′us). It is characterized by comprising an emphasizing means.

The fourth image processing apparatus of the present invention, with respect to the original image signal Sorg representing an image, after setting structural elements Bi _n of a plurality of kinds of sizes and / or shapes different from each other, the plurality of types of structures by performing a morphology operation using the elements Bi _n, and scale factor lambda, image portions and / or the original image signal Sorg changes steep image the image signal varies less than the spatially each structural element Bi _n A plurality of morphological signal calculation means for extracting a plurality of morphological signals Smor _n indicating pixels corresponding to the portion;
A frequency band dividing means for dividing the original image signal Sorg into different frequency components S _n together,
A first conversion table that receives an input of each of the morphological signals Smor _n and outputs a plurality of enhancement coefficients αm (Smor _n ) corresponding to the morphological signals Smor _n ;
Based on the enhancement coefficient αm (Smor _n ) output from the first conversion table, the original image signal is subjected to enhancement processing according to the above equation (5) to obtain the first processed image signal S ′. A first emphasis means;
Unsharp mask signal computing means for obtaining an unsharp mask signal S′us corresponding to a predetermined spatial frequency of the first processed image signal S ′;
A second conversion table for receiving the input of the first processed image signal S ′ and outputting an enhancement coefficient β (S ′) corresponding to the first processed image signal S ′; and the second conversion The second processed image signal Sproc is obtained by applying the enhancement processing according to the above equation (2) to the first processed image signal S ′ by the enhancement coefficient β (S ′) output from the table. Or a second emphasis means, or in place of the second conversion table and the second emphasis means,
A second conversion table that receives the input of the unsharp mask signal S′us and outputs an enhancement coefficient β (S′us) corresponding to the unsharp mask signal S′us, and is output by the second conversion table. The second processed image signal Sproc is obtained by applying the enhancement processing according to the above equation (3) to the first processed image signal S ′ by the enhancement coefficient β (S′us). It is characterized by comprising an emphasizing means.

Fifth image processing apparatus of the present invention, with respect to the original image signal Sorg representing an image, after setting structural elements Bi _n of a plurality of kinds of sizes and / or shapes different from each other, the plurality of types of structures by performing a morphology operation using the elements Bi _n, and scale factor lambda, image portions and / or the original image signal Sorg changes steep image the image signal varies less than the spatially each structural element Bi _n A plurality of morphological signal calculation means for extracting a plurality of morphological signals Smor _n indicating pixels corresponding to the portion;
A frequency band dividing means for dividing the original image signal Sorg into different frequency components S _n together,
Receiving said input of each morphology signal Smor _n, the outputs of the corresponding plurality of enhancement coefficient αm _{n (Smor n)} to each of the frequency components S _n, different from the plurality of first conversion table to each other,
A plurality of emphasis coefficients outputted respectively by a first conversion table of the plurality of αm _{n (Smor n),} a first processed by performing enhancement processing in accordance with the equation (6) with respect to the original image signal First enhancement means for obtaining an image signal S ′;
Unsharp mask signal computing means for obtaining an unsharp mask signal S′us corresponding to a predetermined spatial frequency of the first processed image signal S ′;
A second conversion table for receiving the input of the first processed image signal S ′ and outputting an enhancement coefficient β (S ′) corresponding to the first processed image signal S ′; and the second conversion The second processed image signal Sproc is obtained by applying the enhancement processing according to the above equation (2) to the first processed image signal S ′ by the enhancement coefficient β (S ′) output from the table. Or a second emphasis means, or in place of the second conversion table and the second emphasis means,
A second conversion table that receives the input of the unsharp mask signal S′us and outputs an enhancement coefficient β (S′us) corresponding to the unsharp mask signal S′us, and is output by the second conversion table. The second processed image signal Sproc is obtained by applying the enhancement processing according to the above equation (3) to the first processed image signal S ′ by the enhancement coefficient β (S′us). It is characterized by comprising an emphasizing means.

  The structural element Bi (i = 1, 2,..., M) means a set of M structural elements B prepared as structural elements B having different directions in the two-dimensional plane. In the case of M = 1, it means a vertically and horizontally symmetrical element, and in the present invention, it is expressed as a structural element Bi including a multiple structural element in which i ≧ 2 and i = 1. The scale factor λ means the number of times the Minkowski sum calculation and Minkowski difference calculation are performed, and the degree of smoothing increases as the number is increased.

  The first image processing method / apparatus of the present invention applies a morphological operation using the structural element Bi and the scale factor λ to the original image signal Sorg representing the image, thereby spatially converting the image signal into the structural element Bi. Only the image portion that varies more and / or the image portion where the change of the original image signal Sorg is steep is extracted. The extracted image portion includes a desired image portion, but also includes radiation noise which is a so-called high frequency component. Here, the morphological signal Smor corresponding to radiation noise is very small, while the morphological signal Smor for a desired image portion shows a larger value than the morphological signal Smor for this radiation noise.

  The morphological signal is then applied to the original image signal Sorg such that the image signal spatially varies less than the structural element Bi and / or the image portion where the change of the original image signal Sorg is steep. By performing the first enhancement process according to Smor, it is possible to obtain a first processed image signal S ′ in which the image portion is selectively enhanced. As described above, the first processed image signal S ′ is an enhancement process depending on the morphology signal.

  Next, a first process is performed on the first processed image signal S ′ so as to emphasize an image portion corresponding to a desired frequency band in the obtained first processed image signal S ′. By performing the second enhancement process according to the completed image signal S ′, the enhancement process depending on the first processed image signal S ′ such as a density value can be performed.

  As described above, according to the first image processing method and apparatus of the present invention, the enhancement of the radiation noise unnecessary for the image interpretation is suppressed by the enhancement processing depending on the morphological signal (referred to as the first enhancement processing). It is possible to effectively enhance only the image portion that is spatially smaller than the structural element Bi as the signal component and / or the image portion in which the original image signal Sorg has a sharp change.

  Further, the enhancement processing depending on the first processed image signal (referred to as the second enhancement processing) can also enhance the image signals in other frequency bands that cannot be enhanced by the first enhancement processing. .

  As described above, according to the first image processing method / device of the present invention, it is possible to efficiently emphasize only a specific image portion of interest without enhancing components unnecessary for image interpretation such as noise components. it can. Then, by performing the enhancement process depending on the signal such as the density on the first processed image signal, it is possible to suppress the occurrence of artifacts caused by the dispersion value dependent enhancement process.

  It should be noted that by emphasizing the frequency bands to be emphasized by the first enhancement process and the frequency bands to be enhanced by the second enhancement process, the two frequency bands can be emphasized with different degrees of enhancement. it can.

  Furthermore, according to the first image processing method and apparatus of the present invention, in the morphological calculation, the processing for calculating the maximum value or the minimum value is performed instead of the processing for calculating the variance value of the image signal, so that the calculation time is shortened. can do.

  Hereinafter, Equation (8), which is an example of the morphological operation, will be specifically described. For simplicity, λ = 1 and i = 1.

According to the morphological calculation for the density value Sorg which is an image signal having a high density and high signal level, for example, (2) for the image data having the density value Sorg distribution as shown by the solid line in FIG. ), The density value S _i of a certain pixel of interest is obtained by performing the Minkowski sum operation on three linear structural elements B as shown in FIG. _Is converted to S _i ′ employing the maximum value S _{i + 1} of By performing this calculation for all the pixels, the image data is converted into image data indicated by a broken line in FIG.

Next, considering the Minkowski difference due to the structural element B of the density value obtained by the calculation of the Minkowski sum, the density value S _i ′ of the target pixel indicated by the broken line in FIG. This is converted into S _i ″ (= S _i ) that employs the minimum value S _i−1 ′ among the three pixels adjacent to each other as the center. By performing this calculation for all the pixels, the distribution of the density value Sorg ″ is changed. It is converted into image data indicated by the alternate long and short dash line in FIG. In the image data indicated by the alternate long and short dash line, the image portion of the change in the signal value finer than that of the structural element B disappears from the original image data of the solid line, and the change in the signal value larger than that of the structural element B This image portion indicates that the original state before the operation is maintained. That is, the above process (closing process) is a process of smoothing the image density distribution from the high density side.

  Thus, the value Smor obtained by subtracting the value obtained by the closing process from the original image signal Sorg represents the image portion of the fine signal value change erased by the closing process.

  Here, the image signal originally has a position (x, y) which is a two-dimensional element and a signal value f (x, y) which is a third dimension element. However, in the above description, it is easy to understand. Therefore, the one-dimensional image signal distribution curve that appears in a predetermined section of the image developed two-dimensionally has been described.

  Accordingly, each image processing method / apparatus of the present invention is obtained by enlarging and applying the above description to a two-dimensional image, and the structural element Bi shown in Expression (8) performs two-dimensional morphological operations on such a cross section. This means a set of i structural elements B prepared as structural elements B having different orientations in the two-dimensional plane when applied to a surface in an enlarged manner.

  Further, as a result of performing the closing process on all these structural elements Bi, an image portion whose extension direction coincides with any of the structural elements Bi and changes more greatly than the size thereof is expressed by the expression (8). Since the value of the second term is Sorg itself, the value of Smor is zero, and the portion is not emphasized.

  The second image processing method / apparatus of the present invention performs the morphological operation using the structural element Bi and the scale factor λ on the original image signal Sorg representing the image, as in the first image processing method / apparatus described above. As a result, only the image portion where the image signal fluctuates spatially smaller than the structural element Bi and / or the image portion where the change of the original image signal Sorg is steep is extracted.

  Next, by subtracting the ultra-low spatial frequency component Sus from the original image signal Sorg, only a relatively high frequency component from which the frequency components below the ultra-low spatial frequency are removed from the original image signal Sorg is extracted. be able to. The extracted relatively high frequency component also includes radiation noise which is a so-called high frequency component.

  Here, an enhancement coefficient αm (Smor) based on a preset morphological signal is obtained according to the obtained morphological signal, and the image signal is enhanced by the obtained enhancement coefficient αm (Smor).

  The enhancement coefficient αm (Smor) based on the morphological signal is, for example, as shown in FIG. 2, the enhancement coefficient αm is zero in a range where the morphological signal Smor corresponding to the radiation noise (granular region) is small. Since the morphological signal Smor corresponding to 1 takes a large value to some extent, it belongs to the “signal region” shown in FIG. 2 and the enhancement coefficient αm has a large value.

  Then, the emphasis coefficient αm (Smor) is multiplied by the relatively high frequency component in accordance with the equation (1), whereby the original image signal S ′ is obtained by emphasizing the relatively high frequency component of the original image signal. Is output as This relatively high frequency component includes a high-frequency component unnecessary for image interpretation diagnosis such as the radiation noise described above, but the enhancement coefficient αm (Smor) corresponding to such an unnecessary high-frequency component is zero or very small. As a result of multiplication by the enhancement coefficient αm (Smor), high frequency components unnecessary for image interpretation diagnosis are not enhanced.

  This is illustrated in FIG. That is, in the original image signal Sorg as shown in FIG. 16A, frequency components in a high frequency band (a frequency band having a spatial frequency higher than lowpas1) are separated according to the size of the unsharp mask, and the signal component thereof Only the emphasis coefficient αm (Smor) is selectively emphasized, and only the signal component as shown in FIG. 5B is converted into the enhanced first processed image signal S ′.

  Next, also for the first processed image signal S ′, a non-sharp mask signal S′us corresponding to an ultra-low spatial frequency is obtained. For example, the enhancement coefficient β (S ′) shown in FIG. Therefore, by multiplying a relatively high frequency component of the first processed image signal S ′, a relatively high frequency component of the first processed image signal S ′ is included in the first processed image signal S ′. It is output as a second processed image signal Sproc that has been subjected to enhancement processing according to the magnitude.

  This is because frequency components in the high frequency band (frequency band having a higher spatial frequency than lowpas2) are separated from the first processed image signal S ′ as shown in FIG. 16B according to the size of the unsharp mask. Thus, the entire image signal in the high frequency band is emphasized by the enhancement coefficient β (S ′) and converted into the second processed image signal Sproc as shown in FIG.

  Thus, according to the second image processing method and apparatus of the present invention, it is possible to efficiently emphasize only a specific image portion of interest without enhancing components unnecessary for image interpretation such as noise components. it can. Then, by performing the enhancement process depending on the signal such as the density on the first processed image signal, it is possible to suppress the occurrence of artifacts caused by the dispersion value dependent enhancement process.

  It should be noted that the non-sharp mask process for the original image signal Sorg and the non-sharp mask process for the first processed image signal S ′ are different in size between the two frequency bands. Can be emphasized by changing the degree of emphasis.

  Furthermore, according to the second image processing method and apparatus of the present invention, the morphological calculation performs the process of calculating the maximum value or the minimum value instead of the process of calculating the variance value of the image signal, thereby reducing the calculation time. can do.

Third image processing method, apparatus of the present invention, in the image processing method and devices described above, by dividing the original image signal Sorg into different frequency components S _n together, each frequency obtained by the division for components S _n, by each enhancement by separate emphasis coefficient αm n _(Smor), the desired size in the respective frequency components S _n, can be adjusted selectively emphasis degree structures shape . Then, the first processed image signal S ′ after the enhancement process is subjected to a signal value (density) -dependent enhancement process to obtain a second processed image signal Sproc, whereby the dispersion value dependent enhancement is performed. It is possible to suppress the occurrence of artifacts generated in the processing. Other operations and effects are the same as those of the above-described methods and apparatuses.

The fourth image processing method, apparatus of the present invention, in the image processing method and devices described above, by using a plurality of types of structural elements Bi _n mutually different size and / or shape, a plurality of sizes to obtain a desired obtain a plurality of morphology image signals Smor _n corresponding to the and / or the image portion of the shape, thereby obtains a plurality of emphasis coefficients αm (Smor _n), whereas a plurality of types also unsharp mask corresponding to the structural elements Bi _n The original image signal is separated into a plurality of types of frequency bands, and the image signal of the corresponding frequency band is multiplied by the enhancement coefficient αm (Smor _n ) corresponding to each morphological image signal Smor _n. The emphasis process is performed separately, and the sum of the emphasis is obtained as shown in Equation (5), thereby changing the emphasis degree for each of the plurality of spatial frequency bands. Obtaining a first processed image signal S '. The obtained first processed image signal S ′ is subjected to the signal value (density) -dependent emphasis processing of the above equation (2) to obtain a second processed image signal Sproc.

  Accordingly, even when there are a plurality of types of desired image portions to be enhanced instead of one size and shape, each image portion can be enhanced with a desired enhancement degree, and the above-described present invention can be applied. As in each of the image processing methods / apparatuses described above, it is possible to efficiently emphasize only a specific image portion of interest without enhancing components unnecessary for image interpretation such as noise components. Then, by performing the enhancement process depending on the signal such as the density on the first processed image signal, it is possible to suppress the occurrence of artifacts caused by the dispersion value dependent enhancement process. In addition, in the morphological calculation, the processing for calculating the maximum value or the minimum value is performed instead of the processing for calculating the variance value of the image signal, so that the calculation time can be shortened.

  The fifth image processing method / apparatus of the present invention is a combination of the third image processing method / apparatus of the present invention and the fourth image processing method / apparatus, and the functions and effects thereof are as follows. This is the same as the third image processing method / device and the fourth image processing method / device.

  Hereinafter, an image processing apparatus for specifically carrying out the image processing method of the present invention will be described with reference to the drawings.

FIG. 1 is a schematic block diagram showing an image processing apparatus according to this embodiment. The illustrated image processing apparatus includes a first low-pass filter (denoted as lowpass 1 in the drawing) 11 for obtaining an unsharp mask signal Sus corresponding to an extremely low spatial frequency of an original image signal Sorg representing an image, and an original image signal. Morphological signal indicating that the original image signal Sorg is a pixel corresponding to an image portion that fluctuates smaller than the structural element Bi by performing a morphological operation using the structural element Bi and the scale factor λ on Sorg. A morphological filter (Morphology-filter in the drawing) 12 for extracting Smor and a first conversion table 13 which receives an input of the morphological signal Smor and outputs an enhancement coefficient αm (Smor) according to the morphological signal Smor. And the enhancement coefficient αm (Smor) output from the first conversion table 13, the original image signal So The arithmetic elements 14a, 14b, 14c that perform arithmetic processing according to the following formula (1) to rg, and the image portions that fluctuate spatially smaller than the multiple structural elements Bi as a result of the arithmetic operations performed by these arithmetic elements are emphasized. A second low-pass filter (denoted as lowpass 2 in the drawing) 21 for obtaining a non-sharp mask signal S′us corresponding to an ultra-low spatial frequency of the processed first processed image signal S ′; The second conversion table 23 which receives the input of the processed image signal S ′ and outputs the enhancement coefficient βd (S ′) corresponding to the first processed image signal S ′ and the second conversion table output Arithmetic elements 24a, 24b, 24c that obtain the second processed image signal Sproc by performing the arithmetic processing according to the following equation (2) on the first processed image signal S ′ by the enhancement coefficient βd (S ′); It is the structure provided with.

  Here, the original image signal Sorg representing an image may be read in advance from a radiographic image by a predetermined image reading device and stored in a predetermined storage means, or may be input directly from the image reading device. There may be.

In addition, the radiation image in the present embodiment is a mammogram, and the first low-pass filter 11 sets, for example, a blur mask composed of a pixel matrix of 3 columns × 3 rows for the original image signal Sorg, and the following formula (14) ( The blur mask signal Sus obtained according to (set N = 3) is output.

  The blur mask uses a simple average of the pixel values in the mask as shown in the above formula (14), or a mask corresponding to the distance from the center pixel as shown in the lower right matrix in FIG. It is also possible to use a pixel in which the weight of the pixel value is changed.

The morphological filter 12 performs arithmetic processing on the original image signal Sorg according to a closing process shown in the following equation (8) by using a structural element B composed of a pixel matrix of 5 columns × 5 rows and a scale factor λ, for example. An image signal corresponding to an image portion in which the value of the original image signal Sorg is smaller than that of the surrounding image portion and fluctuates spatially smaller than that of the structural element Bi (for example, a minute calcification portion indicating breast cancer in a mammogram). Is input, a large value morphological signal Smor is output. On the other hand, an image corresponding to an image portion in which the value of the original image signal Sorg is larger than that of the surrounding image portion or spatially fluctuates more than the structural element Bi. When a signal is input, a morphological signal Smor having an extremely small value is output. The structural element B is preset according to the shape and size of a desired microcalcification portion to be subjected to the emphasis process.

  For example, as shown in FIG. 2, the first conversion table 13 sets the output to 0 (zero) for a region having a small morphological signal value Smor that is a radiation noise region, and corresponds to a desired image portion such as a calcified shadow. The region where the morphological signal value Smor is extremely large is fixed to the upper limit value of αm (Smor), and these intermediate regions are set so as to increase monotonously as the morphological signal value Smor increases.

  The arithmetic element 14a is an arithmetic element for subtracting the output Sus of the first low-pass filter 11 from the original image signal Sorg, and the arithmetic element 14b is an arithmetic element for multiplying the output of the arithmetic element 14a and the output of the first conversion table 13; The element 14c is an arithmetic element that adds the original image signal Sorg and the output of the arithmetic element 14b.

  The second low-pass filter 21 sets a blur mask composed of a pixel matrix of 29 columns × 29 rows, for example, for the first processed image signal S ′, and follows the above equation (14) (set to N = 29). The obtained blur mask signal S'us is output.

  For example, as shown in FIG. 3, the second conversion table 23 fixes the first processed image signal to an upper limit value of βd (S ′) for a region where the first processed image signal S ′ is somewhat large. In a region where S ′ is smaller than this, it is set so as to monotonously increase as the first processed image signal S ′ increases.

  The arithmetic element 24a is an arithmetic element that subtracts the output S'us of the second low-pass filter 21 from the first processed image signal S '. The arithmetic element 24b is an output of the arithmetic element 24a and an output of the second conversion table 23. The arithmetic element 24c is an arithmetic element for adding the first processed image signal S 'and the output of the arithmetic element 24b.

  Next, the operation of the image processing apparatus of this embodiment will be described.

  When the original image signal Sorg is input to the image processing apparatus, first, the first low-pass filter 11 sets a blur mask composed of a pixel matrix of 3 columns × 3 rows for the original image signal Sorg, and the above equation (14). A calculation according to (N = 3) is performed to output the first blur mask signal Sus. Since the first blur mask signal Sus is a blur mask composed of a pixel matrix of 3 columns × 3 rows, the first blur mask signal Sus has a relatively high frequency.

  The arithmetic element 14a subtracts the first blur mask signal Sus which is the output of the first low-pass filter 11 from the original image signal Sorg, and outputs only the high frequency component (Sorg-Sus) of the original image signal Sorg.

  In parallel with the operation of the first low-pass filter 11 and the arithmetic element 14a, the morphological filter 12 applies a structural element B composed of a pixel matrix of 5 columns × 5 rows to the original image signal Sorg and a scale factor λ. Thus, the arithmetic processing according to the above equation (8) is performed, and the morphological signal Smor corresponding to the characteristic shape of the image portion and the fluctuation of the signal value is output. Here, when the image portion is a microcalcification portion having an abnormal shadow, a large value morphological signal Smor is output, and in other cases, a very small value morphological signal Smor is output.

  The morphological signal Smor output from the morphological filter 12 is input to the first conversion table 13, and the first conversion table 13 sets the enhancement coefficient αm (Smor) according to the magnitude of the input morphological signal Smor. Output. In the present embodiment, the substantially upper limit value is output as the enhancement coefficient αm (Smor) in the microcalcification portion, and substantially zero is output as the enhancement coefficient αm (Smor) in the other portions.

  The arithmetic element 14b multiplies the output (Sorg-Sus) from the arithmetic element 14a by the output αm (Smor) from the first conversion table 13 and weights the high-frequency component (Sorg-Sus).

  Next, the arithmetic element 14c adds the original image signal Sorg and the output from the arithmetic element 14b, and outputs a first processed image signal S '. The output first processed image signal S ′ is a signal obtained by emphasizing a minute calcified portion of the high-frequency component (Sorg−Sus) of the original image signal by the above-described processing.

  Next, the first processed image signal S ′ in which the minute calcified portion is enhanced is converted into the first processed image signal S ′ in which the second low-pass filter 21 is enhanced. On the other hand, a blur mask composed of a pixel matrix of 29 columns × 29 rows is set, and a calculation according to the above equation (14) (set to N = 29) is performed to output a second blur mask signal S′us. The second blur mask signal S'us is a blur mask made up of a relatively large pixel matrix of 29 columns and 29 rows, and therefore has a relatively low frequency.

  The arithmetic element 24a subtracts the second blur mask signal S'us, which is the output of the second low-pass filter 21, from the first processed image signal S ', thereby obtaining a high frequency signal in the first processed image signal S'. Only the component (S'-S'us) is output. However, this high frequency component includes even a frequency component lower than the high frequency component (Sorg-Sus) of the original image signal Sorg described above.

  In parallel with the operation of the second low-pass filter 21 and the arithmetic element 24a, the first processed image signal S 'is input to the second conversion table 23, and the second conversion table 23 is The enhancement coefficient βd (S ′) corresponding to the size of one processed image signal S ′ is output.

  The arithmetic element 24b multiplies the output (S'-S'us) from the arithmetic element 24a and the output .beta.d (S ') from the second conversion table 23, and performs a high frequency component (S'-S'us). ). Next, the arithmetic element 24c adds the first processed image signal S 'and the output from the arithmetic element 24b, and outputs a second processed image signal Sproc. The output second processed image signal Sproc has the high frequency component (S′−S′us) of the first processed image signal S ′ enhanced by the above-described processing, and particularly the first processed image signal. As the signal S ′ increases (for example, as the density increases in a negative image), the degree of enhancement increases.

  Therefore, according to the image processing apparatus of the present embodiment, an image portion having a frequency higher than the second frequency corresponding to the size of the blur mask set by the second low-pass filter 21 is enhanced, and this enhancement processing is performed. Radiation noise in the first frequency band (second frequency <first frequency) corresponding to the size of the blur mask set by the first low-pass filter 11 in the higher frequency band than the set second frequency It is possible to further enhance the microcalcification portion while suppressing the emphasis. This is particularly true for image portions that are also to be enhanced in an intermediate frequency band between the second frequency and the first frequency that are outside the first frequency band (eg, a mass that is another morphological feature of breast cancer). When there are shadows, etc.), they can be emphasized individually with a desired degree of enhancement, which is very useful in computer-aided image diagnosis and the like.

  Note that, as the second conversion table 23 in the present embodiment, an output βd (S′us) corresponding to the blur mask signal S′us of the first processed image signal S ′ output from the second low-pass filter 21 is used. You may use, and the effect similar to the said embodiment can be acquired. In this case, as shown in FIG. 4, instead of inputting the first processed image signal S ′, the output from the second low-pass filter 21 can be input to the second conversion table 23. That's fine.

  The input signal to the morphological filter 12 is not limited to the original image signal Sorg, but is a difference signal between the original image signal Sorg and the first blur mask signal Sus that is an output signal from the first low-pass filter 13. (Corresponding to the above-mentioned high-frequency component (Sorg-Sus)) may be used. In this case, as shown in FIG. 5, the configuration may be such that the output value of the arithmetic element 14a is input to the morphological filter 12.

  Further, the first enhancement process for obtaining the first processed image signal S ′ may be performed in multiple stages for each frequency band.

That is, for example, as shown in FIG. 6, three low-pass filters 11a, 11b, and 11c of the same type are connected in series, and arithmetic elements 14a ₁ and 14a ₁ are connected between output signals and input signals of the respective low-pass filters 11a, 11b, and 11c. performing the calculation of each difference by 14a _2, 14a _3. This is because the frequency component contained in the output signal (blur mask signal) is low every time one low-pass filter is passed, so by calculating the difference between the signals before and after passing through one low-pass filter, It is possible to separate the image signal with the highest frequency component contained in the output signal of the low-pass filter that has passed.

Specifically, as shown in FIG. 6, the blur mask signal after passing through the low-pass filter 11a through which the original image signal Sorg first passes is referred to Sus1, and the blur mask signal after passing through the second low-pass filter 11b is referred to as Sus2,3. th the unsharp mask signal after passing through the low pass filter 11c and SUS316, when the operation element for calculating a difference between the signals before and after the passage of the low-pass filter 14a _1, 14a _2, 14a ₃ to, each processing element 14a ₁ , the output signal _S H of _{14a 2, 14a 3, S M} , S L is as shown below, respectively.

S _H = Sorg -Sus1
S _M = Sus1-Sus2
S _L = Sus2-Sus3
Accordingly, a relatively high frequency (f> f _H component frequency (the f _H) The following frequency components are excluded, including the unsharp mask signal Sus1 for the output signal S _H, the unsharp mask signal to the output signal S _M Sus2 frequency (f and _M) or less of the frequency components included in the output signal S _M to are also excluded frequency or higher frequency components contained in the unsharp mask signal Sus1 with frequency components are excluded f _H ≧ f> containing the f _M, and the frequency components included in the output signal _{S L} similarly becomes _{f M} ≧ _{f> f L} when the frequency contained in the unsharp mask signal Sus3 was _{f L.}

Furthermore, enhancement coefficients αm ₁ (Smor), αm ₂ (Smor), αm ₃ corresponding to the respective frequency bands f _H , f _M , f _L corresponding to the morphological signal Smor are received by receiving the output of the morphological filter 12. (Smor) (hereinafter referred to simply as αm ₁ , αm ₂ , αm ₃ for simplicity) and three first conversion tables 13a, 13b, 13c for outputting, and the respective conversion tables 13a, 13b, 13c An arithmetic element 14b ₁ that performs an operation of multiplying each emphasis coefficient αm ₁ , αm ₂ , αm ₃ and each output _SH , S _M , S _L from the arithmetic elements 14a ₁ , 14a ₂ , 14a ₃ for each corresponding frequency band. 14b ₂ and 14b ₃ and arithmetic elements 14c ₁ , 14c ₂ and 14c ₃ for adding the outputs of the respective arithmetic elements 14b ₁ , 14b ₂ and 14b _{3 and} the output of the third low-pass filter 11c.

Each of the first conversion table 13a, 13b, 13c, for example, as shown in FIG. 7, in the high frequency band _{f H} (A), in the medium frequency band _{f M} is (B), the low frequency band _{f L} , as shown, respectively (C), the enhancement coefficient .alpha.m 1 is output according to morphology signal values Smor corresponding to each frequency _band, .alpha.m _2, and outputs the .alpha.m _3. For convenience of explanation, hereinafter, the conversion table 13a is referred to as a high frequency conversion table, the conversion table 13b is referred to as a medium frequency conversion table, and the conversion table 13c is referred to as a low frequency conversion table.

The morphological signal Smor output from the morphological filter 12 is input to the three first conversion tables 13a, 13b, and 13c, respectively. Among them, the high-frequency conversion table 13a according to the input morphology signal Smor, the aforementioned outputs emphasis coefficient .alpha.m ₁ corresponding to the high frequency components S _H, the morphology signal Smor conversion table 13b for the middle-frequency input Correspondingly, emphasis coefficient outputs emphasis coefficient .alpha.m ₂ corresponding to frequency components S _M in the above, the low-frequency conversion table 13c is in response to the inputted morphology signal Smor, corresponding to the low frequency components S _L of the above αm ₃ is output respectively.

Incidentally, the tentatively .alpha.m ₄ the emphasis coefficients corresponding to the super-low frequency components S _{LL (=} Sus3), may also be arranged to extremely low frequency conversion table for outputting the emphasis coefficient .alpha.m ₄ in accordance with the morphology signal Smor, the In the embodiment, since it is not necessary to individually emphasize the very low frequency component S _LL , it is omitted as αm ₄ = 1 (constant). That is, among the signals divided into n frequency components, the enhancement degree of the remaining one frequency component is controlled relatively by lowering the enhancement degree of (n−1) frequency components to obtain a suppression degree. Because it can be done.

The emphasis coefficients αm ₁ , αm ₂ , and αm ₃ that are outputs from the conversion tables 13a, 13b, and 13c are respectively input to the corresponding arithmetic elements 14b ₁ , 14b _2, and 14b ₃ . On the other hand, the above-described high-frequency component S _H , medium-frequency component S _M , and low-frequency component S _L are also input to the calculation elements 14 b ₁ , 14 b ₂ , 14 b ₃ from the corresponding calculation elements 14 a ₁ , 14 a ₂ , 14 a ₃ , respectively. , operation elements 14b _1, 14b _2, 14b ₃ are each set to correspond them (.alpha.m ₁ and _{S H,} .alpha.m ₂ and _{S M,} .alpha.m ₃ and _{S L)} by multiplying the respective operational elements 14c _1, 14c _2, 14c Output to ₃ . This action means that each frequency component is weighted according to the morphological signal.

The arithmetic elements 14c ₁ , 14c ₂ , 14c ₃ sum the outputs from the arithmetic elements 14b ₁ , 14b ₂ , 14b ₃ and add the very low frequency component S _LL that is the output from the third low-pass filter 11c, The first processed image signal S ′ shown in the following equation (23) is output.

  The first processed image signal S ′ is a signal obtained by performing enhancement processing in accordance with the morphological signal for each frequency component with respect to the original image signal Sorg. The size of the structural element B and the size of the blur mask described above. By selectively changing the number of frequency bands to be divided, and by changing the shape of the enhancement function based on the conversion table, only desired signal components included in each frequency band are selectively emphasized with a desired degree of enhancement. Processing can be performed, and finer adjustment of the emphasis can be made compared with the conventional emphasis processing, and the image diagnosis performance can be improved.

  In the following, the second emphasis process is not described because the same process as that of the above embodiment is performed on the first processed image signal S ′.

In addition, said Formula (23) can be represented like Formula (24) as a general formula.

In other words, Expression (23) is _simply omitted in Expression (24), assuming that the enhancement coefficient to be multiplied by the very low frequency component S _LL is 1 (= constant). Thus, as described above also in this embodiment, to set the emphasis coefficient in accordance with the morphology signal to very low frequency components S _LL, for example, as .alpha.m _4, which may be multiplied to the ultra-low frequency components S _LL .

  The second processed image signal Sproc (see FIG. 8B) obtained in this way is, as shown in FIG. 8, included in each frequency component of the original image signal Sorg (see FIG. 8A). Only the signal components can be selectively processed and the emphasis processing can be performed by freely adjusting the emphasis degree.

FIG. 9 shows the original image signal by the enhancement coefficient (αm + αd) obtained by adding the enhancement coefficient αm according to the morphological signal and the enhancement coefficient αd according to the original image signal in the image processing apparatus of the embodiment shown in FIG. FIG. 2 is a schematic block diagram showing an image processing apparatus of an embodiment in which each frequency component S _H , S _M , and S _L is emphasized.

The image processing apparatus shown in FIG. 6 receives the original image signal Sorg and outputs a monotonically increasing enhancement coefficient αd (Sorg) corresponding to the original image signal Sorg in the image processing apparatus of the embodiment shown in FIG. A third conversion table 16 depending on the image signal, and three amplifiers 15a for amplifying the enhancement coefficient αd (Sorg) corresponding to the original image signal Sorg outputted from the third conversion table 16 with different amplification factors. 15b, and 15c, each amplifier 15a, 15b, the outputs of _{15c αd 1 (Sorg), αd} 2 (Sorg), the enhancement coefficient .alpha.m ₁ to αd ₃ (Sorg) and frequency band based on the mode follower biology filter 12 corresponding respectively (Smor), αm ₂ (Smor), and αm ₃ (Smor) are further associated with arithmetic elements 14d ₁ , 14d ₂ , and 14d ₃ that perform addition processing.

Here, the amplifier 15a, 15b, 15c, when the same input value .alpha.d (Sorg) is inputted, 15a outputs the largest emphasis coefficient αd 1 _(Sorg), emphasis 15c smallest coefficients less .alpha.d ₃ (Sorg) is output, and 15b is set to output an intermediate enhancement coefficient αd ₂ (Sorg) between 15a and 15c.

According to the image processing apparatus according to the embodiment configured as described above, the high frequency components S _H, the enhancement coefficient based on the morphology signal Smor αm 1 _(Smor) and enhancement coefficient .alpha.d 1 _(Sorg based on the original image signal Sorg ) And the enhancement coefficient α ₁ corresponding to the sum of the image and the medium frequency component S _M , the enhancement coefficient αm ₂ (Smor) based on the morphological signal Smor and the enhancement coefficient αd ₂ (Sorg) based on the original image signal Sorg. Enhancement processing is performed with an enhancement coefficient α ₂ corresponding to the sum of and the low frequency component S _L is an enhancement coefficient αm ₃ (Smor) based on the morphological signal Smor and an enhancement coefficient αd ₃ (Sorg) based on the original image signal Sorg. enhancement processing is performed by the emphasis coefficient alpha ₃ in accordance with the sum of.

  As described above, the image processing apparatus according to the present embodiment performs the enhancement process by combining the enhancement process based on the morphological signal and the enhancement process based on the original image signal value. Excessive emphasis (overshoot or undershoot) can be suppressed.

Note that the method of combining the enhancement processing based on the morphological signal and the enhancement processing based on the original image signal value is not limited to using the sum of enhancement coefficients for each enhancement processing as in the above embodiment. These corresponding enhancement coefficients may be multiplied together to obtain a new enhancement coefficient α _n .

Furthermore, the arrangement positions of the arithmetic elements 14c ₁ , 14c ₂ , 14c ₃ are not provided at the position for adding to the output signal Sus3 of the low-pass filter 11c but at the position for adding to the original image signal Sorg as shown in FIG. May be. In this case, the first processed image signal S ′ is as shown in the following equation (25).

  Similarly to Expression (23), this Expression (25) means that a signal in which the degree of enhancement is changed according to the morphological signal for each frequency component is obtained with respect to the original image signal Sorg. Each time, it is possible to perform a desired enhancement process only on an image portion of a desired size included in each frequency band.

  In the image processing apparatus of each of the above embodiments, the output from one morphological filter 12 is input to three different conversion tables to obtain enhancement coefficients for each frequency component of the original image signal. It is also possible to adopt a configuration using three different morphological filters in correspondence with the frequency band.

  That is, FIG. 10 shows a schematic block diagram of an image processing apparatus configured to use such a morphological filter.

The image processing apparatus shown in FIG. 10, in the image processing apparatus of the embodiment shown in FIG. 6, in place of the single motor follower biology filter 12, set corresponding to the high frequency components S _H of the original image signal Sorg first motor follower biology filter 12a and the structural element B _M having a size corresponding to the intermediate frequency components S _M of obtaining a first morphology signal Smor1 of the original image signal Sorg by using the magnitude structural elements B _S of which is second motor follower biology filter 12b and, in the original image signal Sorg by using the structure element B _L having a size corresponding to the low frequency components S _L morphology of obtaining a second morphology signal Smor2 of the original image signal Sorg by using a And a third morphological filter 12c for obtaining the signal Smor3, and the high-frequency conversion table 13a includes the first morphological filter 12. In response to input of a first morphology signal Smor1 is output, the first as the emphasis coefficient αm outputs 1 _(Smor1) in accordance with the morphology signal Smor1, the conversion table 13b for the middle-frequency second motor The low-frequency conversion table 13c receives the second morphological signal Smor2 that is the output of the morphological filter 12b and outputs an enhancement coefficient αm ₂ (Smor2) corresponding to the second morphological signal Smor2. In response to the input of the third morphological signal Smor3 which is the output of the third morphological filter 12c, the enhancement coefficient αm ₃ (Smor3) corresponding to the third morphological signal Smor3 is output.

Each structural element B _L , B _M , B _S has the largest B _L , the smallest B _S , and B _M is an intermediate size between B _L and B _S.

By setting the sizes of the structural elements B _S , B _M , and B _L in this way, each of the morphological filters 12a to 12c causes the morphological signal Smor1 having a characteristic value for an image portion having a corresponding size. , Smor2 and Smor3 are output. In accordance with the morphological signals Smor1, Smor2, and Smor3 output from the morphological filters 12a to 12c, the corresponding conversion tables 13a to 13c have enhancement coefficients αm ₁ (Smor1), αm ₂ (Smor2), and αm ₃ ( Smor3) is output.

The outputted respective emphasis coefficient _{αm 1 (Smor1), αm 2} (Smor2), αm 3 is (Smor3), the operation element _{14b 1, 14b 2, 14b 3} , corresponding each frequency component _{S H, S M, S} by being multiplied _L, and emphasis structural elements _{B S, B M,} the image portion corresponding to the magnitude of _{B L,} the emphasis coefficient _{αm 1 (Smor1), αm 2} (Smor2), corresponding to αm ₃ (Smor3) Emphasis processing can be performed for each degree. Since the following operations and effects are the same as those of the image processing apparatus of the above-described embodiment, description thereof will be omitted.

  In the image processing apparatus according to the present embodiment as well, the frequency components of the original image signal are converted into the enhancement process based on the morphological signal and the original image signal value as shown in FIG. It is also possible to adopt a configuration in which an enhancement process is combined with an enhancement process based on the above.

  In addition, the image processing method and apparatus of the present invention can adopt a configuration in which the first conversion table is unique in the configuration having a plurality of morphological filters shown in FIGS. In this case, the configuration of the image processing apparatus shown in FIG. 6 is reversed, and the enhancement coefficient corresponding to each frequency component can be adjusted by a plurality of morphological filters and a single first conversion table. .

1 is a block diagram showing a first embodiment of an image processing apparatus of the present invention. Function graph representing the first conversion table Function graph representing the second conversion table The block diagram which shows 2nd Embodiment of the image processing apparatus of this invention. The block diagram which shows 3rd Embodiment of the image processing apparatus of this invention. The block diagram which shows 4th Embodiment of the image processing apparatus of this invention. Graph of functions representing first conversion table (A) High-frequency conversion table, (B) Medium-frequency conversion table, (C) Low-frequency conversion table Explanatory drawing which shows the effect | action of the image processing method and apparatus of this invention, (A) Original image data, (B) 2nd processed image data The block diagram which shows 5th Embodiment of the image processing apparatus of this invention. The block diagram which shows 6th Embodiment of the image processing apparatus of this invention. The block diagram which shows 7th Embodiment of the image processing apparatus of this invention. Diagram explaining the basic operation of morphological operations A diagram showing structural elements Bi (i = 1, 2,..., M; M = 4) in the morphological filter Explanatory drawing showing skeleton processing Concentration distribution chart for specifically explaining processing by morphological operation Explanatory drawing which shows the effect | action of the image processing method and apparatus of this invention, (A) Original image data, (B) 1st processed image data, (C) 2nd processed image data

Explanation of symbols

11, 21 Low-pass filter
12 Morphological filters
13 First conversion table
14a, 14b, 14c, 24a, 24b, 24c
23 Second conversion table Sorg Original image signal Sus First blur mask signal S 'First processed image signal Sproc Second processed image signal

Claims (6)

By applying a morphological operation using the structural element Bi and the scale factor λ to the original image signal Sorg representing the image, the image portion in which the image signal spatially varies smaller than the structural element Bi and / or the Extracting a morphological signal Smor indicating that the original image signal Sorg is a pixel corresponding to a sharp image portion,
Determining an unsharp mask signal Sus corresponding to a first predetermined spatial frequency of the original image signal Sorg;
The morphology a signal Smor to based rather emphasis coefficient .alpha.m (Smor), morphology signal values Smor is small area of emphasis coefficient .alpha.m emphasis coefficient region morphology signal values Smor is greater than (Smor) αm (Smor) large with emphasis coefficient .alpha.m (Smor) as it is subjected to a emphasis processing in accordance with the following equation (1) with respect to the original image signal, a high frequency component first emphasizing the (Sorg-Sus) Obtaining a processed image signal S ′;

Determining a non-sharp mask signal S′us corresponding to a second predetermined spatial frequency of the first processed image signal S ′;
The first processed image signal S ′ is subjected to the enhancement processing according to the following equation (2) by the enhancement coefficient β (S ′) based on the first processed image signal S ′, and the second processed image signal S ′. The processed image signal Sproc is obtained, or the first processed image signal S ′ is obtained by an enhancement coefficient β (S′us) based on the unsharp mask signal S′us of the first processed image signal S ′. An image processing method for obtaining a second processed image signal Sproc by performing an enhancement process according to the following equation (3) for:

2. The image processing method according to claim 1, wherein the second predetermined spatial frequency <the first predetermined spatial frequency. By applying a morphological operation using the structural element Bi and the scale factor λ to the original image signal Sorg representing the image, the image portion in which the image signal spatially varies smaller than the structural element Bi and / or the Extracting a morphological signal Smor indicating that the original image signal Sorg is a pixel corresponding to a sharp image portion,
The original image signal Sorg is divided into a plurality of mutually different frequency components Sn,
A plurality of different enhancement coefficients αmn (Smor) respectively corresponding to the plurality of frequency components Sn and having a larger morphological signal value Smor than the enhancement coefficient αmn (Smor) of a region where the morphological signal value Smor is small Using the enhancement coefficient αmn (Smor) having a large αmn (Smor) , the original image signal Sorg is subjected to a first enhancement process according to the following equation (4) to obtain a first processed image. Find the signal S '

A second enhancement process according to the following equation (2) is applied to the first processed image signal S ′ by the enhancement coefficient β (S ′) based on the first processed image signal S ′. The second processed image signal Sproc is obtained or the first processed image is obtained by an enhancement coefficient β (S′us) based on the unsharp mask signal S′us of the first processed image signal S ′. An image processing method for obtaining a second processed image signal Sproc by performing a second enhancement process according to the following equation (3) on a signal S ′:

In the first frequency band to be enhanced by the first enhancement process and the second frequency band to be enhanced by the second enhancement process, the second frequency band <the first frequency band. An image processing method characterized by the above. The morphology operation is an image processing method according to claim 1 or 2, characterized in that the operation shown in any one of the following formulas (7) to (12).

By applying a morphological operation using the structural element Bi and the scale factor λ to the original image signal Sorg representing the image, the image portion in which the image signal spatially varies smaller than the structural element Bi and / or the Morphological signal calculation means for extracting a morphological signal Smor indicating that the original image signal Sorg is a pixel corresponding to a sharp image portion;
First unsharp mask signal computing means for obtaining an unsharp mask signal Sus corresponding to a first predetermined spatial frequency of the original image signal Sorg;
In response to the input of the morphological signal Smor, the enhancement coefficient αm (Smor) of the region where the morphological signal value Smor is larger than the enhancement coefficient αm (Smor) of the region where the morphological signal value Smor is small is larger in accordance with the morphological signal Smor. A first conversion table for outputting an emphasis coefficient αm (Smor),
The original image signal is subjected to enhancement processing according to the following equation (1) using the enhancement coefficient αm (Smor) output from the first conversion table to emphasize the high-frequency component (Sorg-Sus). a first enhancement means for obtaining a first processed image signal S 'obtained by,

Second unsharp mask signal computing means for obtaining an unsharp mask signal S′us corresponding to a second predetermined spatial frequency of the first processed image signal S ′;
A second conversion table for receiving the input of the first processed image signal S ′ and outputting an enhancement coefficient β (S ′) corresponding to the first processed image signal S ′; and the second conversion The second processed image signal Sproc is obtained by applying the enhancement processing according to the following equation (2) to the first processed image signal S ′ using the enhancement coefficient β (S ′) output from the table. A second emphasis means, or
A second conversion table that receives the input of the unsharp mask signal S′us and outputs an enhancement coefficient β (S′us) corresponding to the unsharp mask signal S′us, and is output by the second conversion table. The second processed image signal Sproc is obtained by applying the enhancement processing according to the following equation (3) to the first processed image signal S ′ using the enhancement coefficient β (S′us). With emphasis means,

2. The image processing apparatus according to claim 1, wherein the second predetermined spatial frequency is less than the first predetermined spatial frequency. By applying a morphological operation using the structural element Bi and the scale factor λ to the original image signal Sorg representing the image, the image portion in which the image signal spatially varies smaller than the structural element Bi and / or the Morphological signal calculation means for extracting a morphological signal Smor indicating that the original image signal Sorg is a pixel corresponding to a sharp image portion;
Frequency band dividing means for dividing the original image signal Sorg into a plurality of different frequency components Sn;
In response to the input of the morphological signal Smor , the morphological signal value Smor is an enhancement coefficient αmn (Smor) corresponding to each frequency component Sn, and the morphological signal value Smor is smaller than the enhancement coefficient αmn (Smor) in a region where the morphological signal value Smor is small. large areas of emphasis coefficient αmn (Smor) outputs the emphasis coefficient αmn (Smor) such that large, different from the plurality of first conversion table to each other,
A first enhancement process is performed on the original image signal according to the following equation (4) using a plurality of enhancement coefficients αmn (Smor) respectively output from the plurality of first conversion tables. First enhancement means for obtaining a finished image signal S ′;

Unsharp mask signal computing means for obtaining an unsharp mask signal S′us corresponding to a predetermined spatial frequency of the first processed image signal S ′;
A second conversion table for receiving the input of the first processed image signal S ′ and outputting an enhancement coefficient β (S ′) corresponding to the first processed image signal S ′; and the second conversion A second processed image signal is obtained by applying a second enhancement process according to the following equation (2) to the first processed image signal S ′ by the enhancement coefficient β (S ′) output from the table. A second enhancement means to obtain Sproc, or
A second conversion table that receives the input of the unsharp mask signal S′us and outputs an enhancement coefficient β (S′us) corresponding to the unsharp mask signal S′us, and is output by the second conversion table. The second processed image signal Sproc is obtained by subjecting the first processed image signal S ′ to the second enhanced processing according to the following equation (3) using the enhanced coefficient β (S′us). A second emphasis means,

In the first frequency band to be enhanced by the first enhancement process and the second frequency band to be enhanced by the second enhancement process, the second frequency band <the first frequency band. An image processing apparatus. The image processing apparatus according to claim 4 or 5 , wherein the morphological calculation by the morphological signal calculation means is a calculation represented by any one of the following formulas (7) to (12).

Images