JPH0935056A - Image processor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To output an original image and an emphasized image without making constitution complex and perform an emphasizing process without losing delicate information in the original image. SOLUTION: This image processor is provided with a living body function information calculating means 1 which calculates information regarding a living body function from the inputted original image, a filter processing means 2 which filters the image of the calculated living body function information, a delay means 3 for timing adjustment which is provided in parallel to those living body function information calculating means 1 and filter processing means 2, and an emphasizing process means 4 which emphasizes the original image inputted through the delay means 3 according to the output result of the filter processing means 2, and the emphasizing process is performed for the original image according to an emphasization coefficient calculated as a result of the filter processing for the image regarding the living body function information, so that an observed image having the living body function information emphasized is obtained.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing apparatus for performing image processing such as image enhancement on a biometric image.

[0002]

2. Description of the Related Art Conventionally, in the field of endoscopes, a subject is irradiated with light that has passed through a narrow band filter or a normally visible RGB filter in a time series, and the amount of hemoglobin or hemoglobin oxygen saturation is determined from the obtained image. There are known methods for obtaining information on biological functions such as degree. Recently, the original image is emphasized so that the observed image can be easily identified based on the biological function information.

The emphasizing process includes emphasizing the biofunction information in each pixel centering on the average value of the biofunction information obtained from the original image in one screen. However, in order to emphasize the biofunction information that changes minutely around the average value, it is necessary to increase the degree of emphasis, but before the emphasis, the biofunction information has a large difference from the average value. Causes overexposure and color collapse in the image. Therefore, in order to prevent the inconvenience and clearly observe the biological function information that is minutely changed around the average value, contour enhancement processing or the like is performed.

In addition, images with many dark areas such as luminal organs are
There is a lot of noise. A noise removal filter process such as a median filter is applied to such an image to remove noise in the image.

When processing such as noise removal or contour enhancement on a processed image as described above, conventionally, it is customary to perform the processing on the original image before and after the enhancement processing for calculating the biological function information. Met.

Here, FIG. 17 and FIG. 18 show an example of the configuration of a conventional image processing apparatus that performs a filtering process such as noise removal together with a process of emphasizing biological function information.

In this example, after filtering processing such as noise removal is performed on the input image by the filter processing unit 51,
The enhancement processing section 52 is configured to perform enhancement processing of biological function information in each pixel and output the processed image.

In the configuration in which the filter processing section 51 is arranged in series with the emphasis processing section 52 in this way, if it is desired to output the original image without performing these image processing, the emphasis processing section 5
2 is turned off, and the coefficient of the filter processing unit 51 is rewritten to the coefficient of the through output as shown in FIG. 17, or the delay 53 for timing adjustment is used as shown in FIG. It was necessary to provide another path and switch by the switch 54.

[0009]

As described above, in the configuration of the conventional image processing apparatus, when it is desired to output the original image as it is without performing the filter processing or the emphasis processing, the filter coefficient of the filter processing unit is output through. It was necessary to rewrite to the coefficient of 1 or to switch by providing another path that does not perform filter processing. Therefore, a circuit such as a memory for timing adjustment is required only for rewriting the coefficient of the filter and for outputting the original image. Providing these increases the scale of the entire image processing apparatus and complicates the configuration. It becomes one of the factors.

Further, if the noise removal filter process is directly performed on the original image before the biofunction information is emphasized, the entire image looks like a blur, and the fine mucous membrane in the original image. There was a case that sacrificing the structure pattern of.

The present invention has been made in view of these circumstances, and it is possible to output an original image and an emphasized image without complicating the structure, and without losing fine information in the original image. An object of the present invention is to provide an image processing device capable of performing emphasis processing.

[0012]

An image processing apparatus according to the present invention is an image processing apparatus for performing image processing on an input original image, and biological function information calculating means for calculating information on biological function from the original image. And filter processing means for filtering the image relating to the biological function information, and enhancement processing means for performing enhancement processing on the original image based on the filter processing result. It is possible to output the original image and the emphasized image without making it complicated, and it is possible to perform the emphasis process without losing fine information in the original image.

[0013]

Embodiments of the present invention will be described below with reference to the drawings. First, the schematic configuration of the present invention will be described with reference to FIG. The image processing apparatus includes a biological function information calculating unit 1 that calculates information related to a biological function from an input original image, and a filter processing unit that filters an image of the biological function information calculated by the biological function information calculating unit 1. 2, a delay means 3 provided in parallel with the biological function information calculating means 1 and the filter processing means 2 for timing adjustment, and input via the delay means 3 based on the output result of the filter processing means 2. Enhancement processing means 4 for performing enhancement processing on the original image to be processed.

The operation in the configuration of FIG. 1 will be described by taking the case of processing an endoscopic image as an example. First, in the biological function information calculating means 1, from the original image obtained by the endoscope,
The concentration of hemoglobin that governs the colors in the image, the oxygen saturation, and the like are calculated for each pixel to obtain an image relating to hemoglobin, which is one of biological function information. Next, the filter processing unit 2 performs filter processing such as noise removal on the image of the biological function information. After that, based on the result of the filter processing, processing such as color enhancement is performed on the original image by the enhancement processing means 4. At this time, the delay unit 3 holds the original image for a predetermined period,
Adjust the timing with the filtered output.

With this configuration, even when it is desired to directly output the original image as it is without performing the emphasis process,
Since it is possible to output the original image by turning off only the enhancement processing means 4, the means for rewriting the coefficient of the filter as in the prior art, the path bypassing the filter processing section, and the timing for outputting the original image. The circuit scale can be reduced without the need for adjusting means.

Further, since the filter processing such as noise removal is performed on the biological function information image and the original image is not directly subjected to the filter processing, a problem such as blurring of the entire image is eliminated, and the original image is eliminated. It is possible to output a clear image while maintaining the fine information included in the image. Further, by smoothing the error of the hemoglobin amount calculated by the noise, it is possible to reduce problems such as emphasizing spike noise of the original image.

Next, a specific configuration example of the image processing apparatus will be shown below as an embodiment.

2 to 4 relate to the first embodiment of the present invention, FIG. 2 is an overall configuration diagram of an endoscope apparatus including an image processing apparatus, and FIG. 3 is a block diagram showing an internal configuration of the image processing apparatus. FIG. 4 is an explanatory diagram showing an example of filter coefficients of a smoothing filter provided in the filter processing unit.

As shown in FIG. 2, the image processing apparatus of this embodiment is provided in an endoscope apparatus including an electronic endoscope 11. The electronic endoscope 11 has an elongated, for example, movable insertion portion 12, and a large-diameter operation portion 13 is connected to the rear end of the insertion portion 12. A movable universal cord 14 extends laterally from the rear end of the operation portion 13, and the connector 1 is attached to the tip of the universal cord 14.
5 are provided.

The electronic endoscope 11 is connected via a connector 15 to a video processor 16 having a light source section. Further, an image processing device 18 is connected to the subsequent stage of the video processor 16, and a monitor 17 is connected to the subsequent stage.

At the distal end side of the insertion portion 12 of the electronic endoscope 11, a rigid distal end portion 19 and a bending portion 20 adjacent to the distal end portion 19 and capable of being curved rearward are sequentially provided. By rotating the bending operation knob 21 provided on the operation portion 13, the bending portion 20 can be bent in the horizontal direction or the vertical direction. In addition, the operation unit 1
An insertion port 22 that communicates with a treatment tool channel provided in the insertion portion 12 is provided at 3.

Next, the structure of the image processing device 18 will be described with reference to FIG. In the image processing device 18, R (red), G (green), and B sent from the video processor 16 are sent.
A / D converters 23a to convert each color signal of (blue) (hereinafter, referred to as RGB signal) into a digital signal
23c is provided. A / D converters 23a-2
Inverse γ correction units 24a to 24c, which are lookup tables (LUTs), are provided in the subsequent stage of 3c, and inverse γ correction conversion is performed on the RGB signals converted into digital signals. Outputs of the inverse γ correction units 24a to 24c are branched into two paths, and a frame memory 25 for timing adjustment and logarithmic conversion units 26a to 26c formed of lookup tables are connected to the subsequent stages.

A matrix circuit 27 is provided at the subsequent stage of the logarithmic conversion units 26a to 26c, and the amount of pigment representing the amount of hemoglobin as the biological function information in the logarithmically converted image signal is calculated. In the subsequent stage,
A filter processing unit 28 having a smoothing filter for removing noise is provided, and an emphasis amount calculation unit 29 including an ROM and an average value calculation unit 30 are provided at the subsequent stage. In the filter processing unit 28, the matrix circuit 27
After the output signal is filtered to remove noise, the average value calculation unit 30 calculates the average value of the pigment amount, and the average value signal is transferred to the enhancement amount calculation unit 29 to perform the biological function. The amount of emphasis of information is calculated. At this time, the amount of emphasis can be adjusted by the emphasis level signal sent from the front panel or the like. An emphasis coefficient conversion unit 31 including a ROM is provided in the subsequent stage of the emphasis amount calculation unit 29, and the RGB amount of the original image is emphasized by converting the emphasis amount calculated by the emphasis amount calculation unit 29. An emphasis coefficient for processing is required.

At the subsequent stage of the emphasis coefficient conversion unit 31, image conversion units 32a to 32c including a ROM are provided.
The output signal line of the frame memory 25 is connected to the input terminals of the image conversion units 32a to 32c together with the output signal line of the enhancement coefficient conversion unit 31. Image conversion units 32a to 32
D / A converters 33a to 33c are provided in the subsequent stage of c, and their output terminals are connected to the monitor 17. In the image conversion units 32a to 32c, the inverse γ correction unit 24a to
The timing-adjusted RGB signals sent from the frame 24c through the frame memory 25 are subjected to the emphasis processing based on the emphasis coefficient from the emphasis coefficient conversion unit 31, and after the γ correction is performed, D / A converters 33a to 33
It is adapted to be converted into an analog signal by c and output to the monitor 17.

Next, the operation of the image processing apparatus 18 of this embodiment will be described in more detail.

R sent from the video processor 16
The GB signal is converted into digital signals by the A / D converters 23a to 23c, and then the inverse γ correction units 24a to 24c.
Inverse γ correction conversion is performed by and is sent to the frame memory 25 and the logarithmic conversion units 26a to 26c. Then, in the logarithmic conversion units 26a to 26c, after performing logarithmic conversion for each of the R, G, and B color signals, the matrix circuit 2
The subtraction process is performed at 7, and the hemoglobin amount IHb represented by the following equation is calculated for each pixel. As a result, a distribution image of the amount of hemoglobin is obtained.

IHb = log R−log G (1) The calculated hemoglobin amount is input to the filter processing unit 28, and a smoothing filter for removing noise having a filter coefficient as shown in FIG. 4 is used. Filter processing is performed. By this filter processing, noise included in the hemoglobin amount distribution image output from the matrix circuit 27 is reduced, and the noise component of the original image is not emphasized.

The filter used in the filter processing section 28 is not limited to the smoothing filter, and a filter such as a median filter for suppressing noise components may be used, and these noise components are suppressed. Instead of the filter, the filter processing may be performed using a filter that emphasizes the biological function information such as a contour emphasis filter.

The amount of hemoglobin for each smoothed pixel is
It is sent to the average value calculator 30 and the emphasis amount calculator 29. In the average value calculation unit 30, the hemoglobin amount calculated for each pixel and the number of pixels are added for one field, respectively, and the added hemoglobin amount is divided by the added number of pixels. By this processing, the average value of the hemoglobin amount in one field in the hemoglobin amount distribution image is obtained.

Then, the calculated average value data is sent to the emphasis amount calculation unit 29, and in the emphasis amount calculation unit 29, the hemoglobin amount for each pixel input from the matrix circuit 27 and the average value calculation unit 30 input 1 Based on the average value data in the field, find the difference from the average value of the hemoglobin amount for each pixel, emphasize this difference with a variable emphasis level input from the front panel, etc. By adding, the emphasized hemoglobin amount (IHb ′) is calculated for each pixel.

The emphasized hemoglobin amount IHb ′ calculated by the emphasis amount calculator 29 is sent to the emphasis coefficient converter 31. The enhancement coefficient conversion unit 31 performs data conversion represented by the following equation to obtain an enhancement coefficient for enhancing the RGB signal of the original image.

Αr (i, j) = εr / (εg−εr) · {IHb (i, j) −IHb ′ (i, j)} αg (i, j) = εg / (εg−εr) · { IHb (i, j) -IHb '(i, j)} [alpha] b (i, j) = [epsilon] b / ([epsilon] g- [epsilon] r) * {IHb (i, j) -IHb' (i, j)} (2) Here, i, j are coordinates of each pixel, εr, εg, εb are extinction coefficients of hemoglobin in respective wavelength bands of R, G, B, αr (i, j), αg (i, j), αb ( i, j)
Is R, G, B at the position (i, j) of each pixel.
It is an emphasis coefficient of each color image.

The data of the emphasis coefficient calculated by the equation (2) is transferred to the image conversion units 32a to 32c. The image conversion units 32a to 32c perform data conversion based on the following equations from the RGB signals whose timings have been adjusted by the frame memory 25 and the enhancement coefficients αr, αg, and αb calculated by the enhancement coefficient conversion unit 31 to obtain the biological data. Performs emphasis processing related to functional information.

R ′ (i, j) = R (i, j) · 10 ^ {αr (i, j)} G ′ (i, j) = G (i, j) · 10 ^ {αg (i, j)} B ′ (i, j) = B (i, j) · 10 ^ {α b (i, j)} (3) where R ′ (i, j) and G ′ (i, j) , B '(i, j) are the newly determined brightness levels of R, G, B at the position (i, j) of each pixel. In the above equation, "^" represents exponentiation.

By the conversion process, the image conversion unit 32a
At 32c, the original image is converted into an image subjected to emphasis processing.

If the original image is not desired to be emphasized, the data for directly outputting the original image is directly stored in the frame memory 25. Therefore, for example, by switching the input address of the ROM. It is possible to switch between through output and enhancement processing output in the enhancement coefficient conversion unit 31.

Therefore, there is no need to change the means for calculating the emphasis coefficient in order to display the original image, and if the filter coefficient of the filter processing unit is fixed to one type of filter coefficient, Rewriting means can also be omitted.

Further, in the image conversion units 32a to 32c, γ correction is also performed after the emphasis processing. The R ', G', and B'color image signals created here are converted into analog signals by the D / A converters 33a to 33c, and then transferred to the monitor 17, and the emphasized image is displayed on the monitor 17. It The emphasized image is displayed on the monitor 17, recorded on an image filing device, and output to a printer, a slide photographing device, or the like.

According to the present embodiment, by emphasizing the biological function information (hemoglobin amount) in the image obtained by the endoscope, the portion with much blood has more blood and the portion with less blood has more. Since it is possible to display as if there is less blood, it is possible to emphasize more effectively and naturally the areas such as blood vessels and lesions where the blood flow state is changing from the surrounding mucosa. It will be possible. In addition, since the noise removal filter processing is performed on the hemoglobin amount distribution image calculated from the original image, which is the basis of the emphasis, the entire image is blurred, and the fine pattern included in the original image such as the structure pattern of the mucous membrane is fine. A clear image can be displayed without losing information or emphasizing spike noise of the original image.

Further, when the emphasis processing is not performed on the original image by performing the filtering processing such as noise removal on the biofunction information image, not on the original image, the emphasis processing is performed by the emphasis processing means. Since it is possible to output the original image as a through image simply by turning it off, the noise removal filter coefficient can be rewritten to a through image coefficient, or a separate path that does not perform the filtering process just to display the original image can be provided for timing. It is possible to prevent the configuration from becoming complicated without the need to adjust.

In this embodiment, the data conversion is performed by using the ROM or the matrix circuit.
Instead of the OM or the matrix circuit, a RAM or a field programmable gate array (FPGA) may be used for the processing. Further, the circuit of the present embodiment may be built in a video processor, an image filing device (not shown), or the like, instead of being independently configured as an image processing device.

In the present embodiment, the RGB image is emphasized based on the amount of hemoglobin pigment, which is the element of the biological function information that is dominant in the endoscopic image, but instead of this, the hemoglobin oxygen saturation is used. Alternatively, emphasis processing may be performed based on information such as a dye or a dye.

Hereinafter, as another embodiment, an example in which the internal configuration of the image processing apparatus provided in the endoscope apparatus of the above-described first embodiment is changed will be shown.

5 and 6 relate to the second embodiment of the present invention. FIG. 5 is a block diagram showing the internal structure of the image processing apparatus, and FIG. 6 is a filter coefficient of a high-frequency emphasis filter provided in the filter processing section. It is explanatory drawing which shows an example.

In this embodiment, an image processing device 35 having a different internal structure is provided in the endoscope device having the same structure as that of the first embodiment.

As shown in FIG. 5, the image processing device 35 is provided with A / D converters 36a to 36c for converting the RGB signals sent from the video processor 16 into digital signals. A / D converter 3
Inverse γ correction units 37a to 37c, which are lookup tables, are provided in the subsequent stages of 6a to 36c, and the inverse γ correction conversion is performed on the RGB signals converted into digital signals, and the outputs are output through two paths. And is sent to the timing adjustment memory 38 and the difference circuit 39.

A filter processing section 40 having a high-frequency emphasis filter is provided at the subsequent stage of the difference circuit 39. After the difference between the RGB color signals is obtained by the difference circuit 39, the filter processing is performed to obtain a high-level signal. Area emphasis is performed. Further, an emphasis coefficient conversion unit 41 including a ROM or the like is provided in the subsequent stage, and an emphasis coefficient for emphasis processing of the RGB signal of the original image is obtained.

Image conversion units 42a to 42c including a ROM and the like are provided at the subsequent stage of the emphasis coefficient conversion unit 41.
Based on the timing-adjusted RGB signals sent from the inverse γ correction units 37 a to 37 c through the memory 38 and the enhancement coefficient from the enhancement coefficient conversion unit 41, the enhancement process is performed on the RGB signals. . Image converter 4
D / A converters 43a to 43 are provided in the subsequent stages of 2a to 42c.
c is provided, and the outputs of the image conversion units 42a to 42c are converted into analog signals and output to the monitor 17. The output image of the image processing device 35 is output to the monitor 17, transferred to an image filing device (not shown) or the like for recording, or output to a printer, a slide photographing device, or the like.

Next, the operation of the image processing apparatus 35 of this embodiment will be described in more detail.

In this embodiment, the A / D converter 36a
Up to 36c, the RGB signals sent from the video processor 16 are converted into digital signals, and then inverse γ correction units 37a to 37c perform inverse γ correction conversion.
Then, the data-converted signal is sent to the memory 38, timing-adjusted with the processed signal via the difference circuit 39, etc., and then sent to the image conversion units 42a to 42c.

The signals whose data have been converted by the inverse γ correction units 37a to 37c are also sent to the difference circuit 39. Then, in the difference circuit 39, the difference calculation is performed by selecting from the R, G, and B color signals according to the light absorption characteristics of the dye representing the biological function information. In the present embodiment, since it is desired to calculate a correlated value for the hemoglobin pigment, the R signal and G
The difference between the signal or the R signal and the B signal is calculated. That is, this difference process can be regarded as a simplification of the hemoglobin amount calculation process in the first embodiment.

The calculated difference value is used as a filter processing unit 40.
Is input to a high-frequency emphasis filter having a filter coefficient as shown in FIG. By this filter processing, it is possible to emphasize the portion where the difference between the R signal and the G signal changes abruptly.

The high-frequency emphasis filter is not limited to the above-mentioned coefficients, and any filter coefficient that emphasizes high-frequency components may be used. Further, the filter used in the filter processing unit 40 is not limited to the high-frequency emphasis filter,
A mid-range emphasis filter may be used so that the structure pattern of the mucous membrane can be effectively emphasized, and instead of the emphasis filter, a smoothing filter for removing noise as in the first embodiment is used for filtering. Processing may be performed.

The high-frequency emphasized difference value is sent to the emphasis coefficient conversion section 41, and in the emphasis coefficient conversion section 41, the same processing as in the emphasis amount calculation section 29 and the emphasis coefficient conversion section 31 of the first embodiment is performed. Data conversion processing is performed to obtain the emphasis coefficient. In the present embodiment, instead of calculating the average value of the hemoglobin amount calculated in the first embodiment within one field, emphasis is performed around a certain reference value obtained statistically in advance. Further, in the emphasis coefficient converter 41, the degree of emphasis can be changed according to the emphasis level input from the front panel or the like.

Then, in the image conversion units 42a to 42c, the enhancement coefficient conversion unit 4 is applied to the RGB signal whose timing is adjusted by the memory 38, as in the first embodiment.
Data conversion is performed based on the emphasis coefficient calculated by 1.
Emphasize the biofunction information. The obtained emphasized image is converted into an analog signal by the D / A converters 43a to 43c, and then transferred to the monitor 17 or an image filing device (not shown).

According to the present embodiment, as in the first embodiment, it is possible to more effectively and naturally express a portion such as a blood vessel or a lesion where the blood flow is changed from the surrounding mucous membrane. It is possible to In addition, even when the emphasis process is not performed, the original image can be directly output by simply turning off the emphasis process in the emphasis processing unit that performs the emphasis process on the original image. It is not necessary to adjust the timing by rewriting to the coefficient of 1 or providing another path that does not perform the filtering process only for displaying the original image.

Further, in the present embodiment, the amount of hemoglobin that is minutely changed around the reference value for which the emphasis level should be strongly set is emphasized by performing the high-frequency emphasis filter processing, thereby increasing the emphasis level. It is possible to effectively emphasize the hemoglobin amount around the reference value without causing whiteout of the image (when halation or whitish mucous membrane overflows) that occurs when strongly setting.

In the present embodiment, the data conversion is performed using the ROM and the differential circuit, but as in the first embodiment, the processing is performed using the field programmable gate array or the like instead of the ROM and the differential circuit. It is also possible to do Further, the circuit of the present embodiment may be built in a video processor, an image filing device (not shown), or the like, instead of being independently configured as an image processing device. In addition, emphasis processing may be performed on biological function information other than the amount of hemoglobin.

7 to 9 relate to the third embodiment of the present invention. FIG. 7 is a block diagram showing the internal structure of the image processing apparatus, and FIG. 8 is a filter of a high-frequency emphasis filter provided in the emphasis filter processing section. Explanatory drawing which shows an example of a coefficient, FIG.
FIG. 6 is an explanatory diagram showing an example of filter coefficients of a noise removal filter provided in a smoothing filter processing unit.

The image processing apparatus 60 of this embodiment has a different internal configuration from the image processing apparatus of the first embodiment, and the RGB signals sent from the video processor 16 are stored in the image processing apparatus 60. Inverse γ correction units 61a to 61c, which are lookup tables that perform inverse γ conversion, are provided. In the present embodiment, the video processor 1
The RGB signals sent from 6 are assumed to be digital data. The RGB image data that has been subjected to the inverse γ conversion has the timing adjustment memory 62 and the logarithmic conversion unit 63a.
~ 63c.

A matrix circuit 64 is provided at the subsequent stage of the logarithmic conversion units 63a to 63c, and an inter-image calculation is performed on the logarithmically converted image data to calculate the hemoglobin amount for each pixel as the biofunction information. It has become. Further, an emphasis filter processing unit 65 and a smoothing filter processing unit 66 are provided in the subsequent stage, and the filter processing is performed by each filter. An emphasis coefficient conversion section 68 is provided via a selection circuit 67 at a stage subsequent to these filter processing sections. The selection circuit 67 selects the processing result of one of the filters and inputs it to the emphasis coefficient conversion section 68. Then, the emphasis coefficient of the intensity designated by the emphasis level is obtained.

Image conversion units 69a to 69c including a ROM and the like are provided at the subsequent stage of the emphasis coefficient conversion unit 68.
Inverse γ correction unit 61 whose timing is adjusted by the memory 62
Based on the RGB signals from a to 61c and the enhancement coefficient from the enhancement coefficient conversion unit 68, the enhancement processing is performed on the RGB signal. Image conversion units 69a to 69c
D / A converters 70a to 70c are provided in the subsequent stage, and the outputs of the image conversion units 69a to 69c are converted into analog signals and output to the monitor 17 or the like to display the emphasized image. Further, the output image of the image processing device 60 is transferred to an image filing device (not shown) or the like for recording, or is output to a printer, a slide photographing device, or the like.

Next, the operation of the image processing apparatus 60 of this embodiment will be described in more detail.

The digital RGB image data sent from the video processor 16 are inverse γ correction units 61a to 61a.
Inverse γ correction conversion is performed by 1c. The data that has undergone the inverse γ correction conversion is logarithmic conversion circuits 63a to 63c.
The logarithmic conversion is performed at, and the inter-image calculation is performed at the matrix circuit 64. Here, the hemoglobin amount for each pixel is calculated in the same manner by the calculation formula shown in the first embodiment.

The calculated hemoglobin amount is input to the emphasis filter processing unit 65 and is filtered by the high frequency emphasis filter having the filter coefficient as shown in FIG. By this high-frequency emphasis filter processing, it is possible to emphasize a contour or the like in which the hemoglobin amount is subtly changing. In the smoothing filter processing unit 66,
The filtering process is performed by a smoothing filter having filter coefficients as shown in FIG. With this smoothing filter processing, it is possible to obtain the effect of suppressing the enhancement of the pixel containing the noise component.

The filter coefficients in each of the filter processing sections are not limited to those listed here, and any filter coefficient can be used as long as the purpose is to perform similar filter processing.

The hemoglobin amount filtered by these filter processing units is input to the selection circuit 67.
Of the results processed by the two filter processing units, a desired filter processing result is selected by an operator's operation instruction or the like, and is input to the emphasis coefficient conversion unit 68.

In this embodiment, the selection circuit 67 is provided to
One of the two filter processing results can be selected, but an image synthesizing circuit, etc. is provided to synthesize the two filtered data based on a predetermined ratio, and to perform superimposition. May be.

Further, in the case of an image with poor contrast or an image with a lot of noise components, a smoothing filter is selected, and an image for which the structure pattern of the mucous membrane is desired to be emphasized, such as an enlarged image of the mucous membrane, is selected separately. Means may be provided to automatically select the filtering result.

In the emphasis coefficient conversion section 68, the same data conversion processing as the processing in the emphasis amount calculation section 29 and the emphasis coefficient conversion section 31 of the first embodiment is performed to obtain the emphasis coefficient in each pixel. In the emphasis coefficient conversion unit 68, a variable emphasis level can be input from the outside or the inside to adjust the degree of the emphasis effect.

Then, in the image conversion units 69a to 69c, as in the first embodiment, the emphasis coefficient conversion unit 6 is applied to the RGB signal whose timing is adjusted by the memory 62.
Data conversion is performed based on the emphasis coefficient calculated by 8.
Emphasize the biofunction information. The obtained emphasized image is converted into an analog signal by the D / A converters 70a to 70c and then transferred to the monitor 17 or an image filing device (not shown).

According to this embodiment, the same effect as that of the first and second embodiments can be obtained, and a desired filter processing can be selected from the two filter processings of high-frequency emphasis and smoothing. Since it is possible, for example, select smoothing filter processing for noisy images, and select high-frequency emphasis filtering if you want to effectively emphasize minute hemoglobin changes such as early cancer. By doing so, more effective image processing can be performed according to the state of the image.

FIG. 10 is a block diagram showing the internal arrangement of an image processing apparatus according to the fourth embodiment of the present invention.

The image processing apparatus 71 of the present embodiment has two filter processing units 6 arranged in parallel in the third embodiment.
The smoothing filter processing unit 66 and the enhancement filter processing unit 65 are connected in series in this order. The configuration of the other parts is the same as that of the third embodiment, and the description is omitted.

As an operation of this embodiment, an operation different from that of the third embodiment will be described here.

The hemoglobin amount obtained in the same manner as in the third embodiment is first input to the smoothing filter processing unit 66, and spike noise and the like are removed by the smoothing filter. This filtering process prevents noise from being emphasized by the emphasis filter processing unit 65 in the subsequent stage. Then, in the emphasis filter processing unit 65, the high-frequency emphasis filter performs high-frequency emphasis processing to emphasize the contours that are finely changed on the hemoglobin distribution image. As with the third embodiment, the filter coefficients in these filter processing units may be any filter coefficients that can achieve their respective purposes.

The filtered hemoglobin amount data is processed by the emphasis coefficient conversion unit 68 in the same manner as in the third embodiment.
Is input to the image processing apparatus, an enhancement coefficient is calculated, and based on the enhancement coefficient, the image conversion units 69a to 69c perform enhancement processing on the original image. The obtained emphasized image is converted into an analog signal by the D / A converters 70a to 70c, and then output to the monitor 17 for display and also output to an image filing device, a printer, a slide photographing device or the like not shown. It

According to the present embodiment, as in the first and second embodiments, it is possible to clearly identify the region where the hemoglobin amount is minutely changed without causing the whiteout or color collapse of the image. It is possible to provide an observable image and prevent noise such as spike noise from being emphasized.

11 to 14 relate to the fifth embodiment of the present invention. FIG. 11 is a configuration diagram of a rotary filter provided in a light source section of a video processor, and FIG. 12 is a diagram of each filter provided in the rotary filter. FIG. 13 is a characteristic diagram showing transmission characteristics.
Is a block diagram showing the internal configuration of the image processing apparatus, and FIG. 14 is an indocyanine green (hereinafter referred to as I
It is a characteristic view showing the light absorption characteristics of CG).

The image processing apparatus of this embodiment is applied to an endoscope apparatus that uses infrared light as a part of illumination light. The video processor provided in the endoscopic device is
By irradiating infrared light having a wavelength near 805 nm, it is possible to obtain information on the deep part of the living mucous membrane. The light source unit in the video processor is provided with a disc-shaped rotary filter 75 on the irradiation optical path as shown in FIG.

The rotary filter 75 is provided with filters 76a to 76c having transmission characteristics as shown in FIG.
These filters 76a to 76c are sequentially inserted in the irradiation optical path so that light of each wavelength in the infrared region, G region, and B region is transmitted and sequentially irradiated. That is, when the filter 76a is inserted on the irradiation optical path, the wavelength light in the infrared region around 805 nm is emitted instead of the wavelength light in the R region during normal visual observation.

An image processing device 80 as shown in FIG. 13 is connected to the video processor having such a light source unit.

In the image processing device 80, an A / D converter 81a for converting the infrared observation image data sent from the video processor from an analog signal to a digital signal.
~ 81c are provided. A / D converter 81a-
Inverse γ correction units 82a to 82c are provided in the subsequent stage of 81c, and inverse γ correction conversion is performed on the infrared observation image data converted into a digital signal, and the output is branched into two paths. It is sent to the timing adjustment memory 83 and the division circuit 84. Then, by the division circuit 84, 80
The ratio of the image by the irradiation light in the infrared region of 5 nm and the other images is calculated.

A smoothing filter processing section 85 and an emphasis filter processing section 86 are provided in this order in the subsequent stage of the division circuit 84.
After the smoothing filter processing unit 85 removes noise from the output of the division circuit 84, the emphasis filter processing unit 86 performs emphasis processing such as edge emphasis. Further, an emphasis coefficient conversion unit 87 is provided in the subsequent stage, and an emphasis coefficient for emphasizing the original image is obtained for each pixel. Here, the degree of emphasis can be changed depending on the emphasis level input from the outside or inside.

Image conversion units 88a to 88c are provided downstream of the emphasis coefficient conversion unit 87, and an inverse γ correction unit 82a is provided.
Up to 82c through the memory 83, the enhancement processing is performed on the original image based on the timing-adjusted original image signal and the enhancement coefficient from the enhancement coefficient conversion unit 87. The γ correction units 89a to 89c and the D / A converter 90 are provided downstream of the image conversion units 88a to 88c.
a to 90c are provided in order, and image conversion units 88a to 88c are provided.
After .gamma.-correction is performed by the .gamma.-correction units 89a to 89c, it is converted into an analog signal by the D / A converters 90a to 90c and output to a monitor or the like so that an emphasized image is displayed. In addition, the image processing device 80
The output image is transferred to an image filing device (not shown) or the like for recording, or is output to a printer, a slide photographing device, or the like.

Next, the operation of the image processing apparatus 80 of this embodiment will be described in more detail.

The living body mucous membrane has the property of relatively easily transmitting light having a wavelength in the infrared region. By utilizing this characteristic, irradiating infrared wavelength light and observing the reflected light,
It is possible to obtain information such as blood vessels running in the deep part of the living mucous membrane and information about the infiltration range of the lesion.

Further, as shown in FIG. 14, ICG, which is a liver function test drug, has a characteristic of specifically absorbing light having a wavelength of 805 nm, and exhibits almost no absorption in the visible light region. By utilizing this characteristic, by irradiating the wavelength light in the infrared region instead of the illumination light in the R region and injecting ICG intravenously during infrared observation, the infrared image assigned to the R image is the ICG.
Since the blood vessel containing the light absorbs the irradiation light, the blood vessel can be observed with good contrast as if it were floating in black.

At this time, since the G image and the B image hardly absorb light due to ICG, images due to changes in the mucous membrane structure or changes due to variations in irradiation light can be obtained.
Therefore, a value proportional to the ICG concentration can be obtained by calculating the ratio of the infrared image in which the ICG absorbs and the image in the wavelength that does not absorb in the dividing circuit 84.

Based on this value, the filtering process is performed in the same manner as in the fourth embodiment, and the enhancement process is performed on the original image as in the first embodiment, so that the brightness of the irradiation light is not affected. An enhanced image can be obtained. Then, from the obtained image, it becomes possible to more clearly observe the running state of the blood vessel, that is, the infiltrated area such as a lesion, and the diagnostic ability can be improved.

According to the present embodiment, by illuminating the infrared light and observing the subject injected with ICG, it is possible to clearly observe the blood vessels running in the deep part of the mucous membrane of the living body. Further, by performing the emphasis process on the original image based on the ratio between the image in the infrared region and the image in the G and B regions where the ICG does not absorb, it is possible to obtain an emphasized image that is not affected by the brightness of the irradiation light. It is possible to more clearly observe the infiltrated area such as a part. Thereby, the diagnostic ability can be improved.

Further, as in the first embodiment, no special circuit is required to output the original image even when the enhancement processing is turned off, and further, the contour enhancement in the filter processing unit is performed to remove fine blood vessels. It is possible to enhance without causing overexposure or color crushing, and it is possible to clearly display an infrared image that tends to be noisy.

15 and 16 relate to the sixth embodiment of the present invention. FIG. 15 is a block diagram showing the internal structure of the image processing apparatus, and FIG. 16 is a characteristic diagram showing absorption and fluorescence characteristics of fluorescein.

The image processing apparatus 91 of the present embodiment has a different internal configuration from the image processing apparatus of the first embodiment, and the endoscope sent from the video processor 16 in the image processing apparatus 91. A / D converter 92a for converting RGB signals of an image from analog signals to digital signals
.About.92c are provided. A / D converter 92a-
Inverse γ correction units 93a to 93c are provided in the subsequent stage of 92c, and inverse γ correction conversion is performed on the RGB image data converted into digital signals. Since the signal from the video processor 16 is γ-corrected, the inverse γ
Inverse γ that restores linear characteristics in the correction units 93a to 93c
Make corrections.

At the subsequent stage of the inverse γ correction units 93a to 93c, a memory 94 for timing adjustment for dividing the output into two paths and dividing circuits 95a and 95b are provided. In the dividing circuits 95a and 95b, the inverse γ correction is performed. The ratio of the color components of R and G, and G and B is calculated for each corrected data. Smoothing filter processing units 96a and 96b are provided at the subsequent stages of the dividing circuits 95a and 95b, respectively, and the output terminals of the smoothing filter processing units 96a and 96b are connected to the dividing circuit 97. And the output end of the smoothing filter processing unit 96b are connected to the selector 98. The outputs of the division circuits 95a and 95b are noise-removed in the smoothing filter processing units 96a and 96b, respectively, and then the ratio is calculated in the division circuit 97, and the output of the division circuit 97 and the output of the smoothing filter processing unit 96b are calculated. Are input to the selector 98.

An emphasis filter processing section 99 and an emphasis coefficient conversion section 100 are provided in order after the selector 98. Either one of the output of the division circuit 97 and the output of the smoothing filter processing section 96b in the selector 98 is provided. Is selected and output, and after the edge emphasis processing is performed by the emphasis filter processing unit 99, the emphasis coefficient conversion unit 100 obtains the emphasis coefficient for emphasizing the original image.

Image conversion units 101a to 101c are provided after the emphasis coefficient conversion unit 100, and timing-adjusted original image signals sent from the inverse γ correction units 93a to 93c via the memory 94. And emphasis coefficient converter 10
Based on the enhancement coefficient from 0, the enhancement process is performed on the original image. Image conversion units 101a to 1
In the subsequent stage of 01c, γ correction units 102a to 102c, D /
A converters 103a to 103c are provided in order, and outputs from the image conversion units 101a to 101c are γ correction units 102a to 102a.
After the .gamma.-correction is performed by 102c, the D / A converters 103a to 103c convert the signals into analog signals, which are output to a monitor or the like so that an emphasized image is displayed. Further, the output image of the image processing device 91 is transferred to an image filing device (not shown) or the like for recording, or is output to a printer, a slide photographing device, or the like.

Next, the operation of the image processing apparatus 91 of this embodiment will be described in more detail.

When a fluorescent substance called fluorescein having absorption and fluorescence characteristics as shown in FIG. 16 is injected into a vein while observing a living mucous membrane, the concentration of fluorescein in blood is changed and fluorescence is emitted. As shown in FIG. 16, when fluorescein irradiates light of wavelength B, it absorbs this light and emits fluorescence. Therefore, in the case of observation by the frame-sequential endoscope apparatus, fluorescence is emitted at the timing of irradiation with the wavelength light in the region B, and at this time, the higher the concentration of fluorescein, the stronger the fluorescence.

Further, during signal processing, regardless of the emission wavelength of the fluorescence, fluorescence is emitted at the timing when the light of the wavelength in the region B is irradiated. Therefore, magenta is superior to the mucosa having a high fluorescein concentration in the obtained image. Be observed. Therefore, the fluorescein concentration distribution can be observed by the change in color tone.

However, since pink is predominantly observed on the living mucous membrane, it is difficult to clearly observe this portion even if the color changes from magenta to magenta. Further, since the amount of fluorescence emitted at this time is minute, the effect cannot be expected so much even if observation is performed as it is.

Therefore, it is possible to clearly observe the concentration distribution of fluorescein by using the image processing apparatus 91 having the configuration as in the present embodiment to perform the emphasis process on the observed image.

The RGB image signal input to the image processing device 91 is converted from an analog signal to a digital signal,
After the γ correction is released, the B component and R
The ratio B / R of the components is calculated. This value is proportional to the concentration of fluorescein, but is also affected by blood volume.
Therefore, the ratio R / G of the R component and the G component is calculated by the division circuit 95b.
By calculating, a value proportional to the blood volume is calculated,
The value that correlates with the concentration of fluorescein is normalized.

The data output from the division circuits 95a and 95b each include a noise component, and in particular, the value correlated with the fluorescein concentration output from the division circuit 95a has a small amount of fluorescence emission, and thus the noise component. It is easily affected by. Therefore, the smoothing filter processing units 96a and 96b respectively perform filter processing using, for example, a median filter or the like to reduce the influence of noise.

The value correlated with the noise-removed fluorescein is also normalized in the division circuit 97 by a value proportional to the noise-removed blood volume. here,
A value that is proportional to the fluorescein concentration with the effect of blood volume removed is calculated. For the fluorescein concentration distribution image calculated in this manner, for example, in order to clarify the boundary portion of a region where the fluorescein concentration is significantly higher than the surrounding mucous membrane, such as blood vessels, contour emphasis or the like is performed in the emphasis filter processing unit 99. To do.

Then, in the emphasis coefficient conversion section 100,
An enhancement coefficient is obtained from the fluorescein concentration distribution image. At this time, in order to facilitate the observation, the enhancement coefficient conversion unit 100 enlarges the difference amount from the preset reference value for the fluorescein concentration distribution image whose edge is enhanced by the enhancement filter processing unit 99. . Further, in the emphasis coefficient converter 100, the emphasis level can be changed by an emphasis level setting switch or the like provided on the front panel of the image processing apparatus or the like, and by this emphasis level setting, the difference amount from the above-mentioned reference value is set. Change the degree of expansion.

Next, in the image conversion units 101a to 101c, for the RGB signals of the original image whose timing is adjusted by the memory 94, based on the enlarged fluorescein density distribution calculated by the emphasis coefficient conversion unit 100,
Emphasize the biofunction information. This emphasis processing is almost the same as that in the first embodiment, and conversion into an image having a newly determined fluorescein concentration distribution is performed.
That is, when the fluorescein concentration distribution is calculated from the original image by the above procedure, the fluorescein concentration distribution enlarged by the emphasis coefficient conversion unit 100 is obtained.

Further, in the present embodiment, not only the fluorescein concentration distribution can be emphasized but also the blood volume can be emphasized. In this case, the selector 98
By selecting the output from the smoothing filter processing unit 96b, it is possible to obtain an emphasized image in substantially the same manner as in the fourth embodiment.

According to the present embodiment, by emphasizing the change caused by the weak fluorescence emission, it is possible to easily observe the fluorescence observation image which was difficult to observe as it is. Further, by performing filter processing such as contour enhancement on the fluorescein concentration distribution, it is possible to clearly observe the boundary portion, which has not been able to be observed so clearly until now.

Further, when the fluorescent agent is injected, if the monoclonal antibody having the ability to accumulate in the tumor is chemically bonded to the fluorescent agent, the intravenously injected fluorescent agent accumulates in the tumor. When excited by the excitation light, it emits fluorescence. Therefore, since the fluorescence intensity around the tumor becomes strong, it becomes easy to find a lesion. By using this inspection method in the configuration of the present embodiment, it becomes possible to more clearly observe the boundary of the lesion.

In the present embodiment, fluorescein has been described, but it is also possible to use other fluorescent agents such as hematoporphyrin and adreamycin, and to emphasize them with a configuration considering their absorption and fluorescence characteristics. is there.

As described above, in the embodiment of the present invention, the emphasis based on the hemoglobin pigment amount mainly correlating with the blood volume has been described. However, as in the fifth embodiment and the sixth embodiment, a living body other than the hemoglobin pigment is used. It is also applicable to dyes that can provide information about their function.

Further, in each of the above-described embodiments, the case where the invention is applied to the frame-sequential endoscope apparatus is described, but also in the simultaneous endoscope apparatus, after the obtained image is converted into the RGB image. The same process may be performed. Further, the CCD is not limited to an electronic endoscope having a solid-state image sensor at the tip of the insertion portion, but may be replaced by an eyepiece of an endoscope such as a fiberscope or a rigid endoscope capable of macroscopic observation, or by replacing with an eyepiece. It can also be applied to an endoscope that is used by connecting an external television camera having a solid-state imaging device such as.

As described above, in the present embodiment, mainly in an endoscopic image, an image relating to hemoglobin, that is, an image relating to hemoglobin, that is, the concentration of hemoglobin that governs color and the oxygen saturation are calculated for each pixel. The information image regarding the biological function is obtained, the image is subjected to filter processing such as noise removal, and based on the processing result, processing such as color enhancement is performed on the original image in the enhancement processing unit. This allows fine biological function information to be clearly observed by performing edge enhancement without causing whiteout or color crushing, and also allows the original image to be output as a through image without performing any enhancement processing. Even if you want to do so, it is possible to output the original image by turning off only the biofunction information emphasis processing unit, so you do not need a path or timing adjustment means that bypasses the filter processing unit as in the past. It is possible to reduce the circuit scale.

Further, since the noise removal filter processing is performed on the biofunction information image and the noise removal or other filter processing is not performed directly on the original image, the problem that the entire image is blurred is solved. Also, since errors such as the amount of hemoglobin due to noise can be removed, the problem of emphasizing spike noise is reduced, and it is possible to perform emphasizing processing while retaining fine information in the original image.

[Supplementary Notes] (1) In an image processing apparatus for performing image processing on an input original image, a biological function information calculating unit for calculating information on biological function from the original image, and an image on the biological function information. An image processing apparatus, comprising: a filter processing unit that performs filtering on the original image; and an emphasis processing unit that performs an emphasis process on the original image based on the filter processing result.

(2) The image processing apparatus described in appendix 1, wherein the filter processing means performs filtering for the purpose of removing noise.

(3) The image processing apparatus according to appendix 1, wherein the filter processing means performs filtering for the purpose of emphasizing information relating to biological functions.

(4) The image processing apparatus according to appendix 1, wherein the information on the biological function is calculated by the inter-image calculation of the original image in the biological function information calculating means.

[0120]

As described above, according to the present invention, it is possible to output an original image and an emphasized image without complicating the structure, and to perform an emphasizing process without losing fine information in the original image. There is an effect that an image processing device capable of performing

[Brief description of drawings]

FIG. 1 is a block diagram showing a schematic configuration of the present invention.

FIG. 2 is an overall configuration diagram of an endoscope apparatus including an image processing apparatus according to an embodiment of the present invention.

FIG. 3 is a block diagram showing an internal configuration of the image processing apparatus according to the first embodiment of the present invention.

FIG. 4 is an explanatory diagram showing an example of filter coefficients of a smoothing filter provided in a filter processing unit of the first embodiment.

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

FIG. 6 is an explanatory diagram showing an example of filter coefficients of a high-frequency emphasis filter provided in a filter processing unit of the second embodiment.

FIG. 7 is a block diagram showing an internal configuration of an image processing apparatus according to a third embodiment of the present invention.

FIG. 8 is an explanatory diagram showing an example of filter coefficients of a high-frequency emphasis filter provided in an emphasis filter processing unit of the third embodiment.

FIG. 9 is an explanatory diagram showing an example of filter coefficients of a noise removal filter provided in the smoothing filter processing unit of the third embodiment.

FIG. 10 is a block diagram showing an internal configuration of an image processing apparatus according to a fourth embodiment of the present invention.

FIG. 11 is a configuration diagram of a rotary filter provided in a light source unit of a video processor according to a fifth embodiment of the present invention.

12 is a characteristic diagram showing transmission characteristics of each filter arranged in the rotary filter of FIG.

FIG. 13 is a block diagram showing an internal configuration of an image processing apparatus according to a fifth embodiment of the present invention.

FIG. 14 is a characteristic diagram showing absorption characteristics of indocyanine green (ICG), which is a liver function test agent.

FIG. 15 is a block diagram showing an internal configuration of an image processing apparatus according to a sixth embodiment of the present invention.

FIG. 16 is a characteristic diagram showing absorption and fluorescence characteristics of fluorescein.

FIG. 17 is a block diagram showing a first configuration example of a conventional image processing apparatus.

FIG. 18 is a block diagram showing a second configuration example of a conventional image processing apparatus.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 ... Biological function information calculation means 2 ... Filter processing means 3 ... Delay means 4 ... Enhancement processing means 11 ... Electronic endoscope 16 ... Video processor 17 ... Monitor 18 ... Image processing apparatus 25 ... Frame memories 26a-26c ... Logarithmic conversion part 27 ... Matrix circuit 28 ... Filter processing unit 29 ... Enhancement amount calculation unit 30 ... Average value calculation unit 31 ... Enhancement coefficient conversion unit 32a to 32c ... Image conversion unit

Claims (1)

[Claims] 1. An image processing apparatus for performing image processing on an input original image, a biological function information calculating unit for calculating information on biological function from the original image, and filtering on the image on biological function information. An image processing apparatus comprising: a filter processing unit that performs the above process; and an emphasis processing unit that performs an emphasis process on the original image based on a result of the filter process.