JPH0236836A - Electronic endoscope image processing device - Google Patents

Abstract

PURPOSE:To aim at increasing the amount of information suitable for processing an image, by separating an image into several color signals, and by providing a means of calculating a histogram for thus separated color signals and a characteristic converting means for performing a function conversion in accordance with thus calculated histogram. CONSTITUTION:In the case of image processing in an image processing device 30, required signal data are transmitted from memory sections 46G, 46R, 46B in response to control signals for controlling the timing of transmission delivered from a control signal generating section 47. The signal data transmitted to work memory 51 is subjected to a histogram creating process, a characteristic converting process for density distribution of thus obtained color signals, and the like by means of a computing and processing device 52, and resulted data are transmitted to the work memory 51. If a density gradation conversion is carried out as a histogram conversion, an amount of effective information may be obtained wavelength ranges R, B which have been effectively used conventionally.

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial Application Field 1] The present invention relates to an endoscopic image processing device equipped with means for decomposing an image into a plurality of color signals and independently varying the output level of the decomposed color signals. .

[Prior art] In recent years, by inserting an elongated insertion section into a body cavity,
Endoscopes are widely used to observe diseased areas within body cavities without requiring incisions, and to perform treatment using treatment tools if necessary.

In addition to the optical endoscope (fiberscope) that uses an image guide as an image transmission means, there are also
Electronic endoscopes using solid-state imaging probes such as OD (hereinafter referred to as
Electronic endoscopes (electronic endoscopes) have come into practical use.

Also, an IC1fffl image camera was connected to the eyepiece of the fiberscope using a solid-state imaging device such as a COD, so that it could be displayed on a monitor.

Figure 19 is filed by the applicant in patent application No. 61-207432.
Most of the previous examples proposed in this issue have almost the same configuration.

Image No. 15 output from the CCD 1 forming the photographing FA means is input to the amplifier 2 and amplified to a voltage level within a predetermined range. Thereafter, the signal enters the γ correction circuit 3 and is subjected to γ correction. In the case of the RG B-side sequential method, the signal after γ correction is converted into a digital product by a)7 narrow-to-digital converter (Δ/D converter) 1, and then converted to a digital product by a selector switch 5.
It is recorded in 6R26G and 6B. These meals are 6R and 6G.

The image data recorded in 6B are simultaneously read out by applying a timing signal outputted from the control signal generator 7, and are converted into color signals R, G, and B in nonononalogram quantities by the D/A converter 8, respectively. The image 111 is output from the output terminal 9.The image 111 is displayed on the display screen of the monitor by the color signals R, G, B output from the output terminal 9 and the synchronization signal output from the synchronization signal generation circuit 10. The image of the subject is displayed.

On the other hand, the frame-sequential illumination means when performing frame-sequential imaging uses the R, G, and B R, G, [-
3 and irradiates the light guide 15 with the illumination light. However, from the other end of this light guide 15, towards the subject -
I.J.R.

Sequentially illuminated with light of wavelengths G and B, and under this illumination CC
A subject image is formed on the imaging surface of the DI.

In the preceding example, the input and output pins are lr and I- for all R, G, and B signals. What kind of influence such a system has on the reproducibility of endoscopic images will be explained below with reference to FIG. 20.

Figure 20 (8) shows the distribution of the ornamental stone levels of the 97 subjects.

In general, the brightness level of a subject viewed through an endoscope is mostly due to the nasal component and less due to the blue component.

FIG. 20 CB) shows the histograms of output values after digital conversion for R, G, and B for the luminance level range shown in FIG. 20 (8). Since the characteristics at the time of input are determined by C based on the average brightness level of the subject, the bias in the brightness level of the subject is output as is.

As is clear from this, the G (g signal is outputted in a normal manner).However, the primary signal is biased towards a high level and leads to saturation.On the other hand, the B signal is biased towards a low level and is easily buried in noise.

[Problems to be Solved by the Invention] It is clear from FIG. 20 that <r J: In the prior example, the characteristics at the time of input are adapted to the average brightness level of the subject.

Therefore, there is a drawback that the R43 signal is biased toward a high level, and the 1] signal biased toward a low level. These drawbacks reduce the amount of effective information contained in the image.Also, since the R signal is biased towards high levels, there are problems such as masking minute level signals and making their identification an in-house job. arise.

The present invention was made in view of the points mentioned above.
The purpose of this invention is to provide an endoscopic image processing device that can obtain an effective amount of information from a wavelength range that has been difficult to utilize effectively.

[Means and effects for solving the problem] The present invention includes a means for decomposing an endoscopic image into a plurality of color signals, a means for extracting a histogram for each of the decomposed color signals, and a means for calculating a histogram for each color signal. Based on the histogram,
By providing a characteristic conversion means that performs function conversion, it is possible to easily obtain an amount of information suitable for image processing and to perform image processing with a good S/N ratio.

[Example] Hereinafter, the present invention will be specifically explained with reference to the drawings.

FIGS. 1 to 8 relate to the first embodiment of the present invention.
The figure is a block diagram of the endoscopic image processing device of the first embodiment, Figure 2 is the overall block diagram of the first embodiment, and Figure 3 is a block diagram of the endoscopic image processing device of the first embodiment. , Fig. 4 is a characteristic diagram showing an example of a calculated histogram, Fig. 5 is a characteristic diagram showing a histogram of an image obtained by temperature conversion from the linear His 1 Hegerum, and Fig. 6 is an arbitrary histogram. For the density levels of the two points -(t characteristic diagram, Figure 7 shows the histogram of the image obtained by performing m degree conversion, flattening characteristic conversion is performed) 1] - Char 1 to Figure 8, Figure 8 7 is a characteristic diagram showing a histogram of an image obtained by the processing of FIG. 7. FIG.

As shown in FIG. 2, the (endoscope) image processing device 821 of the first embodiment includes an electronic endoscope 24 into which an elongated insertion section 23 is inserted into the body cavity of a cow body 22, and mirror 24]-
An observation device 26 to which the connector 25A of the universal cord 25 is connected has a function of processing video signals, a color monitor 27 connected to this observation device 26 and capable of displaying video signals in color, and a branch from the connector 25△. A suction device 29 is connected to the electronic endoscope 2/I via the Ic suction coil 28, and the observation device 26 is connected to the R9G, B
It is comprised of an image processing device fffi 30 that generates a histogram from the color signals and performs conversion and adjustment of 111 degree IIPi tone characteristics for each color signal.

As shown in FIG. 1, in the electronic endoscopy @24, a light guide 31 that transmits illumination light is inserted into the insertion portion 23 of the set head, and the light guide 31 inserted through the insertion portion 23
The rear end side is further marked with J universal code 25 as shown in Figure 2.
It is penetrated inside. Therefore, the connector 25Δ is observed with an observation device.
26, illumination light is also supplied from the light source section 32.

That is, the light source unit 32 irradiates the white illumination light from the white lamp 33 onto the incident end of the light guide 31 through the rotary filter 35 that is rotationally driven by the motor 34 . This rotary filter 35 is equipped with filters 35R, 35G, and 35B that transmit light of each wavelength of red, green, and blue.
At the input end of the
illumination light is sequentially supplied. Therefore, this lie I guide 31 is transmitted to the other end side, and the other end (output)
R and G toward a subject such as an affected area within the living body 22 from l).

Emit the illumination light B. The object illuminated with R, G, and B light is captured by an objective lens 37 disposed within the distal end of the insertion section 23.
Accordingly, the CC disposed on the focal plane of the objective lens 37
An image is formed on the first most image plane of D38. The C0D 38 outputs a photoelectrically converted electrical signal by applying a drive signal from the driver 39.

The image signal output from this CCD 38 is transmitted to the amplifier 4
1, and an electrical signal in a predetermined range (for example, 0 to 1)
After being amplified to volts), the A/D converter 4, which forms the signal processing section 13 in the observation device 2G, passes through the signal cable 42.
4) J is received. This A/D converter 44 converts the signal into a hanged digital signal, which is then sent to the selector switch 4.
5 and is input to the memory unit 716.

The changeover switch 45 has a number of changeover contacts equal to the number of color transmission filters 35R, 35G, and 35B of the rotary filter 35 (that is, three contacts), and the two contacts are successively activated in response to a changeover signal from the control signal generator 47. Can be switched. Each of these contacts connects three memories 46R and 46G forming the memory section 46.
, 46B, and the digital signals are recorded in the memories 46R146G and 46B through the turned-on contacts.

Above memory 4.6R, 4.6G, 4.68V
It is connected to the image processing device 30. This image processing device 30 is composed of a working memory 51, an arithmetic processing device 52, an auxiliary storage device 53, an external output device 54, and an analog switch 50, which are memories 46R, 46G,
46B.

When performing image processing in both image processing devices 30, the working memory 51 is connected to the memories 46R and 46G by a control signal that controls the transfer timing from the control signal generator 47.
, 46B, the required signal data is transferred. The signal data transferred to the working memory 51 is processed by the processing unit 52 to generate a histogram, convert the characteristics of the density distribution in the obtained color signal data, etc., and use the results for working purposes. Forward to email address 51. Also,
The signal data subjected to characteristic conversion processing can be stored in the auxiliary storage device 53 as required, and the Hiss 1 to Grams etc. can also be printed out using an external output device 5/I such as Blink.

In addition, if you do not want to perform the above image processing, turn on Analyzer 46R, 46G/1.6B.
The signal data stored in 1]/A converter 55a,
Appear to 55b and 55c at the same time.

The ablog color signals converted by each D/A ml inverter 55a, 55b, 55c are further γ-corrected by gamma correction circuits 56a, 56b, 55c, and output from the output terminal 57 to the monitor 27 side, where they are displayed in color. Ru.

Further, the control signal generating section 47 outputs a timing signal to the synchronizing signal generating circuit 58.
The SYNC signal is output from the SYNC signal.

The control signal generator 47 also outputs a signal for controlling the rotation of the motor 3/I that rotates the rotary noiler 35. Further, this control signal generating section 47 controls the on/off of the ab-1'g switch 50, and also controls the arithmetic processing unit 5.
2, the arithmetic processing unit 52
11 to the control signal generating section 47, so that efficient data transfer and zero can be carried out.

The operation of the first embodiment configured in this way is described below in 31.
I will clarify.

As shown in FIG. 1, the image signal from CCl') 38 is amplified by an amplifier/1.1 to a voltage in a predetermined range, e.g. This image signal is digitized by an A/D converter/171I at a certain quantization level (for example, 8 bits). This digital signal is transmitted to the CCD 38 based on the control signal from the control signal generator 47 via the changeover switch 45.
When the image signal imaged is under red (R) illumination, it is stored in the R signal memory 46R, and when it is under green (G) illumination, it is stored in the G signal memory 46G. (B) When the signal is 1 due to illumination, it is stored in the B signal memory 46B.

Each memory = 1.6R, /1.6G, 46B has independent human output, and the power and output can be performed at the timing of each monologue.

The image signal data stored in each of the memories 46R, 46G, and 46B is input to the image processing device 30. In this case, if the mode does not perform image processing (for example,
When the selection switch is selected (“turn off”), the image processing device @ 30 sends a J signal to the control unit 41 section 47 to indicate that it is not in a state where image processing is to be performed, and this signal Based on this, the analog switch 50 is turned to A, and the memory 46
The image signal data stored in R, 46 is D/A in the same 1h.
The signals are input to converters 55a, 55b, and 55c, and converted into a single color signal Shingetsu. At that time, the γ correction circuits 56a, 56
Color signals R, G, B that have a predetermined γ characteristic at b, 56c
and is outputted to the monitor 27 and displayed on the monitor 27 along with the 5YNC signal of the synchronization signal light generation circuit 58.

On the other hand, when the image processing device 30 is set to the image processing mode, it performs the process of generating the J histogram shown in FIG.

The process of calculating a histogram will be explained with reference to FIG.

The calculation of the histogram will be explained in FIG.

Now, convert the image into a two-dimensional front row 1mage (Xsize, Y
s'1e), X5ize means the image size in the X direction, and y5ize means the image size in the Y direction.

In addition, Hiss 1 to Gram is one-dimensional front row 11ist (Lev
When expressed as el), Levcl means the number of density levels (for example, 1] If the childization level is 8 bits, 256 levels from O to 255).

First, process the image 1mage (×5ize, Y
5ze) is transferred to the working memory 51, and the memory contents (for example, provided in the working memory 51, for storing His 1 hegerum when gradually released) are transferred to the working memory 51.
l i st (represented by Level )) is initialized (assigned to O). In addition, prepare X as a variable that indicates the position in the X direction, 1y as a variable that indicates the position in the Y direction, and ρ as a variable that indicates the density level of the image, and initialize each (substitute O). 3° Then, perform the following processing.

(1) Increment y by 1 count.

(2) Initialize X (substitute O)'? I-ru.

(3) Increase x by 1 count.

(4), 1 image of 1 page indicated by x and yr in Q
(X.

Substitute the fuchisho level of y).

(5) Upload Mist (, Q) by 1 count].

(6) If x < X 5ize, return to (3),
In other cases, proceed to (7).

(7) If y<Y 5ile, return to (1), and if it is Sen, the process ends.

Through the processes (1) to (7) above, the one-dimensional front row 11
A histogram is created on ist. Figures 3 and 4 show an example of a histogram created in this way. The horizontal axis is the concentration level (O・~255 in this example)
, and the vertical axis represents the frequency. Next, referring to FIG. 5, we will explain the histogram conversion, for example, 1li11 anti-synthesis conversion processing from the histogram '14 generated in the above process.

The processes (1) to (7) above are carried out in the memories 46R, 4 (3).
When applied to the signal data of G and /16B, the result is as shown in Fig.
), (b), and (c).13 These histograms are for normal endoscopic images, and as can be seen from these (a), (b), and (c), , King's histogram is only a part of the falsified m degree bell number (O ~ 255 if the quantization level is 8 pi 1-). An image like a mouth has a low -1 contrast and is difficult to distinguish. The current level of hanging is 8 bits, and IJ is the maximum of Hiss 1 Hegelam! The li degree level is ρmaX, the minimum wi degree level is 9min,
At the same time, the original image is displayed as Image(x.

It is expressed as y). Image Ima after performing density gradation conversion F
ge' (x, y) is calculated by the following equation.

Image' (x, y) -255(Imag
e(x,y)−,Om1n)/(,Qma
x −, Q m1n)・(1) Here, 1≦X≦
X s+ze1≦y≦Y 5IZe.

This allows the normal narrow concentration level range [ρmin, fl
The images 1ma (le (x, y) in Fig. 5 (a), (b), and (c) in maxl are respectively (a'
), (b'), (C"), the level range [, G min, j maxl is expanded and translated, so that it is distributed over the entire concentration level [0, 255], Contrast is emphasized.Furthermore, an effective amount of information can be obtained from wavelength ranges (R, B) that have not been effectively utilized in the past.

Figure 6 shows the Fuchima level range [ρmin,, fl ma
Instead of xl, any two points β1. ρ2 is specified, and the density gradation conversion of equation (1) is performed within the specified density level range [, G+, j21''('').

Next, a description will be given of how the histogram flattening process is performed using the apparatus of the first embodiment. This conversion is performed so that the height of the histogram becomes the same in all peaks. Now, if we subtract 8 pixels from the hanging level, the number of pixels Dot that one density value has can be obtained by the following equation.

Dot - (X 5ize x y 5ize) /
256 =-(2) Also the cumulative density function CD(i) in His(-grams.

For O≦i≦255, the following formula 'C is obtained.

In order to perform the above flattening process, the process shown in the flowchart of FIG. 7 is performed.

Image to be processed 1meg (X 5ize, Y
5ize) to the working memory and convert it]! I! Later image 1 m a jJ (i(X 5ize, Y
5ize). Also,
As shown in FIGS. 5(a), 5(b), and 5(c), it is assumed that the histogram t of the image exists in l1iSt(1-evel). Furthermore, X is a variable representing the position of the force plane, y is a variable indicating the position in the Y direction, and a working variable is i 1. i2. j 1. Prepare j2 and initialize each.

Next, the following sub-processes (1) to (7) are performed.

(1) Cover i2 with 1 count amplifier.

If 2>255 then (7)l\, if 12≦255 then (3)
Move to.

(3) Take out a new m-pricing bell and substitute j21\.

(4) If , j2- j 1 <1, return to (1) l\; otherwise, proceed to (5).

1≦Ima(le(X, V) < i 2 For all X and y, Image' (x, y >
Substitute j2-1 into .

Substitute 12 for 1 and j2 for jl, and return to (1).

1≦+maOc(x, V) < l2 for all X and y, l1lla(le' (X,
Substitute J2-1 for V).

Through the steps (1) to (7) above, Image (
An image is created in which the histodarum is flattened (X 5ize, Y 5ize). From this, each image (x, y) of the histogram shown in Figure 8, (a), (b)', and (c), is shown in Figure 8 (a'), (b'). ) (C'), the flattened image Image' (x.

y).

1ε) In other words, the part where the 8m distribution is dense (the mountain part in His 1-gram)
is expanded, and areas where the density distribution is sparse (troughs in the histogram) are compressed. Since the dense portion has a larger number of pixels than the sparse portion, the number of pixels expanded for the entire image increases, and the contrast is emphasized to 1.

In the embodiment of the present application, since the output signal level is changed for each color signal, the color balance is the best.

Although it is optimal for performing image processing, it is difficult to output it to the monitor as it is (but it is difficult to output it in natural colors. Therefore, the image processing device 30 normally uses AU, and only when necessary does it output the control signal from the control signal generator 47. As a result, it is normally possible to display natural colors on the monitor as in the conventional example.

Further, the converted image can be stored in the auxiliary storage device 53, and the histogram can be output to an external output device 54 such as a printer.

In the embodiment of the present application, a frame-sequential electronic endoscope using RGB signals has been described, but the present invention can be similarly applied to a single-panel electronic endoscope that encodes a digital video signal.

FIG. 9 shows the main parts of the second embodiment of the present invention.

In this second embodiment, when a newly installed switch 71 is operated in the first embodiment shown in FIG. Control is performed to change the amount of illumination light of 32.

That is, after performing illumination with the normal amount of illumination light, the control signal generation unit 47 outputs a control command to reduce the amount of illumination light to the aperture control circuit 72, and the aperture blades 73 and the aperture blades 73
The aperture amount of the aperture device 75 consisting of the aperture motor 71 that rotates the aperture motor 71 is increased.

When one frame worth of image is captured in this state, the control signal generating section 47 then outputs a control command to the diaphragm control circuit 72 to greatly overshadow the amount of illumination light, thereby reducing the amount of light shielded by the diaphragm blades 73. Then, images are taken under a large amount of illumination light. Note that signals captured under different amounts of illumination light are transferred to the working memory 51 and subjected to image processing.

The @ image under the normal illumination light amount is the same as in the first embodiment. In other words, the distribution of R, G, and B color signals is as shown in FIG. 10(a).

On the other hand, when the amount of illumination light is reduced, the histogram of an image captured under red illumination light has a characteristic close to that of normal A5, as shown in FIG. 10(b).

When the amount of illumination light is further increased, Fig. 10 (C)
The blue histogram is close to a normal distribution.
Cru. Therefore, the amount of information that is missing in a normal image can be obtained.

Next, a third embodiment will be described in which the effect of histogram conversion is improved while maintaining the color tone obtained by the endoscope.

The configuration of this third embodiment is based on the first embodiment shown in FIGS. 1 and 2.
This is similar to the example. In this embodiment, the image processing method in the image processing mode of the image processing device 30 is completed, and the image processing method shown in FIG. ) to obtain the processed image.

That is, when the image processing device 30 shown in FIG. 1 is set to the image processing mode, the menus 46R, /16G, /! 1
．． 6B of data are sequentially transferred to the working memory 51.

For example, the R data transferred to the working memory 51, the arithmetic processing unit 52 takes in two pixel data, compares them, leaves the larger value (smaller value), and performs a comparison operation. The maximum value (minimum value) is detected by performing this on the entire R data. By performing similar comparison operations on other G data and B data, the JR
, G, B, respectively, maximum value ITI a X and minimum value m
Detect n.

, set an appropriate number of levels between the maximum value max and the minimum value min in L, determine which level corresponds to the data sequentially read out from the working memory 51, and change the level to that level. - Increment the count value of the histogram calculation counter by one.

By performing this processing on the entire R data, it is possible to calculate a histogram based on the R signal. Similarly, other G
-5'- data, by performing this on the B data, the 11th
The second process in Figure 12 or Figure 12 (same figure for (a))
A histogram like that shown in (b) will be obtained.

Next, apply the maximum value max and minimum value m to the above R data.
The above-described flattening process is performed on the range of in. The other G data and B data are also flattened in the same way, and as shown in FIG. 12(C), R,
Obtain flattened histograms for the G and B signals.

Then, the R, G, and B data flattened by turning on the analog switch 50 are transferred from the working memory 51 to D/A=
1 inverter 55a.

The signal is output to the monitor 27 (see FIG. 2) via 55b and 55c.

By performing the above-mentioned flattening process, the average value M of the color signals R, G, and B shown in FIG. The average value M' also does not change much.

Therefore, in the first embodiment, in which the levels after histogram conversion of R, G, and B are almost the same, the entire image is white (
In contrast, in this example, the existing color tones are integrated and the histogram conversion is performed. The advantages of f6. ．． In addition, the above-mentioned "+3" embodiment can be used to perform simple processing with the strength of the lean-in 1 herast.

In the above three actual flow examples, the average of the image before the calculation of the hiss 1-gram (self M and the average after the flattening process (0M') are J, which performs the 1! flattening process so that there is almost no change in /υ, It's good to do it.

Figure 13 shows the fourth embodiment of the present invention (image processing).
Indicates Ll-.

The configuration of this third embodiment is similar to that of the '1st embodiment, but the image processing is different.

As shown in FIG. 13, the average values of R, G, BiiEii(m) are calculated.
To transfer the R, G, and B data of 6 to the working mailbox 51, calculate the sum of the general levels of each 1 dog, G, and B data, and calculate the maximum and minimum. The average value M can be obtained by dividing by the number of level sections.

Next, create a conversion data with the average value M at the inflection point using the following formula (% formula %)
-ru. In this case, the average value M is set to a signal range of O・-1 by 9=
The average value m is set to 11.

In other words, the signal range is O~1, the average value is r+, the input is represented by X, the output is represented by y, p is a real number, for example, p=
It can be set to 4. The characteristics of the deflection curve in this case are as shown in Fig. 1'1.

By performing temperature conversion using the above conversion table,
The R, G, and B data in Figure 15 (a) are shown in Figure 15 (C).
The image processed as follows is displayed as 1q.

In this fourth embodiment, in J3, A-melon conversion is performed so that the average values of R, G, and B do not substantially change.

FIG. 16 shows an endoscopic image processing device 8 according to a fifth embodiment of the present invention.
1 to J0 This device 81 is different from D/J0 in the prior example shown in FIG.
A conversion processing unit 82 is connected between the A converter 8 and the image output terminal 9.
The configuration is such that the configuration is similar to that of the preceding example shown in FIG. 19 above.

The configuration of the conversion processing section 82 is shown in FIG. 17.

It is converted into an analog by the D/A converter 8.

The G and B signals are input to the window circuit 91 and the first
Powered by amplifier 92. This window circuit 91 is a circuit that passes only a valid image signal portion, for example, values from 10 to 254, and the output of this window circuit 91 is input to an average value calculation circuit 93, and from this circuit 93. The average value signal is applied to the gain control terminals of the first amplifier 92 and the second amplifier 94.

Accordingly, the first amplifier 92 performs gain control based on the average value signal so that the average value becomes the center of the signal level. On the other hand, the second amplifier 94 performs gain control to return the gain control of the first amplifier 92 to its original state.

The output signal from the first amplifier 92 is input to an A/D converter 95, converted into a digital signal, and then applied to the address end of a table 96 made up of ROM and written to that address. Read memory contents J0 This table 96
The data read from is input to the D/A converter 97,
After being converted into an analog signal, it is amplified by the second amplifier 94 and output from the output terminal 9.

In the daple 9G, data of the IJ and U-1111 degree conversion characteristics shown in FIG. 18(C) are written.

The above A/D converter 95, table 96, D/A converter 9
7, the timing of A/D conversion, readout of the daple 96, and D/Δ conversion is controlled by a clock from a clock generator 98.

The operation of this fifth embodiment will be explained below.

The operation up to the D/A converter 8 is the same as that shown in FIG. 19, and its explanation will be omitted.

Thus, the input signals that have passed through the D/A converter 8, that is, the R, G, and B signals, are each input to the conversion processing section 82. The input signal is processed by the window circuit 91, and when the input signal level is O~255, the input signal is 10-245.
Only through J-0, that is, particularly bright parts and particularly dark image parts are removed.

The output of this window circuit 91 is input to an average value calculation circuit 93, and the signal is integrated by this circuit 93 to calculate the average brightness of the image. That is, the average brightness of the effective portion of the image is calculated. The output of this average value calculation circuit 93 is marked 1 at the gain control end of the first amplifier 92.
[1], and the gain of this amplifier 92 is adjusted so that the average value is at the center of the signal level. As a result, Figure 18 (a
), the level distribution of the input signal shown in (b)
As shown in A, the gain is adjusted according to the level distribution of the output signal.

The output of the first amplifier 92 is digitized by an A/D converter 95. This conversion is performed in synchronization with the clock of the clock generator 98. However, the A/D converter 95
is applied to the address end of the table 96, and the output of the first
With the conversion characteristics shown in Figure 8 (C)? ! l11 degree conversion converts rows.

The signal whose density has been converted by the daple 96 is returned to an analog signal by the D/A converter 97 at the next stage, and the level distribution of the signal in this case is as shown in FIG. 18(d). Become.

The output of this D/A converter 97 is input to a second amplifier 94, and the second amplifier 94 reversely adjusts the gain adjusted by the first amplifier 92.

Thus, the signal shown in FIG. 18(e) is obtained.

The signals R, G, and B that have been subjected to the above conversion processing are output from the conversion processing section 82.

Also in this fifth embodiment, the average values of the densities of rRG and B before and after the conversion process are approximately equal.

The above-described embodiments are not limited to images obtained by an electronic endoscope, and may be applied to images obtained by a fibre-pass-1-wave (or rigid endoscope).

Further, in the third embodiment, the maximum value and minimum value are calculated from the temperature distribution, but at that time, the value at a predetermined number, for example, 5% of the total number of pixels, may be returned to the maximum value and minimum value. .

Further, in 5 steps of density level (for 0 to 255), the 1fi1 degree value from the center may be returned to the maximum value and minimum value.

In the 4.5th embodiment, the conversion daple is composed of smooth curves, but it may be approximated by two or more straight lines.

In the fifth embodiment, a conversion processing section 82 is provided after the gamma circuit 3.
However, even if the gamma circuit 3 is left behind, it is still not possible.

In the fifth embodiment, analog signals are used in a part of the conversion processing section 82, but all may be digital signals.

In the embodiment of the present application, an RG13 signal is used, but a color difference signal or brightness/saturation/hue signal may be used.

The value of p, which determines the characteristics of the conversion tape in the fourth and fifth embodiments, is a value that differs by C for each of the RGB signals.

In the third to fifth embodiments described above, it is possible to obtain an image with enhanced colors and the like without changing the average value of A with respect to the original signal and without changing the original color tone of the endoscope.

[Effects of the Invention] As described above, according to the present invention, a means for generating histograms for each signal is provided for a plurality of color signals constituting an image. C1 It is possible to obtain an effective amount of information even from wavelength ranges that have been difficult to use until now. .

[Brief explanation of the drawing]

FIGS. 1 to 8 relate to the first embodiment of the present invention.
The figure is a block diagram of the endoscopic image processing device of the first embodiment, FIG. 2 is the overall block diagram of the first embodiment, FIG. 3 is a flowchart of histogram extraction, and FIG. Figure 5 shows the histogram of an image obtained by performing ll1 degree conversion from the calculated His 1 Hegerum. Figure 6 shows the density level of two arbitrary points. FIG. 7 is a flowchart for flattening characteristic conversion, and FIG. 8 is an image C1q obtained by the processing in FIG. 7. FIG. 9 is a block diagram showing the main parts of the second embodiment of the present invention, FIG. 10 is a characteristic diagram of the histogram obtained by the second embodiment, and FIGS. 11 and 12 are Regarding the third embodiment of the present invention, FIG. 11 is a flowchart of image processing according to the third embodiment, and FIG. 12 shows R, G, and 8 images obtained by image processing. 13 to 15 relate to the fourth embodiment of the present invention, FIG. 13 is a flow chart of image processing in the fourth embodiment, and FIG. 14 is a flowchart of image processing in the fourth embodiment.
The figure is a characteristic diagram showing the function that performs the S degree characteristic, and Figure 15 is the 4th characteristic diagram.
R, G, B11Ti obtained by image processing in Example
1 (Fig. 3 to 18 are related to the fifth embodiment of the present invention, Fig. 16 is a block diagram showing the configuration of the fifth embodiment, and Fig. 17 is a diagram showing the configuration of the fifth embodiment). Figure 18 is a characteristic diagram for explaining the operation of the fifth embodiment, Figure 19 is a configuration diagram of the preceding example,
Figure 0 is a J characteristic diagram showing the luminance distribution and color signal level distribution in the prior example. 21... Endoscope image processing device 22... Electronic endoscope 27... Monitor 32... Light source section 52... Arithmetic processing device 54... External output device 26... Observation device 30 ...Image processing device 51...Working mechanism 53...Auxiliary storage device

Claims (1)

[Claims] a signal processing means for decomposing an image from an endoscope into a plurality of color signals; a means for calculating a histogram for each of the plurality of color signals; and a means for calculating a histogram for each of the plurality of color signals; 1. An endoscopic image processing device, comprising means for independently setting each output level of a signal.