JPH05333891A - Automatic reading device - Google Patents

Abstract

PURPOSE:To obtain the automatic reading device which outputs an effect sound matching the scenery of the contents of a book or projects an image showing the scenery simultaneously with a synthesized speech of a read of the contents of the book. CONSTITUTION:An optical character read part 13 after optically reading characters of the book 2 segments the individual characters to generate character-recognized text data. A vocalizing device part 14 while vocalizing the text data automatically detects an image key word showing the scenery in the text data by an image key word detection part 23 and reproduces and outputs effect sound data corresponding to it by an effect sound data reproduction part 26. In another way, a speech synthesizing process and display control part automatically detect the image key word showing the scenery in the text data by the part 23 and draws an corresponding image as a picture. Therefore, a reader is given a feeling of presence and the concrete image.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an automatic reading apparatus for automatically recognizing printed characters of a book such as a novel and outputting it as synthetic voice.

[0002]

2. Description of the Related Art Conventionally, as such an automatic reading device, there is a device for a blind person having an appearance as shown in FIG. This device includes a drawer mechanism (1) for setting the book (2) in the device, an operation panel (3), a Braille explanation board (4) for explaining the operation procedure to the blind, and a recording device. Tape mounting part (5) of cassette tape and headphones (6)
And are provided outside.

Next, FIG. 13 showing the system configuration of this system.
The operation will be described below. The blind person himself sets the book (2) on the main body of the device by the pull-out mechanism section (1) and the operation panel (3), and prepares for reading (A) such as wearing headphones (6). Then, in the optical character reading unit (7), C
When the CD sensor optically reads the print character of the book (2) and inputs the image data, the character recognition unit cuts out the image data for each character and recognizes each character by collation with a standard pattern. Further, in the sentence analysis unit (8), the dictionary is searched by the VLSI memory to analyze the structure of the sentence and the like to obtain the text data of the sentence. The text data is output from the headphone (7) as synthetic voice with intonation and accent by the voice synthesizer (9) and recorded on the cassette tape (51).

[0004]

Since the above-mentioned automatic reading device can read aurally, a blind person can know the contents described in various ordinary books (2) other than Braille without bothering others. It's extremely convenient. However, since the description content of the book (2) is expressed only by synthesized speech, which is the same as a stick reading, the persuasive power and the realism of a concrete image such as when a human reads and listens to the book (2). It is scarce. Therefore, it is easy to get tired and is not used permanently, and even if it is used for learning for children other than blind people, it cannot be used at all because it is a plain dry reading of sentences as described above. ..

Therefore, according to the present invention, a sound effect that matches the scenery, scene, etc. of the description content is output at the same time as the synthesized sound that interprets the description content of the book, or the description together with the synthetic sound that interprets the description content of the book. It is a technical object to provide an automatic reading device capable of displaying images such as scenery and scenes of contents in a video.

[0006]

The present invention has an automatic reading device configured as follows as a technical means for achieving the above-mentioned problems. That is, an optical character reading unit that creates image data of a text of a book by recognizing image data obtained by optically reading printed characters of a book for each character, and a sounding device that outputs the text data as a synthetic voice. An image for determining whether or not a voice synthesizing unit for converting the text data into a synthetic voice and a preset image keyword indicating a scene of a sentence are present in the text data in the sounding device unit. Keyword detecting means, sound effect data detecting means for reading sound effect data corresponding to the image keyword detected by the image keyword detecting means from the sound effect data memory, and sound effect data for reproducing the read sound effect data. A reproducing means, and the reproduced sound effect and the synthesized voice from the voice synthesizing means are mixed and output. It is configured as characterized by comprising an audio output means that.

Alternatively, a voice synthesizing process / display control unit is provided in place of the above-described sounding device unit, and the voice synthesizing process / display control unit is preset with voice synthesizing means for converting the text data into synthetic voice. Image key word detecting means for determining whether or not an image keyword indicating a scene of a sentence exists in the text data, and image data searching for searching image data corresponding to the image keyword detected by the image keyword detecting means And a signal processing control means for controlling the video data retrieved by the video data retrieval section to be read out from the video data memory and displayed on the image display section, and at the same time to output the synthesized voice by the voice synthesis means from the sounding means. It is characterized by having.

[0008]

In the former automatic reading device, in the optical character reading unit, the text on the paper surface of the book is captured as image data by photoelectric conversion by, for example, an image sensor, and individual characters are cut out from this image data and cut out. The character recognizing unit recognizes which character is the recognized character and stores the recognized character as text data. Then, in the sounding device section, the text data is read and input for each reading unit up to the punctuation mark ".", And the image keyword detecting means detects whether an image key word indicating the scene of the sentence exists in the read text data. If there is an image keyword in the text data, for example, the sentence of the text data is "the sound of rain suddenly becomes intense."
Then, "sound of rain" is an image keyword, and the sound effect data detection means searches the sound effect data corresponding to the image keyword, and the sound effect data corresponding to the image keyword stored in advance in the sound effect data memory. The data is read out, and this sound effect data is reproduced by the sound effect data reproducing means, and at the same time, the text data is converted into the synthetic sound by the sound synthesizing means, and the synthetic sound and the reproduced sound effect are mixed to produce a speaker or the like. Is output from the voice output means. Therefore, a sound effect that matches the scene of the read sentence is output at the same time, so that it is possible to convey a sense of presence and a concrete image to the reader.

On the other hand, in the latter automatic reading device, the optical character reading section is the same as that of the above-mentioned automatic reading device. Then, when the voice synthesis processing / display control unit reads out and inputs the text data for each reading unit, the image keyword detection unit determines whether or not an image keyword indicating the scene of the sentence exists in the read text data. If there is an image keyword in the text data,
For example, if the sentence of the text data is "It was Snow Country.", "Snow Country" is the image keyword, and the video data search means searches the video data corresponding to the image keyword of "Snow Country" to display the scene of the snow country. The image data of the "snow scene" to be expressed is drawn on the image display unit, and the text data is converted into synthetic voice by the voice synthesis unit and output from the speaker. Therefore, since an image that matches the scene of the text being read is displayed at the same time, it is possible to convey the sense of presence and its concrete image to the reader, which is particularly suitable for children's learning.

[0010]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Preferred embodiments of the present invention will be described in detail below with reference to the drawings. FIG. 1 shows an external appearance of an embodiment of the present invention. A book insertion tray (10) for setting a book (2) in an apparatus is provided at the lower center of the front surface, and speakers are provided on both left and right sides of the tray (10). (11) are respectively arranged, and the operation portion (12) is provided on the upper surface. FIG. 2 shows a functional block diagram of this embodiment, which is roughly composed of an optical character reading section (13), a sound producing section (14) and a control circuit (15). Each of these components will be described based on the description of the operation by the flowcharts of FIGS. 4 and 5. FIG. 4 shows the operation of the optical character reading unit (13), which converts the text on the paper of the book (2) into
The character reading control unit (1) is provided as image data by photoelectric conversion by the image sensor of the image input unit (19).
7) (step S1). Character reading controller (1
7) cuts out individual characters from the input image data, and recognizes the cut-out characters by referring to the character recognition unit (18) (step S2). The recognized character is temporarily stored as text data in the text data memory (16) (step S3). Then, until it is judged that the reading operation by the optical character reading unit (13) is completed (step S4), the page feed mechanism unit (20) reads the book () every time the recognition of all characters in the input image data is completed. After automatically turning the page of 2), the process jumps to step S1 and the same routine as described above is repeated.

FIG. 5 shows the operation of the sounding device section (14),
The pronunciation control unit (21) has a text data memory (16).
The stored text data is read and input for each reading unit up to the punctuation mark "." (Step S6). The image keyword detection unit (23) is connected to the pronunciation control unit (2
It is determined whether or not the image keyword indicating the scene of the sentence exists in the text data read in 1) (step S7). This image keyword is set corresponding to each sound effect data stored in advance in the sound effect data memory (25). If the image keyword does not exist, the text data is converted into synthetic speech by the speech synthesis unit (22) and the speaker (1
It is output from 1).

On the other hand, when an image keyword is present in the text data, for example, if the sentence of the text data is "the sound of rain suddenly became severe."
“Sound of rain” is an image keyword, and the sound effect data detection unit (24) searches for sound effect data corresponding to the above-mentioned image keyword (step S8). That is, the sound effect data detection unit (24) detects the sound effect data number "1" corresponding to the image keyword of "rain sound" based on the data search table as shown in FIG. Further, as shown in FIG. 6B, the sound effect data of “raining sound” stored in advance in the sound effect data memory (25) as the sound effect data number “1” is read.

Subsequently, the above-mentioned read-out sound effect data of "rainy sound" is reproduced by the sound effect data reproducing section (26), and at the same time, the text data is synthesized by the voice synthesizing section (22). Is converted to (step S9),
The synthesized voice and the reproduced sound effect are mixed and output from the speaker (11) (step S10). Then, it is determined that the reading of all the text data from the text data memory (16) is completed (step S11).
Until this is done, the process jumps to step S6 and the same routine as described above is repeated to read the remaining text data. Therefore, a sound effect that matches the scene of the read sentence is output at the same time, so that it is possible to convey a sense of presence and a concrete image to the reader.

A control circuit (15) controls the optical character reading section (13) and the sound producing section (14) individually and controls the transfer of text data between them. Therefore, the optical character reading unit (13) is independent of the signal processing status of the sounding device unit (14) and the page break mechanism unit (2).
0) can be driven to read the next page, so that it is possible to prepare for the reading operation of the next page. By performing such processing, it is possible to read continuous pages over many pages.

FIG. 3 shows a block embodying the above-mentioned embodiment, in which the same or equivalent parts as those in FIG. 2 are designated by the same reference numerals and should be supplemented to the above description of FIG. Only matters will be explained. In the optical character reading unit (13), the print character of the book (2) is read by the image sensor (2
A program ROM for cutting out each character from the image data converted into a digital signal by the A / D converter (27) after being read in 8) and recognizing the cut out character
(29), dictionary ROM (30) and work RAM
Character reading control CPU (17) connected to (31)
It is processed and processed under the control of.

On the other hand, in the sound producing unit (14), an image keyword is detected from the text data recognized as characters, and the sound effect data corresponding to the detected image keyword is searched, and then the D / A conversion unit (40). The analog signal is converted into an analog signal, and at the same time, the digital signal processor (39) synthesizes voice based on the text data, and the D / A converter (41) converts the voice into an analog signal. Each analog signal with the sound effect data is mixed by the amplifier (42) and output through the speaker (43). These operations are performed by a program ROM (35), a work RAM (36), a voice synthesis dictionary ROM, and a sound effect data ROM (38).
Is controlled by the tone control CPU (21) connected to
It is processed.

In the above embodiment, the page break mechanism section (20) is used.
Although the book (2) is provided and the pages are automatically turned without taking out the book (2) one by one, the book (2) may be turned manually. Further, as the sound effect data, either a natural sound or a synthesized sound may be used. Furthermore, in the block of FIG. 3, the case where three CPUs (17), (21), and (32) are used is illustrated, but one CPU can be shared and controlled, and the effect sound is generated. The reproduction time length can be set as appropriate according to the content of the read sentence.

FIG. 7 shows the appearance of another embodiment of the present invention in which an image image is used instead of the sound effect of the above embodiment.
In the figure, the same or equivalent parts as those in FIG. 1 are designated by the same reference numerals, and the difference in appearance is that an image image display section (a liquid crystal display device) is provided above the main insertion tray (10). 44) is only provided. A functional block diagram of this is shown in FIG. 8, in which the same or equivalent parts to those in FIG. 2 are designated by the same reference numerals. And
The only difference is the configuration in which a voice synthesis processing / display control unit (45) is provided in place of the sounding device unit (14). The voice synthesis process / display control unit (45) controls the whole voice synthesis process / image display control unit (46), voice synthesis unit (22), and image keyword detection unit (23).
A video data search unit (47) for searching a video data number based on the detected image keyword, and a video data memory (49) in which predetermined video data is stored in advance.
And the image display section (4
4) and.

The optical character reading section (13) is similar to that of FIG. 2 and operates according to the flow chart of FIG. Next, the operation of the voice synthesis processing / display control unit (45) will be described with reference to FIG.
This will be described with reference to the flowchart of FIG. The voice synthesis processing / image display control unit (46) reads out and inputs the text data stored in the text data memory (16) for each reading unit up to a punctuation mark "." (Step S13). Image keyword detector (2
3) is a control unit (46) for voice synthesis processing and image display
It is determined whether or not there is an image keyword indicating the scene of the sentence in the text data read in (step S14). This image keyword is set corresponding to each video data stored in advance in the video data memory (48). If the image keyword does not exist in the text data, the text data is converted into synthetic speech by the speech synthesis unit (22) and output from the speaker (11) (step S17).

On the other hand, when an image keyword is present in the text data, for example, if the sentence of the text data is "It was Snow Country.", "Snow Country" is the image keyword, and the video data retrieval unit (47) The video data corresponding to the above-mentioned image keyword is searched (step S15). That is, the video data search unit (47) detects the video data number "1" corresponding to the image keyword "Snow Country" based on the data search table as shown in FIG. As shown in (b), the image data of "snow scene" representing the scene of the snow country, which is stored in advance in the image data memory (48) in association with the image data number "1", is read.

Subsequently, the voice synthesis processing / image display control unit (46) causes the image image display unit (44) to draw the read video data of the "snow scene" (step S16). The text data is converted into synthetic speech by the speech synthesis unit (22) and the speaker (11)
Is output (step S17). Then, until it is determined that the reading of all the text data from the text data memory (16) is completed (step S18),
The process jumps to step S13 and the same routine as described above is repeated to read the remaining text data.
Therefore, an image that matches the scene of the text being read is drawn at the same time, so that it is possible to convey the sense of presence and its concrete image to the reader, and it is particularly suitable for applications such as learning for children. ..

FIG. 9 shows a block diagram embodying the above-mentioned embodiment, in which the same or equivalent parts as those of FIGS. 3 and 8 are designated by the same reference numerals, and the description of FIG. Only items that should be supplemented will be explained. The optical character reading section (13) and the control circuit (15) are the same as those in FIG. Then, the voice synthesis processing / display control unit (45) detects an image keyword from the text data subjected to character recognition, searches for video data corresponding to the detected image keyword, and then displays this video data in a display driver (50). ) Is displayed on the image display section (44), and the digital signal processor (3
9) The voice is synthesized by the D / A converter (41).
Is converted into an analog signal by the, and the analog signal is output from the speaker (11) through the amplifier (42). These operations are performed by the program ROM (35) and the work RA.
M (36), a voice synthesis dictionary ROM and a video data ROM (49) connected to the pronunciation synthesis and display control CP
It is controlled and processed by U (43).

As the video data, moving image data other than landscape may be used in addition to image data created based on pictures and photographs.

[0024]

As described above, according to the automatic reading apparatus of the present invention, after the characters of a book are optically read, the characters are cut into individual characters and the recognized character data is output as a synthetic voice, and the text data is also output. Since it is configured to automatically recognize the description of the scenery and the scene inside and output the sound effect that represents that scene at the same time, the sound effect that matches the scene of the sentence that is read aloud will be output at the same time to the reader. You can convey the feeling and its concrete image.

In addition, the text data obtained by optically reading the characters in the book are output as synthesized voice, and the scenery and the depiction contents of the scene in the text data are automatically recognized and an image representing the scene is displayed. Since it is configured as a drawing, it is possible to convey the presence and specific image to the reader by displaying an image image that matches the scene of the sentence to be read aloud, and it is an advantage that it can be used especially as a learning device for children. is there.

[Brief description of drawings]

FIG. 1 is a perspective view showing an appearance of an embodiment of the present invention.

FIG. 2 is a functional block diagram of the same.

FIG. 3 is a block diagram of the embodied configuration of the same.

FIG. 4 is a flowchart showing the operation of the optical character reading unit.

FIG. 5 is a flowchart showing the operation of the sound producing device section.

6 (a) and 6 (b) are explanatory views of a data search table for image keyword search and a storage state of sound effect data.

FIG. 7 is a perspective view showing the outer appearance of another embodiment of the present invention.

FIG. 8 is a functional block diagram of the same as above.

FIG. 9 is a block of the embodied configuration of the above.

FIG. 10 is a flowchart showing the operation of the voice synthesis processing / display control unit of the above.

FIG. 11A and FIG. 11B are explanatory diagrams of a data search table for image keyword search and a storage state of video data.

FIG. 12 is a perspective view showing an appearance of a conventional blind automatic reading device.

FIG. 13 is a system configuration diagram of the same.

[Explanation of symbols]

Two 11 Speaker (sound output means) 13 Optical character reading section 14 Sounding device section 22 Speech synthesis section (speech synthesis section) 23 Image keyword detection section (image keyword detection section) 24 Sound effect data detection section (Sound effect data detection section) ) 25 sound effect data memory 26 sound effect data reproducing section (sound effect data reproducing means) 44 image image display section 45 voice synthesis processing / display control section 46 voice synthesis processing / image display control section (signal processing control section) 47 video Data search unit (video data search means) 48 Video data memory

Claims (2)

[Claims] 1. An optical character reading section for creating image data of a text of a book by recognizing image data obtained by optically reading printed characters of a book for each character, and outputting the text data as a synthetic voice. And a voice synthesizer for converting the text data into a synthesized voice, and whether or not an image keyword indicating a preset sentence scene is present in the text data. Image keyword detecting means for determining, sound effect data detecting means for reading sound effect data corresponding to the image keyword detected by the image keyword detecting means from the sound effect data memory, and reproducing the read effect sound data. The sound effect data reproducing means, and the reproduced sound effect and the synthesized voice from the voice synthesizing means are mixed. An automatic reading device, comprising: 2. A voice synthesizing process / display control unit is provided in place of the sounding device unit according to claim 1, and the voice synthesizing process / display control unit converts the text data into synthetic voice. Means, an image key word detecting means for determining whether or not an image keyword indicating a scene of a preset sentence exists in the text data, and video data corresponding to the image keyword detected by the image keyword detecting means. And a control unit for controlling the video data retrieval unit for retrieving the video data retrieved by the video data retrieval unit from the video data memory to be displayed on the image display unit, and for outputting the synthesized voice by the voice synthesis unit from the sounding unit. An automatic reading device comprising a signal processing control means.