JP2000078530A - Information recorder, information recording method and recording medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To input a keyword in voice and to record it in relation to video/audio data. SOLUTION: A voice recognition section 22 recognizes a word with respect to a voice entered by a voice entry section 21 by using a word dictionary 23 and gives the word to a dictionary retrieval section 24. The dictionary retrieval section 24 retrieves information of this word from the word dictionary 23 and gives the result of retrieval to a retrieval result display section 25. The retrieval result display section 25 displays plurality of retrieval results sequentially, a voice input section 26 confirms one retrieval result and gives it to an optical disk recorder recording section 27. The optical disk recorder recording section 27 records the confirmed retrieval result to an optical disk 28 in relation to video/audio data.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an information recording apparatus, an information recording method, and a recording medium. The present invention relates to an information recording apparatus, an information recording method, and a recording medium that enable the above.

[0002]

2. Description of the Related Art In a conventional VTR type video camera, a RET (return) key is provided in a camera portion so that a marker for a portion which is a "trigger" for editing can be input during shooting. ing. Therefore,
By inputting the marker, when editing is performed later, the portion where the marker is input can be easily called, and the editing operation can be facilitated.

[0003]

However, for example, when taking a picture of the operation status with a video camera or the like during the operation, it is difficult for the doctor to operate the camera and input the marker. Even if you ask your assistant,
There is a problem that it is difficult to input a marker as expected by a doctor.

[0004] Therefore, it is desirable to automatically "edit" or to take important screen capture timing during surgery by voice input if possible.

[0005] In this case, unless the key word is a pre-registered word, it becomes impossible to perform an automatic keyword search (collating with a pre-registered word at the time of voice input). Since it is necessary to concentrate on the operation during the operation, it is necessary to automatically input the keyword as metadata (additional data) to the recorder in addition to the video / audio by voice input.

The present invention has been made in view of such a situation, and a keyword is input by voice input during shooting with a video camera, and a keyword is added to video / audio so that a keyword search can be performed later. It is intended to be able to be additionally recorded.

[0007]

An information recording apparatus according to the first aspect of the present invention recognizes video input means for inputting video, audio input means for inputting audio, and audio input by the audio input means. Voice recognition means for converting to a word corresponding to voice, generation means for generating a code corresponding to the word recognized and converted by the voice recognition means, video input by the video input means, and video generated by the generation means Recording means for correlating a code corresponding to the selected word with a predetermined recording medium. The recording means may record the sound input by the sound input means on a recording medium together with the video and the code. Further, it is possible to further provide a storage unit that stores the word and information about the word in association with each other, and the voice recognition unit can convert the voice into the word based on the word stored in the storage unit. A storage unit that associates and stores the word and information about the word; a search unit that searches the storage unit for the information of the word recognized by the voice recognition unit; and a display that displays the information of the word searched by the search unit. Means may be further provided. A storage unit that stores the words and information related to the words in association with each other; a search unit that searches the storage unit for the words recognized by the voice recognition unit; and a display unit that displays information on the words searched by the search units. When a plurality of information is retrieved by the retrieval means, a plurality of information is sequentially displayed on the display means, and a selection means for selecting one information from the plurality of information is further provided. The code of the word corresponding to the selected information can be recorded on a recording medium. Further, a search formula generating means for generating a search formula from the code and the logical code may be further provided, and the recording means may associate the code and the search formula with a video and record the code and the video on a recording medium. The information recording method according to claim 7, wherein a video input step of inputting a video, a voice input step of inputting a voice, and a voice that recognizes the voice input in the voice input step and converts the voice into a word corresponding to the voice. A recognition step, a generation step of generating a code corresponding to the word recognized and converted in the voice recognition step, and associating a video input in the video input step with a code corresponding to the word generated in the generation step Recording on a predetermined recording medium. A recording medium according to an eighth aspect is characterized by recording a program capable of executing the information recording method according to the seventh aspect. In the information recording device, the information recording method, and the recording medium according to the present invention, a video is input, a voice is input, the input voice is recognized, the voice is converted into a word corresponding to the voice, the voice is recognized, and the conversion is performed. A code corresponding to the selected word is generated, and the image is associated with the code and recorded on a predetermined recording medium.

[0008]

FIG. 1 is a diagram showing an example of the configuration of an embodiment of an optical disk camera device to which the information recording device of the present invention is applied. FIG. 1 shows a state in which it is used in a medical video capture and is mounted at the center of the ceiling lighting. The operation of the optical disk camera device 1 (hereinafter abbreviated as a camera as appropriate) is performed by a doctor's assistant staff or the like.

The optical disk camera device 1 includes a lens unit 11 that collects light from a subject and forms an image of the subject,
The microphone 13 includes a microphone 13 that collects surrounding sounds and keywords or the like pronounced by a doctor, an image formed by the lens unit 11, and a camera unit 12 that records sound collected by the microphone 13.

The optical disk camera device 1 has a driver 14 for rotating the optical disk camera device 1 around the Z axis and a driver 15 for rotating the optical disk camera device 1 around the Y axis. Thus, the optical disk camera device 1 can be directed in a predetermined direction.

The optical disk camera apparatus 1 recognizes voice input when capturing an image of an operation or the like, and extracts a corresponding word by referring to a word dictionary 23 in which words and word information are registered in advance. Then, using that as a keyword, the video recording /
The data is inserted as additional data (metadata) into the system data (FIG. 3) inside the frame or field of the audio recording. Thus, after shooting, a frame in which the target video / audio is recorded can be extracted by keyword search.

Also, keywords and logical codes (AND / O
R), a search formula may be generated and recorded in the system data field. For example, keywords such as heart / lung / bypass / surgery are AND / O
By generating a search formula by linking with R and recording it in the system data field, a specific video image can be extracted by specifying the search formula.

FIG. 2 is a block diagram showing an example of an electrical configuration inside the optical disk camera device 1 shown in FIG.

As shown in FIG. 1, an optical disc camera device 1 receives a voice, converts the voice input into predetermined voice data, and converts the voice data converted by the voice input unit 21 into a voice. Recognition voice recognition unit 22
A dictionary search unit 24 for performing a dictionary search on words of the voice recognition result recognized by the voice recognition unit 22; and a search result display unit 25 for displaying information of the words searched by the dictionary search unit 24 to the user. And These voice input unit 21, voice recognition unit 22, dictionary search unit 2
4. The search result display unit 25 constitutes a dictionary search device.

In the optical disc camera apparatus 1, the voice recognition unit 22 uses the voice recognition template for each word used for voice recognition of the voice data supplied from the voice input unit 21 and the dictionary search unit 24 searches for each word. It has a word dictionary 23 in which word information is registered in the same word entry. Here, the voice input unit 21 and the voice input unit 26 constitute the microphone 13 in FIG.

For voice recognition, it is faster to register a language to be used in advance and determine which language is closest to the registered language. Therefore, when a plurality of (for example, two or three) selections occur for the same voice, it is necessary to determine one of them. This determination is input by voice from the voice input unit 26.

The voice input unit 26 responds to the selection of a plurality of candidates of the recognition result displayed as a result of the voice recognition by the search result display unit 25 with YES or NO, or OK.
Alternatively, the NG determination is input by voice. At this time, the voice input from the voice input unit 21 is ignored. Therefore, the voice input unit 26 outputs YES / N
It is assumed that a simple speech recognition circuit for recognizing O, OK / NG is provided.

For example, when the user pronounces "ishi" and inputs the sound from the voice input unit 21, the voice recognition unit 22
From 3, the word "is" is recognized. Next, the dictionary search unit 24 searches the word dictionary 2 for the word “Ishi”.
From 3, search for "doctor", "will", "stone", "hanging", and the like. Then, the search results are sequentially displayed in the search result display unit 2.
5 is displayed. On the other hand, by inputting YES or NO, or OK or NG by voice from the voice input unit 26, a desired word can be selected and determined from a plurality of search results.

FIG. 3 shows an example of a frame format of the optical disk 28 on which video / audio data is recorded. Each frame includes a synchronizing unit for synchronizing, a sync block ID storing an ID indicating a frame order, video data, and audio data. The system data is 640 bytes (B) (= 320 * 2) in the video data and 40 bytes (= 20 * 2) in the audio data, for a total of 68 bytes.
0 bytes (B) / field, ie, 1360 bytes /
Enter the frame. The outer code correction and the inner code correction are error correction codes, respectively.

In the system data of the video data, a keyword input by voice is recorded, and in the system data of the audio data, a keyword, a search formula, progress of “YES / NO determination” described later, Data that could not fit in the data is recorded.

Therefore, it is possible to automatically record 680 characters in the case of 2-byte characters (full-width) and 1360 characters in the case of 1-byte characters (half-width) as metadata in the system data per frame. At this time, if a keyword is extracted from the recognized speech and only the keyword is recorded, it is possible to compress and record the number of used memories as a result.

Next, the operation will be described. During the operation of the doctor, the assistant operates the optical disc camera device 1 to photograph the operation. During photographing, a doctor and an assistant or the like input sound to be recorded on the optical disc camera device 1 by the sound input unit 21.
Enter more.

The voice input to the voice input unit 21 is supplied to a voice recognition unit 22 after being converted into digital voice data. The speech recognition unit 22 uses the word dictionary 23 to
Speech recognition is performed on the speech data supplied from the speech input unit 21, and a word as a recognition result is input to the dictionary search unit 24.
To supply.

The dictionary search unit 24 converts the word as a recognition result supplied from the speech recognition unit 22 into a word dictionary 23.
Is used to perform a dictionary search. Then, word information as a search result, for example, in the case of Japanese, information such as kanji,
In addition, when a plurality of words are searched, information such as homonyms is used, and when the speech recognition is ambiguous, a plurality of search results are set as candidates for the speech recognition results, and the search result display unit 25 is used.
To supply.

The search result display unit 25 displays the word information supplied from the dictionary search unit 24. At this time, if there are a plurality of words that have been voice-recognized, the information of the words is displayed one by one in order, and the system waits for the right or wrong indication.

For example, “YES / NO” or [OK /
Wait for the word "NG" to be input from the voice input unit 26. The voice input unit 26 has a simple voice recognition function, and when “YES” or “OK” is input, the information of the currently displayed word is correct in the search result display unit 25. Is supplied. Upon receiving this signal, the search result display unit 25 determines that the currently displayed word is a correct recognition result, and displays that word in the camera unit 1.
2 to the optical disk recorder recording unit 27.

On the other hand, when "NO" or "NG" is input, the voice input unit 26 supplies a signal to the search result display unit 25 indicating that the information of the currently displayed word is incorrect. I do. Upon receiving this signal, the search result display unit 25 determines that the currently displayed word is not a correct recognition result, and displays the next word. Then, it waits again for the word “YES / NO” indicating the right or wrong to be input from the voice input unit 26 by voice.

From the voice input unit 21, for example, when / where / who / what kind of operation / operation was performed is input.
To be more specific, a voice prompt reads, Kobayashi Kotoshi is the best friend, Fuji's Kyoju is going to meet you
XXXX, man, twenty-six, ", the speech recognition unit 21 extracts a word corresponding to the input speech from the word dictionary 23 and supplies the word to the dictionary search unit 24. The dictionary search unit 24 searches the word dictionary 23 for detailed information on the words recognized by the speech recognition unit 22.

For example, for "Keiodaigaku", "Keio University" registered in the word dictionary 23 in advance is searched. When there are a plurality of search results, as described above, the plurality of search results are sequentially displayed, and one of them is selected by voice. Similarly, other words are searched from the word dictionary 23 and corresponding information is extracted. In this way, the keyword corresponding to the input voice is determined.

In the case of the above example, the following keywords are determined as a result of the speech recognition. That is, “0821198/1
000-1200 / Keio University School of Medicine / Shinanomachi / Lecturer Kobayashi / Professor Fujino / Skull molding / Plastic insertion / Patient name:
***** / male / 25 years old. "

Here, each keyword is represented by an alphabet (A, B, C, D, E, F, G, H, I, J, K) or a number (1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 1
Instead of 1), for example, a search formula such as ((A + B) * E * I) may be created and supplied to the optical disk recorder recording unit 27 together with the keyword. In the above retrieval formula, “+” represents a logical sum, and “*” represents a logical product.

The optical disk recorder recording unit 27 writes the keyword supplied from the search result display unit 25, that is, a code corresponding to one or a plurality of keywords, in the system data shown in FIG. Also, the search result display section 2
When the search formula is supplied from 5, the search formula is also written after the keyword.

At this time, as described above, the search result display unit 25 displays one or a plurality of voice-input keywords (words) for each video / frame (still image) so that a keyword search can be performed later. ) And, if necessary, a search formula to the optical disk recorder recording unit 27. For example, as in the above example, “0821198 / 1000-1200 / Keio University School of Medicine / Shinanomachi / Lecturer Kobayashi / Professor Fujino / Skull Molding / Plastic Insertion / Patient Name: **** / Male / 25
Age ”and a search expression ((A + B) * E *
I) is supplied to the optical disk recorder recording unit 27.

The optical disk recorder recording section 27 records the keyword and the search formula in the system data of video data and the system data of audio data of the frame shown in FIG.

In this way, for each frame (still image) constituting a video, the keyword input by voice can be recorded, and if necessary, can be recorded while generating a search formula.

After photographing, for example, “0821198/10
00-1200 / Lecturer Kobayashi / Patient name: **** / ", the optical disk 28 of the camera unit 12 is searched for a frame in which the same keyword as the inputted keyword is recorded in the system data, and Video or still image to be extracted is extracted. When there are a plurality of applicable videos or still images, multiple images are displayed on a small screen obtained by dividing the screen.
Here, by instructing a predetermined video or still image,
In the case of a video, the designated video is reproduced, and in the case of a still image, the video is displayed on the entire screen.

As described above, at the time of photographing, a word input by voice is recognized, a dictionary search is automatically performed on the word, and the search result is recorded in association with the currently recorded video and audio. Then, a search based on the keyword can be performed later, so that the editing work becomes easy and the editing can be performed at high speed. In addition, archiving can be easily performed via a keyword search.

Further, since the keyword is input by voice, it can be used even in a situation where hands cannot be released, such as a doctor during surgery.

In the above embodiment, the case where the present invention is applied to a medical camera has been described. However, the present invention can be applied to automatic editing, automatic archiving, high-speed archiving, and the like.

In the above embodiment, the case where the present invention is applied to an optical disk camera device has been described. However, the present invention can be applied to other image pickup devices such as a digital video camera.

Further, the frame format in the above embodiment is an example, and the present invention is not limited to this.

[0042]

As described above, according to the information recording apparatus, the information recording method, and the recording medium according to the present invention, video and audio are input, the input audio is recognized, and the audio is converted into words. Since a code corresponding to the word that has been voice-recognized and converted is generated, and the video is associated with the code and recorded on a predetermined recording medium, the video and the audio are recorded by a video camera or the like. In addition, a keyword can be input by voice input, and a keyword can be added to video and audio, and the video and audio can be recorded, and a desired video and audio can be searched for later using the keyword.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration example of an embodiment of an optical disk camera device to which an information recording device of the present invention is applied.

FIG. 2 is a block diagram showing an example of an internal electrical configuration of the embodiment of FIG. 1;

FIG. 3 is a diagram illustrating an example of a frame format.

[Explanation of symbols]

 Reference Signs List 1 optical disk camera device 11 lens unit 12 camera unit 13 microphone 14, 15 driver 21, 26 voice input unit 22 voice recognition unit 23 word dictionary 24 dictionary search unit 25 search result display unit 27 optical disk recorder recording unit 28 optical disk

Claims (8)

[Claims] 1. A video input unit for inputting a video, a voice input unit for inputting a voice, a voice recognition unit for recognizing the voice input by the voice input unit and converting the voice into a word corresponding to the voice. Generating means for generating a code corresponding to the word recognized and converted by the voice recognizing means; the video input by the video input means; and the video corresponding to the word generated by the generating means. An information recording apparatus, comprising: recording means for recording a code on a predetermined recording medium in association with a code. 2. The information recording apparatus according to claim 1, wherein the recording unit records the audio input by the audio input unit together with the video and the code on the recording medium. 3. A storage unit for storing the word and information related to the word in association with each other, wherein the voice recognition unit converts the voice into the word based on the word stored in the storage unit. The information recording device according to claim 1, wherein 4. A storage unit for storing the word and information about the word in association with each other, a search unit for searching the word information recognized by the voice recognition unit from the storage unit, and a search by the search unit The information recording apparatus according to claim 1, further comprising: a display unit configured to display information of the word that has been set. 5. A storage unit that stores the word and information related to the word in association with each other; a search unit that searches the storage unit for the word recognized by the voice recognition unit; Display means for displaying the information of the word; and selecting means for sequentially displaying a plurality of pieces of information on the display means and selecting one piece of information from the plurality of pieces of information when a plurality of pieces of information are searched by the search means. The information recording apparatus according to claim 1, further comprising: a recording unit that records a code of a word corresponding to the information selected by the selection unit on the recording medium. 6. A search expression generation unit for generating a search expression from the code and the logical code, wherein the recording unit associates the code and the search expression with the video and records the image on the recording medium. The information recording device according to claim 1, wherein 7. A video input step of inputting a video, a voice input step of inputting a voice, a voice recognition step of recognizing the voice input in the voice input step and converting the voice into a word corresponding to the voice. A generation step of generating a code corresponding to the word recognized and converted in the voice recognition step; and the video input in the video input step;
A recording step of correlating the code corresponding to the word generated in the generation step with the code and recording the code on a predetermined recording medium. 8. A recording medium on which a program capable of executing the information recording method according to claim 7 is recorded.