JPH01212077A - Signal processing circuit for electronic endoscope - Google Patents

Abstract

PURPOSE:To apply for an animation picture and to obtain a picture good in S/N to a low illuminance by disposing the detecting means of the change in time of an image pickup signal and variably controlling the characteristic of a cyclic type noise removing filter. CONSTITUTION:A variable gain amplifier (multiplier) 11 for raising a gain to the object of the low illuminance and amplifying, a movement detector 12 for detecting the scale in the change of the time of a picture and the cyclic type noise removing filter 13 for variably controlling the characteristic of the filter according to the output of this movement detector 12 are provided. Namely, the noise removing filter capable of detecting the change in the time of the picture by the movement detector 12, deciding the coefficient of the filter 13 from this value and the gain setting value of a prestep amplifier 11 and being adapted to the brightness of the object and the movement thereof is constituted. Thereby, the deterioration of the picture to the animation picture is reduced and the picture good in the S/N to the object to the low illuminance can be obtained.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a signal processing circuit for an electronic endoscope that generates an image signal with a high S/N ratio even for a subject with low illuminance.

[Prior Art] In recent years, even in endoscopes, electrical signals photoelectrically converted using a solid-state imaging device such as a COD are transmitted using an electric cable, without transmitting an optical image using an image guide formed of a fiber bundle. Electronic endoscopes (hereinafter referred to as electronic endoscopes) that display color information on a monitor device have come into practical use.

Since the signal captured (imaged) by this electronic endoscope is an electrical signal, it is easy to record it on a VTR, etc., or process the signal to create an image that is easy to diagnose, so it is likely to become more popular in the future. .

By the way, even in electronic endoscopes, it is necessary for the insertion part to be as small as possible in order to reduce the pain caused to the patient, but when COD is used in an endoscope, there is a considerable amount of damage at the tip of the insertion part. This occupies space, and in order to maintain the outer diameter within a certain value, the effective cross-sectional area of the light guide becomes small, which may result in insufficient illumination light σ. In addition, due to the demand for smaller diameters, it is desirable that the light guide has a small cross-sectional area, and depending on the usage situation, the amount of illumination light may be insufficient, and the object may be illuminated only at low illuminance. Further, even if sufficient illumination can be provided in normal use, when performing fluorescence observation using laser beam irradiation, the illumination will be of very low intensity, and the signal level of the jl image signal will be extremely low.

By the way, since the random noise component included in the COD output signal takes a certain constant value regardless of the illuminance of the subject, the S/N of the COD output signal becomes a smaller value as the illuminance of the subject is lower. Therefore, when increasing the amplification of the signal processing circuit to observe a subject in low illumination,
Even if the average voltage level of the image signal increases, the signal S/
Since N does not change, the observed image is buried in image noise and becomes difficult to see.

Therefore, when imaging a subject in low illumination, the amplification degree of the signal processing circuit is increased and the noise of the image signal is removed.
Use means to improve /N.

Known methods for removing image noise include (1) a method of taking an average of neighboring pixels, and (2) a method of taking an average of data of the same pixel for a certain period of time.

Method (1) utilizes the spatial correlation of pixels constituting an image, and is effective in removing noise from smooth images. (
Method 2) utilizes the time correlation of images and is very effective in removing noise from images with small 81M changes, that is, still images. In either method, when the number of pixels to be added (the number of images in case (2)) is n, the noise level can be reduced to 1/C.

In addition, method (2), as shown in Japanese Patent Application No. 155794/1982 or No. 109577/1982, uses multiple frame memories to delay the image by one frame scanning time and then performs averaging. It can be realized with.

A conventional example using method (2) above is shown in FIG.

The imaging signal from the CCD 1 is amplified and unnecessary high frequencies are removed in the preprocessing section 2, and then converted into a digital signal by the A/D converter 3, and then sent to the memory controller 4.
The data is stored in the first memory 6a to the fifth memory 6e via the multiplexer 5, which performs switching. The data in these memories 6a to 6e are simultaneously read out, added by an adder 7, converted to an analog signal by a D/A converter 8, converted to a video signal such as the NTSC system by a host processing unit 9, and then displayed on a monitor. Displayed as 10.

[Problems to be Solved by the Invention] However, method (1) essentially applies a spatial low-pass filter to the input image, and the sharpness of the output image deteriorates.

On the other hand, method (2) utilizes the temporal correlation of images as mentioned above, so it is possible to use images with small temporal changes.
In other words, sharpness is maintained for still images, but for images with large temporal changes, that is, moving images, image blur etc. may occur.
The resulting image looks like a photo taken with a slow shutter speed of a fast subject. In addition, the conventional example shown in FIG. 13 using method (2) requires multiple frame memories, and for example, in order to improve the S/N by 6 dB for a still image, it is necessary to frame one frame of four frames. It has the disadvantage that memory must be prepared, the scale of the device becomes large (and it becomes complicated).

The present invention has been made in view of the above-mentioned points, and provides a signal processing circuit for an electronic endoscope that has little deterioration of images for moving images and that can obtain images with good S/N even for subjects in low illuminance. The purpose is to

[Means and effects for solving the problem] As shown in the conceptual diagram of FIG. It is composed of three main elements: a motion detector 12 that detects the magnitude of a change in motion, and a recursive noise removal filter 13 whose filter characteristics are variably controlled by the output of the motion detector 12.

The object is imaged by the CCD 14, and the image signal is amplified by the multiplier 11 through the preprocessing circuit 15, then inputted to the noise removal filter 13, and after noise removal is inputted to a monitor etc. via the host processing circuit 16. . The noise removal filter 13 includes a one-frame delay element 17, a multiplier 19 that is multiplied by the coefficient of a coefficient setter 18 determined by the output of the motion detector 12 and the output of the delay element 17, and both multipliers 11. It consists of an adder 20 to which the output of .19 is added, and a multiplier 22 to which this addition output and the output of the coefficient setter 21 are multiplied.

An adder 23 extracts the difference between the output of the multiplier 11 and the output of the delay element 17, and the difference is input to the motion detector 12.

Note that the output of this delay element 17 is multiplied by a coefficient of the coefficient setter 17 in a multiplier 24 and inputted to an adder 23 .

Note that a multiplier 11 that amplifies the output of the preprocessing circuit 15
is the output of the gain setter 25 and its multiplication coefficient (gain)
is determined.

The noise removal filter 13 configured in a cyclic manner performs time averaging of images with a certain weighting, and a diagram of the principle of this filter 13 is shown in FIG. Here, the one-frame delay element 17 in FIG. 1 is a delay operator 17' represented by z -1, and the multiplier 19 is a coefficient (multiplier) unit 1 represented by α.
The 911 adder 20 includes an adder (operator) 20' and a multiplier 22.
corresponds to the coefficient (multiplier) 22 represented by β.

The input image to the filter at time k is uk.

Letting the output image be yk and the internal state of the filter as xk,
The filter of FIG. 2 can be expressed as follows.

x, to=αxk +uk ・・・・・・■yk
=βxk・・・・・・■However, uk, xk,
yk relates to a certain same pixel in the image, and 0≦α≦1.

From the above formula 0 k　H-1.

xk = Σα ul + αxO...■゛Art 1. Here, x□ is the initial value of =o.

In other words, as the value of the weighting coefficient α is decreased, this filter works to give more importance to the current image signal, and α
As the value of is brought closer to 1, more emphasis is placed on past image signals. Therefore, when the time change of the input image is small, that is, when it is a still image, α is brought closer to 1, and when the time change is large, that is, when it is a video, α is reduced, so that it can be used for both videos and still images. It can be used as a noise filter.

Furthermore, when the subject is bright, that is, when the preamplifier gain is small, the S/N of the input image is better than when it is dark, so α is made smaller to make the filter characteristics suitable for moving images.

In order to function in this way, in the present invention, the motion detector 1
2, the temporal change in the image is detected, and the coefficients of the filter 13 are determined from this value and the gain setting value of the pre-amplifier 11. This allows the noise to be adjusted to both the brightness and movement of the subject. A removal filter is configured to perform signal processing suitable for electronic endoscopes.

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

FIG. 3 shows an electronic endoscope apparatus equipped with a first embodiment of the present invention, and FIG. 4 shows the configuration of a motion detector.

As shown in FIG. 3, an electronic endoscope device 31 equipped with the signal processing device of the first embodiment is an electronic endoscope (
Also referred to as electronic scope. ) 32 and this electronic scope 3
A signal processing circuit 33 of the first embodiment that processes the image signal of No. 2, a light source section 34 that supplies illumination light to the electronic scope 32, and a monitor (not shown) that displays the signal-processed R, G, and B outputs. Consisting of

The electronic scope 32 has an elongated insertion portion for easy insertion into a body cavity, and an objective lens 3 is provided at the distal end of the insertion portion.
5 and a C0D 36 whose imaging plane is arranged on the focal plane of the objective lens 35 is housed. The photoelectric conversion output of this CCD 36 is inputted to an amplifier 38 forming a signal processing circuit 33 via a preprocessing circuit 37.

Further, a light guide 39 is inserted into the electronic scope 32. In this way, the illumination light from the light source lamp 41 housed in the light source section 34 is reflected by the reflecting mirror 42 into a parallel light beam, and the R and G of the rotating filter 43 interposed on the optical path of this parallel light beam.

The R, G, and B illumination lights are color-separated into R, G, and B illumination lights by an eight-color transmission filter, and are condensed by a condenser lens 44 and projected onto a light guide 39 .

Note that the rotary filter 43 is rotationally driven by a motor 45. This light guide 39 transmits illumination light,
The subject is sequentially illuminated with R, G, and B light, and the R, G, and B
The signal imaged using the CCD 36 under color sequential illumination becomes an R, G, and B color sequential signal, which forms a signal processing circuit 33 and an amplifier whose gain is set by an external gain setting device 46. 38.

The signal amplified by the amplifier 38 is converted into a digital quantity by an A/D converter 47 and then processed.

The data is written into the R, G, B memories 4B, 49.degree. 50 via the changeover switch SW1 which is switched in synchronization with the G and B color sequential signals. When one frame is written to each of these R, G, and B memories 48, 49, and 50, they are simultaneously read out, passed through the D/A converter 51, returned to an analog signal, and passed through the host process circuit 52, respectively.
, G, and B color signals and output to an external device (not shown), such as a monitor.

On the other hand, the switch SW1 is a fluorescent observation switch SW2.
3 flilJ, and input the output of the A/D converter 47 to the cyclic noise removal filter 53.

In this case, the changeover switch SW3 provided on the output side of the noise removal filter 53 is also switched at the same time, and the output of the noise removal filter 53 is connected to the output end of the R, G, B memory 48, 49, 50. This allows input to each D/A converter 51. Note that, as the recursive noise removal filter 53, the noise removal filter 14 shown in FIG. 1, for example, is used. The output of the noise removal filter 53 is inputted to the motion detector 55 together with the output of the A/D converter 47 via the coefficient unit 54, the amount of motion is detected from both outputs, and the detected output and the output of the gain setter 46 are is input to the filter coefficient setter 56. The output of this filter coefficient setter 56 is the noise removal filter 5
3 and controls the filter characteristics of the noise removal filter 53.

Incidentally, the motion detector 55 is configured, for example, as shown in FIG.
/D converter 47 output to noise removal filter 53
It consists of an adder 57 that subtracts or inverts and adds the output on the side, and a comparator 58 that compares the absolute value of this output with a similar level and outputs the result.

In the first embodiment configured in this way, the imaging signal is displayed without using the recursive noise removal filter 53 during normal endoscopic observation, while the noise removal filter 53 is used during fluorescence observation. ing. The operation of this first embodiment will be explained below.

During normal endoscopic observation, the light source lamp 41 of the light source section 34
is turned on, and the white light from this lamp 41 is passed through the rotating filter 4.
By rotating the light guide 39, the illumination light is sequentially color-separated into RlG and B illumination lights and is incident on the incident end surface of the light guide 39. The object is then illuminated with the R, G, B color sequential illumination light emitted from the tip surface of the light guide 39, and the object image illuminated with the color sequential illumination light is focused on the imaging surface of the CCD 36. This subject image is photoelectrically converted into an electrical signal, which is input to an amplifier 38 forming a signal processing circuit 33 via a preprocessing circuit 37 . This amplifier 38 is set to a low gain by a gain setter 46 during normal endoscopic observation, and the amplified imaging signal is converted into a digital quantity by an A/D converter 47, and the R, G, and B colors are sequentially illuminated. The data is sequentially written into the RlG and B memories 48, 49, and 50 via the changeover switch SW1, which is switched in synchronization with the . When each frame has been written to the R, G, and B memories 48, 49, and 50, they are simultaneously read out and converted into analog signals by the D/A converter 51.
Further, the signals are converted into R, G, and B color signals through a post-processing circuit 52, and input to a monitor (not shown), where they are displayed in color.

On the other hand, during fluorescence observation, in which a fluorescent agent is applied to the subject and the fluorescence is observed by exciting it with a laser beam, etc., the fluorescence observation switch SW2 provided in the operation section of the electronic scope 32 or the signal processing circuit 33 is turned on. do. When the switch SW2 is turned on, the switch SW1 is also connected to the noise removal filter 53 (that is, the output of the A/D converter 47 is input to the noise removal filter 53), and the switch SW3 is also turned on. , noise removal filter 53
is activated, and its output is equally input to the R, G, and B signal systems.

When the fluorescence observation switch SW2 is turned on, the set value of the gain setter 46 changes to a large value, increasing the gain of the amplifier 38 and working to change the coefficients of the filter coefficient setter 56. On the other hand, the filter output delayed by one frame time from the noise removal filter 53 is input to the motion detector 55 together with the output of the A/D converter 47, and the filter coefficients are determined according to the magnitude of the temporal change in the image. It acts on the filter coefficient setter 56 to change it. In other words, when the amount of motion is small, the performance of the noise removal filter 53 can be increased to improve the S/N ratio, and when the amount of motion is large, the performance of the filter can be decreased to suppress image deterioration due to motion. .

FIG. 5 shows a second embodiment of the invention.

In this second embodiment, the imaging signal of the CCD 36 is inputted to a cyclic noise removal filter 61 via a preprocessing circuit 37 and an amplifier 38, and the output of this noise removal filter 61 is sent via a host processing circuit 62 to a monitor or a subsequent processing circuit. is input to.

This noise removal filter 61 has a slightly different configuration from the noise removal filter 53 shown in FIG. 2, in which the coefficient unit 22' in FIG. This is connected to a coefficient multiplier 63 interposed between the input terminal and the input terminal of the input terminal. Also, the coefficient unit 19 on the output end side of the delay operator 17'
' is a coefficient unit (or multiplier) 6 whose coefficient is α/(1-α)
It's 4th.

The output of the delay operator 17' is 1/1 along with the coefficient unit 64.
The signal is inputted to a motion detector 66 together with the output of the amplifier 38 via a coefficient unit 65 for the coefficient α2.

The output of the motion detector 66 is input to the coefficient switch 68 together with the output of the gain switch 67.
, the coefficients 1-α, α/(1-
The value of α) is varied.

For example, when the amount of motion detected by the motion detector 66 is large, the coefficient α is small; on the other hand, the smaller the amount of motion is, the larger α is, and the random noise removal function of the filter 61 is increased.

The characteristics of the noise removal filter 61 shown in FIG. 5 are as follows: coefficient α
It exhibits the following characteristics depending on the

From FIG. 5, the internal state xk of the filter and the filter output yk are expressed as follows.

xk −αxk−1+ (1−a>uk ・・・■
′yk=xk ・・・■′■′
2 transformation, (1-cxZ-1) X(z) = (1-α) U(z
) =■′, so from equation 0, the frequency characteristic H(jw) of the filter becomes,
Therefore. The frequency characteristic H(jW) of this filter is α-0
, 9, 0, 95, 0, 99.
It is shown in the figure and Table 1. Furthermore, O~π/T and π/T~2π/T
becomes symmetrical.

(Space below) Table 1 Coefficients and filter characteristics (Space below) As can be seen from Figure 6 above, it is clear that this is a filter that selectively passes components that repeat at a period T (image repetition period). . This shows the characteristic that the still image component is passed through as is, and the noise that is unrelated to the period T, such as the random noise component, is cut and eliminated (passage prevention), and this characteristic becomes sharper as α→1 approaches. become

By the way, the motion detector 66 in the second embodiment is as follows:
The structure shown in FIG. 7 can be used. Note that this is similar to that shown in FIG. 4, but the input coefficients are different. As shown in Fig. 7, if the output of addition 2S57 is e, then e-uk - (1/cX2) xk-1-...■Now, there is no change in the input image from time t-1 to t-k. Then, u k = u 1- const(for i=1.
2. −, k) −・・■. Then, from equation 01, it becomes xk - (Zk e=uk −1el °°゛■
becomes.

That is, the input image at time t=1 is compared with the input image at initial time t=1, and the difference is output. In other words, motion is detected from the difference signal between the image at time t and the image before t.

The comparator after this adder 57 takes let,
Compare it with the reference value, generate a switching signal, and switch the coefficient switch 6.
Input on the 8 side.

Table 2 shows an example of the relationship between the motion detection output value and the coefficient α set by this output value.

Note that in Table 2, α-0 is equivalent to the case where no filter is provided.

The coefficient switching described above is determined by the detection output value or detection signal of the motion detector 66, and is also performed by the gain setting level. That is, when the gain is large, the filter characteristic is sharpened, and when the gain is small, the filter characteristic is lowered, or as in the first embodiment, it is turned off to give priority to the real-time video signal.

FIG. 8 shows a signal processing R circuit 81 of a third embodiment constructed using a real-time memory and an autoregressive memory.

In this embodiment, the gain of the blob O-cess section 37' can be controlled by a gain setter 82, and the gain is increased during fluorescence observation or in the case of low-intensity illumination.

The outputs of the R, G, and B memories 48, 49, and 50 are input to adders 84, 85, and 86 that form a cyclic frame correlation filter 83 as a noise removal filter, and are input to the autoregressive memories 87, 88, and 86, respectively. 89 output is interlocking switch S
W and functions as a frame correlation filter. The outputs of the respective adders 84, 85, and 86 are inputted to the (1-β)/β coefficient coefficient units 94 and 95.96 at the next stage via coefficient units 91 and 92.93 for coefficients of β. Each coefficient unit 94
．． The output of 95.96 becomes R, G, B output.

Coefficient units 91, 94; 92.95; 93.96 are the outputs of the coefficient setter 97, and the value of each coefficient β is set therein.

Note that when the switch SW is off, the coefficient multiplier 91°94
: 92, 95 : This is the same as when 93, 96 are not present.

In this third embodiment, during fluorescence observation, the gain of the preprocessing circuit 37' is increased and the switch SW of the filter 83 is turned on. At this time, real time memory 48.4
The image data stored in 9.50 is almost a state in which weak signals are buried in noise, and the image becomes meaningless without a filter. Since the filter 83 integrates all the image data from the time it was turned on to the current time as described above, it is possible to remove the white noise component and reproduce the image that was previously buried in noise.

Also, at this time, R, G, and B lighting are off, so R, G,
Images of each B system whose time is shifted by 1/(3x frame frequency) are averaged on the monitor. This also has a great filtering effect. In this case, R, G,
Add B output.

Furthermore, this embodiment can also be applied to the case where the switch SW is turned on and color imaging is performed, not only in the case of fluorescence observation but also in the case of very weak low-intensity R2O, B illumination.

Further, in this embodiment, by changing the value of the coefficient β for each color of illumination, it is possible to emphasize a specific color.

Further, in the embodiment shown in FIG. 8, the switch S
Filter 83 by switching W between scrutiny mode and general mode
You may use them differently.

In other words, the recursive frame correlation filter 83 has a great effect on still images, but when the subject moves quickly, image quality deterioration such as image blur occurs. Therefore, when observing the affected area carefully as in the observation mode, the angle movement of the scope is minimized, the patient side is also kept stationary, and the switch SW of the filter 83 is turned on. on the other hand,
In the normal inspection mode, the switch sw of the filter 83 is turned off for observation. Switching between these two modes is
As shown in the figure, the switch S provided in the scope operation section 98
All you have to do is use W.

In other words, a normal video is output for a moving image, and an image with a good S/N ratio is output by filtering for a still image.

In addition, in FIG. 8, when capturing and displaying images in monochrome, such as in fluorescence observation, the three switches may not be linked, but one switch and one regression filter may be used. .

FIG. 10 shows a fourth embodiment of the invention.

In this embodiment, for example, the output of an A/D converter is input to a real-time memory section 101, and the output of this memory section 101 is added to the output of an autoregressive memory section 103 in an adder 102. The output of this adder 102 is input to the autoregressive memory unit 103 via the coefficient unit 104 for the coefficient of β, and is output to the next stage side via the coefficient unit 105 for the coefficient of (1-β)/β. .

Both of the memory sections 101 and 103 are each formed by two memories 101a and 101b: 103a and 103b, and an interlocking switch 81: S2.

Both of these memories 101a, 101b and 103a, 1
03b is used by alternating between write and read in one frame.

As shown in FIG. 10, the input image at time k is uk,
When the filtered image is xk and the output image is Vk, this filter has xk-βX k-1+βuk Vk-(1-β) xk /β.

Similar to each of the above-mentioned embodiments, this embodiment also adds image data in the time axis direction, averages the weighting, and smoothes the time axis direction, so it is possible to improve the image quality of still images without deteriorating the resolution. Effective in removing white noise. From the first equation above, xk=βxO+βsβ−nun k÷1. Here, xO is the initial value of the filter.

From this equation, it can be seen that in the case of a recursive filter, past image data is saved. In other words, this filter works as if it were integrating an image in the time axis direction, so it is very effective when used for still image video signals that are weak and buried in white noise.

FIG. 11 shows the main part of a fifth embodiment of the present invention.

In this embodiment, for example, the output of the preprocessing circuit 37 in FIG. 3 is input to the automatic gain setter 111.

This automatic gain enactor 111 is connected to the A/D converter 11
2 and a gain setter 113 whose coefficient is set by the output of the A/D converter 112. When the output of the A/D converter 112 is large, the gain setting coefficient of the gain setter 113 becomes small, and the amplifier 38 The gain is automatically reduced.

Further, the output of this gain setter 113 is sent to the motion detector 55.
It is input to the filter coefficient setter 56 together with the output of .

The rest is the same as the first embodiment. According to this embodiment, the gain of the amplifier 38 is automatically reduced under sufficient illumination, and is increased when the illumination intensity is low.

FIG. 12 shows the main parts of a sixth embodiment of the present invention.

In this embodiment, for example, in the embodiment shown in FIG.
The output of the motion detector 55 is input to the switching controller 121, and the output of the switching controller 121 controls on/off of the noise removal filter 3, for example, as follows.

If the motion detected by the motion detector 55 is small, A
Switch SW4 and switch SW3 on the output side of /D converter 47 are kept in the on state. - On the other hand, when the amount of movement increases slightly, the on-state cycle is shortened to shorten the image accumulation time, thereby controlling deterioration due to image shift.

Incidentally, each of the embodiments described above can be partially combined to form different embodiments.

[Effects of the Invention] As described above, according to the present invention, a means for detecting temporal changes in an image signal is provided, and the characteristics of the recursive noise removal filter can be variably controlled, so that it can be applied to moving images as well. Furthermore, images with good S/N ratio can be obtained even in low light conditions.

[Brief explanation of the drawing]

FIGS. 1 to 4 relate to a first embodiment of the present invention.
Figure 2 is a conceptual diagram of the present invention, Figure 2 is a principle diagram of a recursive filter used in the first embodiment, Figure 3 is a configuration diagram of an electronic endoscope device equipped with the first embodiment, and Figure 4 is A block diagram of a motion detector, FIG. 5 is a block diagram showing a second embodiment of the present invention, FIG. 6 is a characteristic diagram showing filter characteristics of the second embodiment, and FIG. 7 is a block diagram showing a second embodiment of the present invention.
FIG. 8 is a configuration diagram of a motion detector in an embodiment, FIG. 8 is a configuration diagram of a third embodiment of the present invention, FIG. 11 is a schematic diagram showing the main parts of the fifth embodiment of the invention; FIG. 12 is a diagram showing the main parts of the sixth embodiment of the invention; FIG. FIG. 13 is a configuration diagram showing a conventional example. 11... Multiplier (amplifier) 12... Motion detector 13... Noise removal filter 14... C0D 17... 1 frame delay element 19... Multiplication B 20... Adder 24...
... Multiplier 32 ... Electronic scope 33 ...
・Signal processing circuit 34...Light source section 48.49.50
...Memory 53...Recursive noise removal filter 55...Motion detector 56...Filter coefficient setter Fig. 5 Fig. 6 Fig. 11 Procedure? rli official document (spontaneous) January 12, 1986 Mr. Kunio Ogawa, Commissioner of the Patent Office
,. 1. Indication of the case: 1985 Patent Application No. 313277
Relationship with Case No. 7 (Toshibe Shimoyama, Representative of the Patent Applicant) 7.? 1ltiE(7)7@”11010
\1. On the 7th line of the 5th page of the specification, it says “
``...host...'' [...post...
Correct to J. 2. In the first line of page 7 of the statement, "...host...
``...'' will be corrected to ``...post...''. 3. In the third line of page 16 of the statement, correct "...host..." to "...post...". Procedural form (method) 1. Display of case 1985 patent application No. 313277
No. 2, Title of the invention Signal processing circuit for electronic endoscope 3,
Relationship with the case of the person making the amendment Toshibe Shimoyama, representative of the patent applicant

Claims (2)

[Claims] (1) In an endoscope signal processing circuit that uses a solid-state image sensor as an imaging means and has a means for removing noise contained in an image signal output photoelectrically converted by the solid-state image sensor, a cyclic noise removing means and a current and a motion detecting means for detecting the movement of the image from the difference between the image signal of 1 frame and the image signal of one frame before, and the filter characteristics of the cyclic noise removing means are variably controlled by the output of the motion detecting means. A signal processing circuit for electronic endoscopes. (2) The cyclic noise removing means controls the filter characteristics thereof in conjunction with the amplification setting of an amplifier provided upstream of the cyclic noise removing means. signal processing circuit for electronic endoscopes.