JP2002207497A - Electronic endoscopic system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To stabilize the system control by speech of an electronic endoscopic system. SOLUTION: When a speech signal is confirmed to be inputted from a filter circuit connected with a bone transmission type microphone in a step S100, Whether this speech signal is the speech signal indicating the input start of a command for controlling the electronic endoscopic system by the speech or not is checked in a step S102. If the speech signal indicates the command input start and when the speech signal is confirmed to be successively inputted from the filter circuit in step S104, the matching processing of this speech signal and the command term of a vocabulary data base is performed in a step S106. Whether there is the data coinciding with the speech signal within the command terms of the vocabulary data base or not is checked and if there is the command term coinciding with the speech signal, the step is advanced to a step S110 and the signal is processed as the command for controlling the electronic endoscopic system and the corresponding control signal is outputted to a system controller.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to an electronic endoscope system capable of controlling a system by voice.

[0002]

2. Description of the Related Art Conventionally, an electronic endoscope system has a solid-state image pickup device such as a CCD at a distal end thereof, and an electronic scope inserted into a patient's body and a light source for supplying illumination light to an object to be observed. And an image signal processing unit that performs a predetermined process on the image signal photoelectrically converted by the solid-state imaging device. The image signal processed by the image signal processing unit is reproduced as an image by an output device such as a TV monitor connected to the image signal processing unit, and the doctor operates the electronic scope while checking the reproduced image on the TV monitor. .

During operation of the electronic scope, various controls are performed by input means such as a panel switch provided on the image signal processing unit and a keyboard connected to the image signal processing unit. For example, luminance adjustment according to the illuminance of the reproduction screen on the TV monitor, color adjustment according to the observation site, and the like are performed. However, performing such fine adjustment while observing the inside of the patient's body hinders the original operation (technique) of the electronic scope, and is troublesome.

[0004]

In order to solve such a problem, there has been proposed an electronic endoscope system which operates an image signal processing unit and an output device by voice input.
However, a doctor who is operating the electronic scope has to exchange ordinary conversation with the patient, such as communicating with the patient about explanation of the examination status and giving instructions to the nurse. Therefore, in the voice control system, erroneous recognition may occur due to the voice input of these ordinary conversations, which may cause a malfunction of the electronic endoscope system.

An object of the present invention is to solve the above-mentioned problems, and an object of the present invention is to prevent a malfunction of an electronic endoscope system which can be operated by voice input due to erroneous voice recognition.

[0006]

An electronic endoscope system according to the present invention includes an image signal processing unit for performing predetermined processing on image information obtained from a scope, and a predetermined processing performed by the image signal processing unit. An electronic endoscope system including display means for displaying image information, wherein the utterance detection means for detecting an utterance of the operator, and the utterance information detected by the utterance detection means are controlled by a control command of the electronic endoscope system. When the utterance information detected by the command determination means for determining whether there is a command is present and the utterance information detected by the utterance detection means is a keyword indicating the start of command input, the command determination is performed on the utterance information detected by the utterance detection means following the keyword indicating the start of command input. The utterance information is determined to be a control command by the control means for executing the determination processing by the means and the command determination means. Et al., Characterized in that it comprises a command output means for outputting a control signal corresponding to the control command.

[0007] Preferably, the command determination means includes a database storing predetermined command terms, and compares the utterance information with the command terms stored in the database.

Optionally, the control means causes the command determination means to execute determination processing on the utterance information detected by the utterance detection means until the utterance detection means detects a keyword indicating the end of command input.

[0009] Preferably, the utterance detecting means includes a close-talking microphone and a bone conduction microphone.

An electronic endoscope system according to the present invention displays an image signal processing unit for performing predetermined processing on image information obtained from a scope, and displays image information subjected to predetermined processing by the image signal processing unit. An electronic endoscope system comprising: an utterance detecting means for detecting an utterance of an operator; and determining whether utterance information detected by the utterance detecting means is a control command of the electronic endoscope system. A mode switching means for switching between a first mode for executing the control command when it is determined to be a control command and a second mode for not determining whether the utterance information is a control command is provided. .

Preferably, the mode switching means switches to the first mode when the utterance information is a keyword indicating the start of command input, and switches to the second mode when the utterance information is a keyword indicating end of command input.

As described above, according to the present invention, the command determination process is performed on the voice information detected by the utterance detection means after the keyword indicating the command input is input. In other words, depending on whether a keyword has been entered,
A mode in which the command determination process is executed and a mode in which the command determination process is not executed are distinguished. Therefore, a situation in which a normal conversation with a patient or the like is performed and a situation in which a command is input by voice are clearly recognized by the operator, and an opportunity for the operator to utter a vocabulary other than a control command in a mode in which the command determination process is executed. Decrease. As a result, for example, a problem in which a normal conversation with an operator and a patient or the like is erroneously recognized as a control command is eliminated, and the reliability of the system is improved.

[0013] If the configuration is such that the command determination process is performed on the voice information detected by the utterance detection means between the time when the keyword indicating the start of command input is input and the time when the keyword indicating the end of command input is input. In addition, the operator can execute a plurality of commands collectively, and the operability is further improved.

[0014] If the command determination means is provided with a database in which predetermined command terms are stored, and the utterance information is compared with the command terms stored in the database, the accuracy of recognition of input voice information is improved. . In addition, when a new control function is added to the electronic endoscope system, the command term corresponding to the control function is added to the database, so that the electronic endoscope system can be flexibly handled.

[0015]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram of an electronic endoscope system to which the first embodiment according to the present invention is applied.
The scope 10 has a flexible conduit (flexible tube) and is detachably connected to the image signal processing unit 20. Scope 1
An imaging sensor 11 including an objective optical system and a CCD image sensor is provided at the front end side of the reference numeral 0. When the scope 10 is connected to the image signal processing unit 20, the image sensor 11
Is connected to a first-stage signal processing circuit 24 of the image signal processing unit 20 via a CCD buffer circuit (not shown). A light guide 12 is inserted into the scope 10. The light emitting end of the light guide 12 extends to the tip of the scope 10.

The operation unit 13 of the scope 10 is provided with various operation buttons such as a freeze button for stopping a moving image, a still image copy button, and a recording button. By appropriately operating these buttons, an image signal to be processed in the image signal processing unit 20 is recorded.

The system controller 21 of the image signal processing unit 20 is a microcomputer that controls the entire electronic endoscope. That is, the system controller 21
Is a central processing unit (CPU), a program for executing various routines, a read-only memory (ROM) for storing constants and the like, and a write / read for temporarily storing data and the like.
It comprises a readable memory (RAM).

When the scope 10 is connected to the image signal processing unit 20, the incident end of the light guide 12 is provided with a white light source (not shown) such as a xenon lamp or a halogen lamp provided in the image signal processing unit 20. Optically connected to the light source unit 22. Light guide 12
A stop (not shown) for adjusting the light amount of a light beam emitted from the white light source of the light source unit 22 and incident on the incident end of the light guide 12 is provided between the incident end of the light source unit 22 and the light source unit 22. A condensing lens (not shown) for condensing light on the incident end of the light guide 12 is interposed.

The image signal processing unit 20 is provided with a front panel 23. On the front panel 23, various indicator lights and a power switch (S) for switching ON / OFF of a main power supply (not shown) of the image signal processing unit 20 are provided.
W), various switches such as a lighting switch (SW) for controlling lighting of the white light source of the light source unit 22 are provided.

The system controller 21 outputs a control signal to a lamp power supply circuit (not shown) of the light source section 22 based on a signal from the lighting SW. System controller 21
The power supply to the white light source described above is appropriately controlled by the lamp power supply circuit in accordance with the control signal from.

Further, in the light source section 22, a rotary RGB color filter (not shown) is interposed on the light emission side of the above-mentioned white light source. The rotary RGB color filter has a red filter, a green filter, and a blue filter provided at predetermined intervals in a circumferential direction of a disk-shaped support member, and is rotated by a drive motor such as a servo motor or a stepping motor. . When the rotary RGB color filter is rotated, red light, green light and blue light are sequentially emitted from the end face of the light emitting end of the light guide 12, and the object to be observed is sequentially illuminated with red light, green light and blue light. And the optical observation object image of each color is
One objective optical system sequentially forms an image on the light receiving surface of the CCD image sensor. The imaging sensor 11 photoelectrically converts each color optical observation object image formed on the light receiving surface of the CCD image sensor into an analog pixel signal for one frame,
The analog pixel signals for one frame of each color are sequentially read from the image sensor 11. Such an imaging sensor 1
Reading of the analog pixel signal from 1 is performed by a CCD driver 14 provided in the scope 10.

The CCD driver 14 of the scope 10 reads out one frame of analog pixel signals of each color of red (R), green (G), and blue (B) from the image sensor 11, and sends it to the first-stage signal processing circuit 24. Is entered. The first-stage signal processing circuit 24 includes a preamplifier, a band-limiting video filter, a gamma circuit, and the like, and performs predetermined image processing such as amplification of input analog pixel signals and gamma correction.

An analog pixel signal for one frame of each color processed by the first-stage signal processing circuit 24 is converted into a digital pixel signal by an analog / digital (A / D) converter (not shown), and an image memory is provided for each color. 25. The digital pixel signals of each color stored in the image memory 25 are read simultaneously, and a horizontal synchronizing signal, a vertical synchronizing signal, and the like are added to the read digital pixel signals of each color. That is,
Digital pixel signals of three primary colors for one frame are output from the image memory 25 as color digital video signals (R, G, B). The color digital video signal of each color is subjected to processing such as enlargement, reduction, noise reduction, etc., and is converted into a color analog video signal for one frame by a digital / analog (D / A) converter (not shown). , Is input to the subsequent-stage video signal processing circuit 26.
The timing of conversion in the A / D converter and the D / A converter, taking in of the pixel signals from the image memory 25, generation of the synchronization signal, and the like are controlled by the timing controller 27.

The output level of the color analog video signal for one frame of each color is appropriately adjusted after passing through a low-pass filter in a subsequent-stage video signal processing circuit 26 and sent to an analog video printer 30 and an analog RGB monitor 31. , And the optical RGB image of the object under observation is reproduced as a color image on the analog RGB monitor 31.

A keyboard 40 is connected to the image signal processing unit 20 as command input means.
For example, when a command for finely adjusting the screen of the analog RGB monitor 31 is input using the keyboard 40, the command is input to the system controller 21. A control signal corresponding to the command is transmitted from the system controller 21 to the analog RGB monitor 31, and fine adjustment such as illuminance adjustment and contour emphasis on the screen of the analog RGB monitor 31 is performed.

Further, the image signal processing unit 20 has, as command input means, a bone conduction type microphone 50 for directly detecting vibration of the skull when the operator speaks, and an operator provided at the mouth of the operator to speak. A close-talking microphone 60 that collects and detects the sound of the event is connected. The signals detected from the bone conduction type microphone 50 and the close-talking type microphone 60 are transmitted to the image signal processing unit 2.
0 is input to the speech recognition circuit 28.

FIG. 2 is a circuit diagram of the speech recognition circuit 28. The audio signal detected by the bone conduction microphone 50 is input to the filter circuit 281, and the signal characteristics are corrected based on a predetermined correction coefficient. Filter circuit 28
1 and the close-talking microphone 60
Is output to the comparator 282.
The comparator 282 performs a subtraction process of the two audio signals, and outputs a control signal for adjusting the above-described correction coefficient in the filter circuit 281 to the filter circuit 281 so that the result becomes substantially zero. The bone conduction type microphone 50 does not collect so-called background noise other than the utterance of the operator, but has a characteristic that the gain in a high sound range is low. Therefore, by adopting the above-described circuit configuration using the close-talking type microphone 60 together with the bone conduction type microphone 50, it is possible to prevent the background information from being collected and to prevent the sound information from being lost in the high frequency range. As a result, a highly reliable audio signal is output from the filter circuit 281.

The audio signal output from the filter circuit 281 is input to the word matching circuit 283. The vocabulary database 284 stores command terms of control commands. The word matching circuit 283 performs a matching process between the input speech signal and the command term stored in the vocabulary database 284 under predetermined conditions.

Table 1 shows correspondence between command terms stored in the vocabulary database 284 and control operations of the system controller 21.

[0030]

[Table 1]

In the table, item numbers (1) and (2) are analog RGB
This is a command term relating to the brightness adjustment of the screen of the monitor 31. When a voice “Kidappu” is input, the brightness of the screen is increased by a predetermined amount. Quantitative down. That is, the luminance changes stepwise each time a command term relating to luminance is input by voice.

In the table, item numbers (3) and (4) are analog RGB
A command term relating to color adjustment of an image reproduced on the screen of the monitor 31. When a voice "Akaappu" is input, the ratio of the red component in the reproduced image is increased by a predetermined amount, and a voice "Akadaun" is generated. When input, the ratio of the red component is reduced by a predetermined amount. That is, each time a command term relating to red color adjustment is input by voice, the ratio of the red component in the reproduced image changes stepwise.

In the table, item numbers (5) and (6) are command terms relating to the adjustment of the outline emphasis of the reproduced image. When the voice "Enhance" is input, the outline emphasis of the reproduced image is increased by a predetermined amount. , "Enhansuro", the contour enhancement of the reproduced image is reduced by a predetermined amount. That is, every time a command term relating to the adjustment of the contour enhancement is input by voice, the contour enhancement in the reproduced image changes stepwise.

In the table, item numbers (7) and (8) are command terms relating to the operation of a pump for supplying cleaning water to a water supply nozzle (not shown) provided on the scope 10, and the voice of "Ponpon" is output. When it is input, the pump operates and the washing water is sent from the water supply nozzle, and when the sound of “Panpoff” is input, the pump stops and the supply of the washing water also stops.

In the table, item numbers (9) and (10) correspond to the light source unit 22.
Is a command term related to ON / OFF control of the white light source, and when the sound “ranpion” is inputted, the above-mentioned white light source is turned on, the object is illuminated with light, and “ranpoff” is called. When sound is input, the white light source is turned off.

The list of command terms shown in Table 1 is merely an example, and does not exclude other control commands by voice. For example, as a command for controlling the color adjustment of the reproduced image, a command term related to the intensity of blue and green may be further stored.

FIG. 3 is a flowchart showing a processing procedure in the word matching circuit 283 when the operator speaks. If it is confirmed in step S100 that an audio signal has been input from the filter circuit 281, step S10
At 2, it is checked whether the voice signal is a voice signal indicating the start of command input by voice for controlling the electronic endoscope system. In the first embodiment, “komando” is recognized as a voice signal indicating the start of voice input of a command. If it is determined that the input voice signal does not match "Komando", the voice continuously emitted from the operator is determined to be a normal conversation voice, control returns to step S100, and the word collation processing described later is performed. Not. That is,
Unless a voice corresponding to “Komand” is uttered by the operator, system control by voice input is not performed.

If it is confirmed in step S102 that the input audio signal matches "komando",
Proceed to step S104. If it is confirmed in step S104 that the next audio signal following "komando" has been input from the filter circuit 281, the process proceeds to step S106. In step S106, a matching process is performed between the voice signal following "komando" and the command terms (see Table 1) stored in the vocabulary database 284.

In step S108, it is checked whether any of the command terms stored in the vocabulary database 284 matches the voice signal acquired in step S104. If there is a command term in the vocabulary database 284 that matches the voice signal, step S
Proceeding to 110, it is processed as a command for controlling the electronic endoscope system. That is, the corresponding control signal is output from the word matching circuit 283 to the system controller 21.

In step S108, if it is confirmed that the command term matching the voice signal is not stored in the vocabulary database 284, the voice signal is not a control command of the electronic endoscope system, so that the above-mentioned step S110 is performed. Is not executed, and the process returns to step S100.

For example, if the operator feels that the image reproduced on the analog RGB monitor 31 is dark and difficult to observe while operating the scope 10, he speaks “komando” and then “kidoappu”. Thereby, the brightness of the reproduced image on the monitor 31 is increased.

As described above, in the first embodiment,
When the voice of “Komand” is input, a command determination process is performed on a voice signal that is subsequently input. In other words, as long as the sound of “Kandando” is not input, any sound input is not reflected in the system control of the electronic endoscope system. In other words, the operator only needs to say “comand” and then continue to say command terms when trying to execute voice-based system control, or the patient or nurse if he does not say “comand”. Even if the user has a normal conversation with, words in the conversation are not erroneously recognized as command terms.

FIG. 4 is a flowchart showing a processing procedure in the word matching circuit of the electronic endoscope system to which the second embodiment according to the present invention is applied. The system configuration of the electronic endoscope system to which the second embodiment is applied and the configuration of the voice recognition circuit are the same as those in FIGS. Step S
The processing from 200 to S208 is the same as the processing from steps S100 to S106 in the flowchart of FIG. In other words, when "komando" is uttered, matching processing between the subsequent voice signal and the command term in the vocabulary database 284 is performed, and it is checked whether a command term matching the voice signal exists.

In step S208, "komando"
If it is confirmed that the voice signal input following the voice of the command matches one of the command terms stored in the vocabulary database 284, the process advances to step S210 to process the command as a command for controlling the electronic endoscope system. Is done. That is, a corresponding control signal is output from the word matching circuit 283 to the system controller 21. Next, the control returns to step S204, and the subsequent processing is repeated.

If it is confirmed in step S208 that there is no command term matching the voice signal, step S2 is executed.
Proceeding to 12, it is checked whether the voice signal indicates the end of command input by voice. In the second embodiment,
Recognize "SyuRyo" as a voice indicating the end of command input by voice. If it is determined in step S212 that the voice signal is “successful”, it is determined that the control of the electronic endoscope system by voice input is to be terminated. In step S214, the command input mode by voice is terminated. It returns to S200.

On the other hand, in step S212, when it is confirmed that the voice signal is not "successful", the voice signal is neither a control command nor a command completion by voice, that is, a command of the operator. It is determined to be a normal conversation. Therefore, the output of the control signal and the end of the command input mode are not executed, and the control returns to step S204 to wait for the input of a further audio signal. After the input of the audio signal, the subsequent processing is repeated.

As described above, in the second embodiment,
The command determination process is performed on all voice signals that are input after the voice indicating the start of command input is input and before the voice indicating the end of command input is input. In other words, when the keyword "komando" is uttered by the operator, the mode becomes a mode for performing a command determination process on the voice signal input thereafter, and when the keyword "syuryo" is uttered by the operator, A mode in which the command determination process is not performed on the audio signal input thereafter is set.

The operator can continuously perform the system control by voice by repeatedly uttering the command term until the utterance of “Korando” after the utterance of “Komando”. For example, if you utter "komando" and then utter "kidouun", "enhance yes", and "ponpoon" continuously, the analog RGB monitor 3
1 lower the brightness of the reproduced image and strengthen the edge enhancement,
Further, a series of operations of supplying the cleaning water to the water supply nozzle can be continuously performed until the user speaks "success".

Also, as in the first embodiment, the "komando"
As long as the operator does not speak, the words in the conversation will not be erroneously recognized as control commands even if the operator has a normal conversation with the patient or nurse. In addition, even in a mode in which a command determination process is performed on an input audio signal, if the user speaks “success”, the electronic endoscope system will not malfunction even if a normal conversation is subsequently made.

In the first and second embodiments, the keyword for starting the command determination processing is "Komando", and the keyword for ending the command determination processing in the second embodiment is "SyuRyo". It is not limited to. Other vocabulary that is easy for the operator to speak may be used as these keywords.

Further, in the electronic endoscope system to which the first and second embodiments are applied, image processing is performed by a frame sequential method using RGB color filters, but the present invention is not limited to this. The image sensor 11 may be configured to perform image processing by a simultaneous method using a CCD image sensor provided with a color chip filter.

[0052]

As described above, according to the present invention, it is possible to stably control the electronic endoscope system by voice input.

[Brief description of the drawings]

FIG. 1 is a block diagram of an electronic endoscope system to which a first embodiment according to the present invention is applied.

FIG. 2 is a configuration diagram illustrating a word recognition circuit of the electronic endoscope system according to the first embodiment.

FIG. 3 is a flowchart illustrating a processing procedure of word matching in the word recognition circuit of the first embodiment.

FIG. 4 is a flowchart showing a procedure of a word matching process in a word recognition circuit of an electronic endoscope system to which a second embodiment according to the present invention is applied;

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 10 Scope 11 Image sensor 12 Light guide 13 Operation part 14 CCD driver 20 Image signal processing unit 21 System controller 22 Light source part 23 Front panel switch 24 First stage signal processing circuit 25 Image memory 26 Second stage video signal processing circuit 27 Timing controller 28 Voice recognition circuit 31 Analog RGB monitor 50 Bone conduction microphone 60 Close-talking microphone 284 Vocabulary database

Claims (6)

[Claims] 1. An electronic endoscope system for performing predetermined processing on image information obtained from a scope and outputting the image information by an output device, wherein: an utterance detection unit that detects an utterance of an operator; A command determination unit that determines whether the utterance information detected by the detection unit is a control command of the electronic endoscope system; and wherein the utterance information detected by the utterance detection unit is a keyword indicating a command input start. Control means for executing a determination process by the command determination means for utterance information detected by the utterance detection means following a keyword indicating the start of command input; and determining that the utterance information is the control command by the command determination means Command output means for outputting a control signal corresponding to the control command, Electronic endoscope system according to claim. 2. The electronic apparatus according to claim 1, wherein the command determination unit includes a database storing predetermined command terms, and checks the utterance information with command terms stored in the database. Endoscope system. 3. The control means causes the command determination means to execute determination processing on utterance information detected by the utterance detection means until a keyword indicating the end of command input is detected by the utterance detection means. The electronic endoscope system according to claim 1, wherein: 4. The electronic endoscope system according to claim 1, wherein said utterance detection means includes a close-talking microphone and a bone conduction microphone. 5. An electronic endoscope comprising: an image signal processing unit that performs predetermined processing on image information obtained from a scope; and a display unit that displays the image information that has been subjected to predetermined processing by the image signal processing unit. An endoscope system, an utterance detection unit for detecting an utterance of an operator, and determining whether utterance information detected by the utterance detection unit is a control command of the electronic endoscope system, and determining that the utterance information is a control command. An electronic endoscope comprising: a mode switching unit that switches between a first mode for executing the control command when the utterance information is issued and a second mode for not determining whether the utterance information is the control command. system. 6. The mode switching means switches to the first mode when the utterance information is a keyword indicating the start of command input, and switches to the second mode when the utterance information is a keyword indicating end of command input. The electronic endoscope system according to claim 5, wherein: