JP2009100936A - Image processor - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an image processor capable of suppressing production of overshoot and undershoot, maintaining the texture components of images and compressing the dynamic range of the images. <P>SOLUTION: The image processor includes: an image acquisition part for acquiring the images; a component decomposition part for decomposing brightness of respective points on the image acquired in the image acquisition part into a high frequency component and a low frequency component in a spatial frequency on the image; a component calculation part for calculating a new brightness component by maintaining the high frequency component as it is as the absolute value of the high frequency component is larger, reducing it as the absolute value of the high frequency component is smaller and adding it to the low frequency component decomposed in the component decomposition part; and a compression part for compressing the range of the brightness that the image has by changing the brightness of the respective points on the image acquired in the image acquisition part corresponding to the brightness component of the point calculated in the component calculation part. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to an image processing apparatus that compresses a brightness range of an image.

  Conventionally, in the medical field, an elongated tube (optical probe) with a mirror or imaging device attached to the tip is inserted into the subject's body, and the subject's body is photographed to remove tumors or blood clots. Endoscope systems for observation are widely used. By directly photographing the body of the subject, it is possible to grasp the color and shape of the lesion, which is difficult to understand with radiographic images, without damaging the subject externally, and it is necessary to determine the treatment policy etc. Easy information.

  By the way, in an endoscope system, since a strong light is applied to an observation location in a dark body, a difference in brightness of a subject is large, and a signal level width of the brightness of an image generated by an image sensor (image Dynamic range) increases. For this reason, if the captured image is displayed on the display monitor as it is, the dynamic range of the captured image exceeds the signal level width that can be displayed on the display monitor (the dynamic range of the display monitor), and the captured image is displayed on the display monitor. The bright and dark parts of the image are crushed. In order to prevent this, it is necessary to compress the dynamic range of the captured image to fall within the dynamic range of the display monitor before the captured image is displayed on the display monitor.

  As an image compression method for compressing the dynamic range of the image, a method of simply compressing the dynamic range of the image by gradation conversion is known. However, in this image compression method, the amplitude of the image signal decreases. The shadow of the image becomes difficult to see. In addition, a method of decomposing image brightness into a high frequency component having a high spatial frequency and a low frequency component having a low spatial frequency and compressing only the low frequency component is also used. In this image compression method, compared to a case where the image is simply compressed by gradation conversion, a decrease in the amplitude of the image signal can be suppressed, so that a fine shading change (texture) on the image can be maintained. . However, when the dynamic range of the low frequency component decreases, the high frequency component becomes relatively large and the high frequency component is emphasized, so-called overshoot or undershoot occurs at the edge portion of the image. There is a problem that appears.

Regarding this problem, Patent Document 1 describes a technique for compressing a low-frequency component up to a signal level where overshoot and undershoot are not noticeable, and compressing the image range by normal gradation conversion. Patent Document 2 describes a technique for controlling the addition amount of a high-frequency component in accordance with the pixel value of the original image and the gradient of the gradation when compressing the low-frequency component and combining it with the high-frequency component. .
JP 2005-102152 A JP-A-2005-353102

  However, with the technique described in Patent Document 1, although the occurrence of overshoot and undershoot can be suppressed, after all, image compression by normal gradation conversion is performed, so that the texture of the image is lowered. Further, in the technique described in Patent Document 2, since the processing is complicated, the time and load for compressing the range of the image increase, and overshoot and undershoot as a whole image are likely to occur. Since this is not taken into consideration, there is a risk that important textures that are less likely to cause overshoot and undershoot are also compressed. In particular, the endoscope system is required to observe subtle discoloration and irregularities such as the stomach wall, so that there is a problem that a correct diagnosis cannot be made if the texture of the image is lost.

  Such a problem is not limited to the endoscope system alone, but is a problem that occurs in the field of image processing apparatuses that compress the brightness range of an image in general.

  SUMMARY OF THE INVENTION In view of the above circumstances, an object of the present invention is to provide an image processing apparatus capable of suppressing the occurrence of overshoot and undershoot and maintaining the image texture and compressing the dynamic range of the image.

An image processing apparatus of the present invention that achieves the above object includes an image acquisition unit that acquires an image,
A component decomposition unit that decomposes the brightness of each point on the image acquired by the image acquisition unit into a high frequency component and a low frequency component at a spatial frequency on the image;
A new brightness component is calculated by adding the high-frequency component to the low-frequency component decomposed by the component decomposition unit while maintaining the original value as the absolute value of the high-frequency component increases and decreasing it as the absolute value of the high-frequency component decreases. A component calculation unit to perform,
Compression that compresses the brightness range of an image by changing the brightness of each point on the image acquired by the image acquisition unit by an amount corresponding to the brightness component of that point calculated by the component calculation unit And a section.

  Overshoot and undershoot are generated by emphasizing the high frequency component when compressing the low frequency component of the image and combining it with the high frequency component, and are likely to occur at the edge portion of the image where the high frequency component is large. There is a feature.

  According to the image processing apparatus of the present invention, at the edge portion where the absolute value of the high frequency component is large, the compression processing close to the conventional simple tone conversion is executed based on the brightness component close to the brightness of the original image. Therefore, the occurrence of overshoot and undershoot can be suppressed. Also, in the image portion where the absolute value of the high frequency component is small, the brightness range is compressed based on the brightness component close to the low frequency component, so the brightness amplitude due to the high frequency component that bears the texture of the image is the compression ratio. The texture is maintained without depending on etc.

In the image processing apparatus of the present invention, when the image is a color image in which the color of each point on the image is expressed by a combination at each point of the luminance values of the plurality of color components,
It is preferable that the compression unit compresses the range by changing the brightness by multiplying / dividing a coefficient corresponding to the brightness component in common with respect to the luminance values of the plurality of color components.

In the image processing apparatus of the present invention, when the image is a color image in which the color of each point on the image is expressed by a combination at each point of density values of a plurality of color components,
It is preferable that the compression unit compresses the range by changing the brightness by adding or subtracting a value corresponding to the brightness component in common to the density values of the plurality of color components.

  In a color image in which colors are represented by combinations of luminance values of multiple color components, the hue is maintained by multiplying or dividing the luminance values of the multiple color components by a coefficient corresponding to the brightness component in common. In a color image in which the color can be changed by combining the density values of multiple color components, the value corresponding to the brightness component is common to the density values of the multiple color components. By adjusting the brightness, the brightness can be changed while maintaining the hue. As a result, even in a color image, it is possible to suppress the influence on the color and the like and to compress the brightness range with high accuracy by a simple calculation.

  According to the present invention, the occurrence of overshoot and undershoot can be suppressed, and the dynamic range of the image can be compressed while maintaining the texture of the image.

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

  FIG. 1 is a schematic configuration diagram of an endoscope system to which an embodiment of the present invention is applied.

  An endoscope system 1 shown in FIG. 1 guides and irradiates light into the body of a subject P, generates an image signal based on the reflected light, a light source device 20 that emits light, and an optical probe. The image obtained by 10 is subjected to predetermined image processing to generate a medical image obtained by photographing the inside of the subject P, and the medical image generated by the image processing device 30 is displayed on the display monitor 41. And a display device 40 for displaying.

  The optical probe 10 includes a flexible elongated probe unit 11, an operation unit 12 that operates the probe unit 11, and a light / signal guide 13 that connects the light source device 20, the image processing device 30, and the optical probe 10. It is configured. Hereinafter, the side of the optical probe 10 that is inserted into the body of the subject P is referred to as the front end, and the opposite side of the front end is referred to as the rear end.

  The operation unit 12 includes a bending operation lever 121 for bending the probe unit 11, a freeze button 122 for capturing a still image, a color adjustment button 123 for adjusting the color of a displayed image, and the like. Is provided.

  The light / signal guide 13 includes a light guide 131 that transmits light and a signal line 132 that transmits signals. The light guide 131 has a rear end connected to the light source device 20, guides the light emitted from the light source device 20 to the probe unit 11, and guides the light from the irradiation window 11 a provided at the tip of the probe unit 11. Irradiate toward P. The signal line 132 has a CCD 133 attached to the front end, and the rear end side is connected to the image processing apparatus 30. The reflected light reflected from the inside of the subject P by the light irradiated from the irradiation window 11a of the light guide 131 is collected by the optical member 134 provided at the tip of the probe unit 11, received by the CCD 133, and reflected light. A photographed image representing is generated. The CCD 133 has a plurality of light receiving portions arranged side by side, and light is received by each of the plurality of light receiving portions, whereby image data in which an image is expressed by a plurality of pixels is generated. The CCD 133 has a plurality of light receiving portions arranged side by side, and when a light is received by each of the plurality of light receiving portions, a captured image represented by a plurality of pixels is generated. In the present embodiment, a color filter (see FIG. 2) in which R, G, and B colors are arranged in a regular color pattern is attached to the CCD 133 at positions corresponding to the plurality of light receiving units. When the light that has passed through is received by the CCD 133, a color mosaic image in which pixels of R, G, and B colors are arranged in the same color pattern as the color pattern of the color filter is generated. The generated color mosaic image is transmitted to the image processing device 30 through the signal line 132, and the image processing device 30 performs predetermined image processing.

  FIG. 2 is a schematic functional block diagram of the endoscope system 1.

  In FIG. 2, the display monitor 41 and the operation unit 12 of the optical probe 10 are not shown, and only main elements related to the generation of the image signal are shown.

  The light source device 20 also shown in FIG. 1 emits white light and is controlled by the overall control unit 360 of the image processing device 30.

  In addition to the CCD 133 also shown in FIG. 1, the optical probe 10 digitally receives analog image signals generated by the color filter 140 in which each color of R, G, B is arranged in a mosaic pattern with a regular color pattern, and the CCD 133. Are provided with an A / D conversion unit 150 that converts the image signal into an image signal, an imaging control unit 160 that controls processing in various elements in the optical probe 10, and the like. In the present embodiment, the optical probe 10 generates RGB color space image data in which the color of the image is represented by R, G, and B colors and the components of the R, G, and B colors are linear in luminance. It will be explained as a thing.

  The image processing apparatus 30 includes a storage unit 300 that stores a still image generated by pressing the freeze button 122, various parameters, and the like, and gain correction that corrects the gain of the image sent from the optical probe 10. Each pixel of the color mosaic image generated by the optical probe 10 includes: a unit 310; a spectral correction unit 320 that corrects spectral characteristics of the optical probe 10 including the CCD 133; a gamma correction unit 340 that performs gradation correction processing on the image; By interpolating other color components (for example, B and G colors) except for the color component (for example, R color) having the surrounding pixels, each pixel is a mixed color of R, G, and B colors. A display processing unit 41 that generates a represented color image and a signal level width (dynamic range of the color image) of the color image generated by the synchronization processing unit 350. A display adjustment unit 360 that compresses the signal level width within the displayable signal level range (dynamic range of the display monitor 41), an overall control unit 390 that controls processing of the optical probe 10 and the entire image processing apparatus 30, and the like are provided. Yes.

  The display adjustment unit 360 generates a luminance component Y based on an RGB image in which colors are expressed in three RGB colors, and an LPF (low pass filter) that extracts a low frequency component having a low spatial frequency from the luminance component Y. Unit 262, a Y composition unit 363 that generates a composite component by adding a high frequency component to the low frequency component of luminance component Y, a correction amount calculation that calculates a correction amount for compressing the dynamic range of the RGB image based on the composite component 364 and a correction unit 365 that compresses the range of the RGB image. The Y generation unit 361 corresponds to an example of an image acquisition unit according to the present invention, and the low-pass filter unit 262 corresponds to an example of a component decomposition unit according to the present invention. The Y composition unit 363 is an example of the component calculation unit according to the present invention, and the combination of the correction amount calculation unit 364 and the correction unit 365 corresponds to an example of the compression unit according to the present invention.

  FIG. 3 is a flowchart showing a flow of a series of processes until a captured image is displayed on the display monitor 41.

  In the following, a flow of a series of processes from when the subject is imaged until the captured image is displayed on the display monitor 41 will be described according to the flowchart.

  First, the optical probe 10 having a size suitable for the observation location of the subject is mounted on the light source device 20 and the image processing device 30, and the optical probe 10 is inserted into the body of the subject P. The light emitted from the light source device 20 is guided to the tip of the optical probe 10 by the light guide 131 and is irradiated into the body of the subject P from the irradiation window 11a. The reflected light obtained by reflecting the light emitted from the light source device 20 in the body of the subject P passes through the color filter 140 and is received by the CCD 133 to generate a captured image (step S1 in FIG. 3). The generated captured image is digitized by the A / D converter 150 and then transmitted to the image processing apparatus 30 through the signal line 132.

  In the image processing device 30, various image processing for correcting the color of the image is performed on the captured image (step S2 in FIG. 3). First, gain correction is performed in the gain correction unit 310, spectral correction processing is performed in the spectral correction unit 320, and tone correction processing is performed in the gamma correction unit 340, which is then transmitted to the synchronization processing unit 350. In the synchronization processing unit 350, the captured image that is a mosaic color image is subjected to synchronization processing, and each pixel is converted into an RGB image expressed by a mixed color of R, G, and B colors. The converted RGB image is transmitted to the display adjustment unit 360.

  The display adjustment unit 360 compresses the dynamic range of the RGB image so that the dynamic range of the RGB image falls within the dynamic range of the display monitor 41. In the present embodiment, the dynamic range of the entire RGB image is compressed by executing the process of correcting the brightness (signal value) of each of the plurality of pixels constituting the RGB image.

  When the RGB image is input to the display adjustment unit 360, the Y generation unit 361 generates a luminance component Y based on the RGB image (step S3 in FIG. 3). In the present embodiment, the luminance component Y at each point on the image is calculated based on the following formula (1).

Y = 0.299 × R + 0.587 × G + 0.114 × B (1)
The generated luminance component Y is transmitted to the low-pass filter unit 362.

  In the low-pass filter unit 362, a low-frequency component having a low spatial frequency on the image is extracted from the luminance component Y, and a high-frequency component obtained by removing the low-frequency component from the original luminance component Y is generated (step in FIG. 3). S4). The generated low-frequency component is a so-called blurred image, and in the present embodiment, a filter characteristic that reliably separates a texture component necessary for diagnosis as a high-frequency component in a standard medical image is prepared in advance. A low frequency component and a high frequency component are generated based on the filter characteristics. In this embodiment, an FIR filter is applied as the low-pass filter unit 362. However, the component decomposition unit referred to in the present invention applies an IIR filter having a circuit scale smaller than that of the FIR filter, or uses a reduced image. Also good. The generated low frequency component and high frequency component are transmitted to the Y synthesis unit 363.

In the Y synthesis unit 363, the high frequency component is added to the low frequency component transmitted from the low pass filter unit 362 to generate a synthesis component (step S5 in FIG. 3). In the present embodiment, based on the original luminance component Y and the low-frequency component Y _L, the synthesized signal Y _mix is calculated according to the following equation (2).

Y _mix = Y _L + F (Y−Y _L ) (2)
The term (Y−Y _L ) in Expression (2) corresponds to the high frequency component transmitted from the low pass filter unit 362.

  FIG. 4 is a diagram illustrating the function F in the equation (2).

As shown in FIG. 4, the function F is F = 0 when the absolute value of the high-frequency component (Y−Y _L ) of the luminance component Y is equal to or less than the specified value “t”, and the high-frequency component ( If the absolute value of _{Y-Y L)} is greater than the specified value "t" is a linear function of slope a represented by _{F = a × (Y-Y} L) ± t. That is, from equation (2)
When | Y−Y _L | ≦ t, Y _mix = Y _L (3)
When | Y−Y _L |> t, Y _mix = Y _L + a (Y−Y _L ) ± t (4)
Is required. In this embodiment, when the slope a = 1 and the high frequency component is sufficiently large,
When | Y−Y _L | >> t, Y _mix = Y ± t≈Y (5)
It becomes. The specified value “t” is sufficiently smaller than the high frequency component in the edge portion in the standard medical image, and an empirical value is applied so that the high frequency component in the texture portion is included in the range of −t ≦ x ≦ t. . That is, at the edge portion where the absolute value of the high frequency component is large and overshoot or undershoot is likely to occur, a composite component Y _mix close to the original luminance component Y is generated, and the absolute value of the high frequency component is equal to or less than the specified value “t”. In an image portion including an important texture component, the composite component Y _mix = low frequency component Y _L.

The generated composite component Y _mix is transmitted to the correction amount calculation unit 364.

The correction amount calculation unit 364 calculates a correction amount for correcting the brightness of each pixel for compressing the range of the RGB image based on the composite component Y _mix (step S6 in FIG. 3). In the present embodiment, a correction table in which the size of the composite component Y _mix and the correction amount are associated is prepared, and the correction amount calculation unit 364 refers to the correction table based on the composite component Y _mix , A correction amount is acquired.

  FIG. 5 is a diagram illustrating an example of the correction table.

In FIG. 5, the horizontal axis is associated with the magnitude of the composite component Y _mix , and the vertical axis is associated with the correction amount. As shown in FIG. 5, the smaller the composite component Y _mix (that is, the darker the pixel), the larger the correction amount is in the positive direction, and the pixel is corrected to become brighter. Also, synthetic components Y _mix is large (i.e., the pixel is bright), the more the correction amount is increased in the negative direction, is corrected so that the pixel is dark.

In the correction amount calculation unit 364, the correction amount r corresponding to the composite component Y _mix calculated by the Y composition unit 363 is acquired, and the acquired correction amount r is transmitted to the correction unit 365.

  In the correction unit 365, the brightness of each pixel constituting the RGB image transmitted from the synchronization processing unit 350 is corrected with the correction amount r calculated by the correction amount calculation unit 364, so that the entire RGB image is corrected. The dynamic range is compressed (step S7 in FIG. 3). In this embodiment, an RGB image in which the color of each pixel is expressed by a combination of luminance values of R, G, and B colors is generated, and the correction unit 365 has the same correction amount for each of the R, G, and B components. The dynamic range of the RGB image is compressed by adding r.

Here, as described above, the composite component Y _mix is close to the original luminance signal Y in the image portion where the absolute value of the high frequency component is large, and the composite component Y _mix = low in the image portion where the absolute value of the high frequency component is small. It has a frequency component _{Y L.} For this reason, at the edge portion where the absolute value of the high-frequency component is large, image compression processing similar to normal gradation conversion is performed based on the original luminance signal Y, and the occurrence of overshoot and undershoot can be suppressed. it can. In addition, in the texture portion where the absolute value of the high frequency component is equal to or less than the specified value t, the brightness is corrected on the average based on the low frequency component, so that it is possible to maintain the shadow and subtle unevenness of the image. .

  The RGB image in which the dynamic range is compressed is transmitted to the display device 40 and displayed on the display monitor 41 (step S8 in FIG. 3).

  As described above, according to the present embodiment, the occurrence of overshoot and undershoot can be suppressed, the image texture can be maintained and the image range can be compressed, and a high-quality medical image useful for diagnosis can be displayed. can do.

  Above, description of 1st Embodiment of this invention is complete | finished and 2nd Embodiment of this invention is described. Since the second embodiment of the present invention has the same configuration as that of the first embodiment shown in FIG. 2, FIG. 2 is also used in the description of the second embodiment, and is different from the first embodiment. Only explained.

  In the present embodiment, the optical probe 10 generates an RGB image in which each pixel constituting the image is expressed by a combination of density values of R, G, and B colors.

  FIG. 6 is a diagram illustrating an example of a correction table used by the correction amount calculation unit 364 of the present embodiment.

In FIG. 6, the horizontal axis is associated with the magnitude of the composite component Y _mix , and the vertical axis is associated with the correction coefficient r ′ that is multiplied by each pixel value of the original image. In the present embodiment, as the composite component Y _mix is smaller (that is, the pixel is darker), the correction coefficient r ′ is larger than “1”, and the brightness of the pixel after correction is brighter than the pixel before correction. It is corrected to. Further, as the composite component Y _mix is larger (that is, the pixel is brighter), the correction coefficient r ′ is smaller than “1”, and correction is performed so that the pixel becomes darker.

  In the correction unit 365, the dynamic range of the entire RGB image is compressed by multiplying each R, G, B component of each of the plurality of pixels constituting the density linear RGB image by the correction coefficient r ′.

  Thus, in a color image in which the color of each pixel is expressed by a combination of density values of a plurality of colors, the dynamic range of the image is increased by multiplying each color component by a common correction coefficient r ′. Can be compressed.

  Here, the example in which the contrast evaluation value is calculated every time the frame image is captured has been described above. However, the endoscope apparatus of the present invention stores the captured frame image, and the user The calculation of the contrast evaluation value and the determination of the frame image having the maximum value may be executed at the time when the instruction from is received.

  In the above description, an example in which an image processing apparatus that compresses the range of image brightness is applied to an endoscope system has been described. However, the image processing apparatus of the present invention is applied to a normal digital camera, an image display apparatus, and the like. May be.

  In the above description, an example of compressing the brightness range of a color image in which the color of the image is expressed by a combination of R, G, and B colors has been described. A CMYK image in which the color of the image is expressed by a combination of the colors, Y, and K may be used.

In the above description, the example in which the Y component is generated from the RGB image using the equation (1) for calculating a color close to the sense of the human eye has been described.
Y = R + G + G + B (6)
The Y component may be generated using such as. This equation (6) is widely used to approximate the Y component using an RGB image.

1 is a schematic configuration diagram of an endoscope apparatus according to a first embodiment of the present invention. It is a functional block diagram of an endoscope apparatus. It is a flowchart figure which shows the flow of a series of processes until the image | photographed image is displayed on a display monitor. It is a figure which shows the function F in Formula (2). It is a figure which shows an example of the correction table in 1st Embodiment. It is a figure which shows an example of the correction table in 2nd Embodiment.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Endoscope apparatus 10 Optical probe 20 Light source apparatus 30 Image processing apparatus 40 Display apparatus 11 Probe part 12 Operation part 13 Light / signal guide 121 Bending operation lever 122 Freeze button 123 Selection button 131 Light guide 132 Signal 133 CCD
140 Color Filter 150 A / D Conversion Unit 160 Imaging Control Unit 300 Storage Unit 310 Gain Correction Unit 320 Spectral Correction Unit 340 Gamma Correction Unit 350 Synchronization Processing Unit 360 Display Adjustment Unit 361 Y Generation Unit 362 Low-Pass Filter Unit 363 Y Composition Unit 364 Correction amount calculation unit 365 Correction unit 390 Overall control unit

Claims (3)

An image acquisition unit for acquiring images;
A component decomposition unit that decomposes the brightness of each point on the image acquired by the image acquisition unit into a high-frequency component and a low-frequency component at a spatial frequency on the image;
A new brightness is obtained by adding the high-frequency component to the low-frequency component decomposed by the component decomposition unit while keeping the original as the absolute value of the high-frequency component is larger and decreasing the absolute value of the high-frequency component as it is smaller. A component calculation unit for calculating a component;
The brightness range of the image is changed by changing the brightness of each point on the image acquired by the image acquisition unit by an amount corresponding to the brightness component of the point calculated by the component calculation unit. An image processing apparatus comprising: a compression unit that compresses the image. The image is a color image in which the color of each point on the image is expressed by a combination at each point of the luminance values of the plurality of color components,
The compression unit is configured to compress the range by changing the brightness by multiplying / multiplying the coefficient corresponding to the brightness component in common to the luminance values of the plurality of color components. The image processing apparatus according to claim 1. The image is a color image in which the color of each point on the image is expressed by a combination at each point of density values of a plurality of color components,
The compression unit compresses a range by changing brightness by adding or subtracting a value corresponding to the brightness component in common to density values of the plurality of color components. The image processing apparatus according to claim 1.

Images