KR20110015276A - Glasses-type announcement visualization apparatus using photochromic lens and method thereof - Google Patents

Abstract

PURPOSE: A method and a device for visualizing a guide broadcasting using a discolored lens are provided to increase a voice recognition rate and a portability of the guide broadcasting visualizing device. CONSTITUTION: A voice preprocessing unit(152) performs preprocessing about an inputted voice. Through a preset voice pattern, a voice recognition unit(156) recognizes pre-processed voice. A character post-processing unit(158) adjusts output layout of a character corresponding to the recognized voice. According to neighboring brightness, a discolored transparent lens adjusts background color. A character display unit(160) outputs a post-processed character on inner surface of the discolored transparent lens.

Description

Glasses-type Announcement Visualization Apparatus Using Photochromic Lens and Method

The present invention relates to a spectacled announcement broadcasting device using a color changing lens. In detail, the present invention provides an eyeglass type guide visualization apparatus for improving the voice recognition rate for guide broadcasting to a specialized recognition target, and increasing the readability of characters displayed as a result of voice recognition by changing the background color of the color changing lens according to the surrounding environment. It is about a method.

Humans communicate or communicate with each other using a variety of means, the most widely used of which is language. Languages are largely divided into phonetic and textual languages, each of which has its own characteristics. Voice language that relies on hearing can transmit information simply by making a voice, so it can communicate easily and quickly in a face-to-face state. On the other hand, since the text language that depends on the vision has to undergo a literation process, the instantaneous communication is inferior, but it is relatively conservative.

The general public who is fluent in the language can use both the spoken and written languages without difficulty, but some people prefer either one of them or only communicate by one. For example, hearing impaired people cannot communicate by voice language, and visually impaired people cannot communicate by text language. People who have a slow understanding of a specific language, such as foreigners, may prefer a text language that can be understood slowly over a voice language that requires real-time understanding.

On the other hand, the voice language and the text language can convey the meaning more strongly when both are used at the same time than when only one of them is used. Even ordinary people without difficulty in understanding both speech and text language can understand the message faster and more reliably if both are transmitted at the same time. This is why broadcasting and various multimedia contents for general public transmit subtitles along with video and audio. This is because the general public needs to visualize important voice keywords in the flood of information and receive meaning.

If the message is transmitted only in the voice language and is displayed in the form of a text message, the communication efficiency of the hearing impaired and the general public can be improved. This has been mainly studied in the field of hearing aids or voice visualization technology. Prior art in this field includes assistive devices for the hearing impaired, combining voice recognition and display technologies. Some of these devices have been proposed in the form of a wearable device that displays voice recognition results in text on a transparent lens. However, these devices have not been widely used due to problems such as voice recognition rate, portability, and cost.

Meanwhile, some speech recognition technology is currently commercialized, but it does not yet provide a perfect level of recognition for all speech. Also, in general, the wider the range of speech recognition objects, the greater the volume of data required, which entails an increase in the weight and volume of the device. There is a trade-off relationship between the range of voice recognition targets, voice recognition rate, and portability, so that the wider the voice recognition target, the lower the voice recognition rate and portability in general. It is advantageous to provide a high level of speech recognition and portability.

In various places, such as subways, trains, airplanes, government offices, safaris, galleries, museums, and destinations, announcements are mostly delivered in a spoken language for an unspecified number of ordinary people. Because of this, deaf people have been part of the blind spot of this type of announcement, and the public has not been offered the opportunity to communicate more efficiently using both voice and text. Voice announcements, unlike everyday conversations, make it impossible for the deaf to grasp their content using sign language, reading, and writing, and they contain more important information than ordinary conversations. There is a characteristic that it is doing. Nevertheless, in developing a speech visualization system, the prior arts do not provide a subsidiary device that substantially satisfies both portability and speech recognition rate because it does not specialize the range of speech recognition objects.

In addition, the conventional voice visualization technology has a problem in that the text readability of the user varies depending on the surrounding environment in that the text is displayed in the same way in both bright and dark places. Since the voice language is difficult to predict the input time and has a characteristic of being continuously input, the screen of the portable voice display device is continuously changed. In spite of these characteristics of the voice language, if the readability of the character, which is the result of recognition, varies depending on the brightness of the surroundings, the performance of the device cannot be considered stable from the point of view of the person receiving the input voice language in the form of a text language.

SUMMARY OF THE INVENTION In order to solve the above problems, an object of the present invention is to provide an eyeglass type guide broadcasting visualization apparatus and method which improves portability and voice recognition rate by applying voice recognition and visualization technology specifically for guide broadcasting.

In addition, an object of the present invention is to provide an eyeglass guide visual visualization apparatus and method using a color changing lens that improves character readability by adjusting a background color of a lens according to a change in ambient brightness.

In order to achieve the above object, an aspect of the present invention provides a voice input unit for receiving a voice from an external device, a voice preprocessor for performing noise reduction and signal compensation on the input voice, and information on a preset voice pattern. A voice pattern storing unit for storing the voice pattern, a voice recognition unit for recognizing the preprocessed voice using the voice pattern, a text post-processing unit for adjusting the output layout of the text corresponding to the recognized voice, and brightness according to the surrounding brightness It provides a spectacle-type guide broadcasting visualization apparatus using a color change lens including a color change transparent lens, the background color is adjusted, and a text display unit for outputting the post-processed characters on the inner surface of the color change transparent lens.

In addition, the voice pattern is a voice pattern extracted from the actual sound source (guide) of the announcement, and provides a visual guide broadcasting visualization device using a color changing lens that is used speaker-dependent speech recognition techniques for speech recognition.

The discolored transparent lens may include a photochromic or electrochromic material, and the character display unit may provide a spectacles-type broadcast visualization apparatus using a discolored lens formed by including a transparent display panel.

In addition, the transparent display panel is a transparent photonic light emitting display panel including an upper plate and a lower plate formed of a transparent dielectric, a partition formed of glass, a transparent nano phosphor, and an electrode formed of indium tin oxide (ITO). Provided is an eyeglass visualization system using a color changing lens.

The character post-processing unit outputs the character in a light color when the discolored transparent lens is colored in a dark color, and outputs the character in a light color when the discolored transparent lens is colored in a dark color. Provides an eyeglass guide broadcasting visualization device.

In addition, the voice input unit may be disposed on the left and right sides of the guidance broadcasting visualization apparatus, and the text post-processing unit may be disposed inside the color-transparent transparent lens corresponding to a direction in which the input voice is most strongly detected among the voice input units. Provided is a spectacled announcement broadcasting device using a color changing lens that outputs the character on the surface.

In order to achieve the above object, another aspect of the present invention provides a voice input step of receiving a voice from outside, a voice preprocessing step of performing noise reduction and signal compensation on the input voice, and information on a preset voice pattern. A voice recognition step of recognizing the preprocessed voice, a text post-processing step of adjusting an output layout of a character corresponding to the recognized voice, and a background color of the color-transparent transparent lens to which the character is output according to ambient brightness And a character display step of outputting the post-processed character to the inner surface of the color faded transparent lens.

In addition, the voice pattern is a voice pattern extracted from the actual sound source (sound source) of the guide broadcast, the voice recognition step provides a guide broadcast visualization method using a speaker-dependent voice recognition technique.

In addition, the text post-processing step, the broadcast to output the text in a light color when the color fading transparent lens is colored dark, and outputs the text in a light color when the color fading transparent lens is colored dark Provide visualization methods.

According to an exemplary embodiment of the present invention, the voice recognition target is specialized in a guide broadcast form, so that the portability and voice recognition rate of the guide broadcast visualization device may be improved.

In addition, there is an effect that can provide a stable character readability by adjusting the background color of the character display unit in accordance with the external brightness change.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. In the following description of the present invention, when it is determined that a detailed description of a known function or configuration may unnecessarily obscure the subject matter of the present invention, the detailed description thereof will be omitted. The following terms are defined in consideration of the functions of the present invention, and may be changed according to the intentions or customs of the user, the operator, and the like. Therefore, the definition should be based on the contents throughout this specification.

1 is a block diagram of an embodiment of a guide broadcasting visualization apparatus according to the present invention. The present embodiment is implemented in the form of glasses, it is composed of a color change transparent lens 110, glasses leg 120, the voice input unit 130, the power unit 140, guide broadcast visualization inside the glasses leg 120 The unit 150 is included.

In this embodiment, the discolored transparent lens 110, the glasses leg 120 and the voice input unit 130 is formed of a pair (110a, 110b) (120a, 120b) (130a, 130b), which are merged into one or three It may be formed separately. In addition, the power supply unit 140 and the guide broadcast visualization unit 150 may also be formed outside the glasses leg 120. External materials such as the color change transparent lens 110, the glasses leg 120, the voice input unit 130, and the like may be used as materials used in ordinary glasses or lighter than the power unit 140 so that there is no inconvenience in wearing them. The announcement visualization unit 150 may be distributed and mounted on different glasses legs 120 to balance the weight of the entire structure of the glasses. The power supply unit 140 may be a conventional small battery such as a mercury cell, or a dedicated battery designed to be built in the glasses leg 120.

The voice input unit 130 serves to receive voice from the outside and converts the detected voice wave into an electrical signal. The voice input unit 130 may be implemented as a small microphone or a voice detection sensor mounted on the glasses body, or may be separated from the glasses body and may be a separate module that transmits a voice signal through wired / wireless communication with the announcement broadcast visualization unit 150. Can be implemented.

The announcement broadcast visualization unit 150 may be implemented in hardware such as a microprocessor, and includes a voice preprocessor 152, a voice pattern storage unit 154, a voice recognition unit 156, and a text postprocessor 158. ) May be included. The configuration of the guide broadcasting visualization unit 150 will be described in detail later.

The color change transparent lens 110 is a transparent lens formed of a transparent material whose background color is adjusted according to the brightness of the surroundings. In some embodiments, the color-transparent transparent lens 110 may include a photochromic or electrochromic material.

A representative example of the color-transparent transparent lens 110 formed by including a photosensitive material is a dimming lens, also called a photochromic lens or a Corning lens. Dimming lenses use chemical vapor deposition (CVD) or sputtering techniques on glass or plastic surfaces, and are used silver halides such as silver chloride (AgCl), silver bromide (AgBr), and silver iodide (AgI). It is manufactured by forming an oxide thin film of the (Ag) series. The dimming lens has a light color that is almost colorless in a dark place such as an indoor environment, but the color of the dimming lens is dark in a bright place such as an outdoor area. The color of the dimming lens changes depending on the irradiation amount of light because the silver halides dispersed as fine crystals in the lens material react with each other depending on the amount of ultraviolet rays. When a silver halide-based compound receives ultraviolet rays, a coloring reaction occurs in which silver (Ag) and a halogen element are released into the respective fine particles, and when the ultraviolet ray is blocked and receives infrared or visible light, the free silver (Ag) and halogen element are ionized. This results in a decolorization reaction that recombines with silver halide (Ag).

In another embodiment, there is a lens using an electrochromic material, which transmits an electric signal to glass or plastic when ultraviolet rays are detected, so that the color of the substrate to which the electric signal is transmitted may increase cyanity even in external light. It's used to turn brown. In this case, an ultraviolet sensor may be attached to the glasses. In the present embodiment, the screen portion may be made by using an indium tin oxide (ITO) -based glass plate, and may be used to discolor using an electrochromic device material. For example, a method of using the tungsten oxide and giving a predetermined voltage to the substrate as shown in Formula 1 below, may be used to deposit.

W + H₂O₂ → WOx + H₂O + H₂O₂

The character display unit 160 is a transparent display device for outputting the post-processed character on the inner surface of the color change transparent lens. Referring to FIG. 1, in the present exemplary embodiment, the character display unit 160 is formed only on the inner bottom surface of the one color change transparent lens 110b. However, according to the embodiment, the character display unit 160 may include both color change transparent lenses ( All may be formed in the 110, and the location and area to be formed may be variously set. According to an exemplary embodiment, the character display unit 160 may be implemented as a transparent display panel. A detailed configuration of the character display unit 160 in the form of a transparent display panel will be described in detail later.

2 is a block diagram illustrating a schematic configuration of an embodiment of a guide broadcast visualization unit.

The speech preprocessor 152 may perform noise reduction and signal compensation on the speech signal received from the speech input unit 130, and perform feature vector extraction required for recognition. The voice preprocessor 152 may include hardware such as an analog-to-digital converter and a digital signal processor. The preprocessing performed by the speech preprocessor 152 is to increase the speech recognition rate, and includes pre-emphasis, windowing, fast Fourier transform (FFT), and energy analysis. It may include the feature vector to the end point detection of the speech performed through. In addition, by applying various filtering and audio signal processing algorithms to perform various types of noise cancellation and signal compensation, the speech recognizer 156 concentrates purely on phonetic characteristics. It can support to perform voice recognition. The speech preprocessing unit 154 extracts only the phonetic characteristics from the speech signal received through noise reduction and signal compensation, LPC (Linear Predictive Coding), MFCC (Mel Frequency Cepstral Coefficients), high frequency emphasis extraction method, window function. Extraction methods and the like can be used.

The voice pattern storage unit 154 stores information about a preset voice pattern. As the voice pattern stored according to the embodiment, the actual voice pattern extracted from a sound source of a guide broadcast broadcast in public places, safaris, art galleries, museums, tourist attractions, or the like and recorded voices may be used. Herein, the announcement may include all types of sounds for delivering information or warning messages in various forms such as alarm sounds as well as human voices. The recognition rate of speaker-dependent speech recognition technology that recognizes the voice of a particular speaker is generally higher than that of the speaker-independent speech recognition technology that recognizes the input voice regardless of who the dialogue is. . Considering this, when the actual sound source of the announcement used in public places such as subways, trains, airplanes, and public offices is stored as a voice pattern, the voice recognition rate of the entire system can be increased to the recognition rate of the speaker-dependent voice recognition technology. There is an advantage.

The speech recognizer 156 recognizes the preprocessed speech using the stored speech pattern. As a speech recognition algorithm, various techniques such as template pattern matching based on speaker-dependent speech recognition technology may be used. In this case, the speech recognition unit 156 compares the template speech pattern stored in the speech pattern storage unit 154 with the pattern of the feature vector or the end point extracted from the preprocessed speech, and among the speech patterns stored in the speech pattern storage unit 154. The most accurate speech pattern is selected as the recognition result. As a template pattern matching algorithm, DTW (Dynamic Time Warping), HMM (Hidden Markov chain), ANN (Artificial Nural Network) may be used. According to an exemplary embodiment, the speech recognizer 156 may include a first similarity between the pattern of the preprocessed voice and the voice pattern having the highest similarity among the stored voice patterns, and the second highest similarity between the preprocessed voice pattern and the stored voice pattern. The second similarity between the voice patterns may be implemented to be strictly processed as if the voice is recognized only when there is a difference of a predetermined level or more.

The text post-processing unit 158 adjusts the output layout of the text corresponding to the recognized voice. The output layout of the characters may include the number and position of characters displayed on a line, the font, size and color of the displayed characters. By adjusting the settings regarding the output layout of the text, the user can more effectively see the text in which the voice recognition result is converted. According to an exemplary embodiment, the character post-processing unit 158 may be configured to interlock with a color change of the color changing transparent lens 110 that reacts to a change in ambient brightness. In this case, for example, when the discolored transparent lens 110 is colored in a dark color, the color of the text is brightened, and when the transparent color is bleached transparent, the color of the text is darkened, thereby ensuring stable character readability without being sensitive to changes in the surrounding environment. It becomes possible. In another embodiment, the character display unit 160 is provided at both sides of the color-transparent transparent lens 110, and the intensity of the signal detected by the voice input unit 130 is different from the voice detected by the voice input unit 130. If the signal is greater than the strength of the signal, the color corresponding to the strong color of the transparent lens 110 in the direction in which the signal is strongly detected may display a character corresponding to the result of the recognition of the voice. In this case, there is an advantage in that not only the result of speech recognition but also information on the direction in which voice is heard can be delivered.

3 illustrates a screen displayed by the announcement broadcasting visualization apparatus of the present invention in various environments.

3A is a screen when used in a platform of a train station during the day. Since the ambient brightness is a bright day, the discolored transparent lens 110 is colored, and since the voice is more strongly detected at the right side of the user, the character display unit 160 of the right discolored transparent lens 110a displays a letter corresponding to the recognized voice. A train comes in ”is displayed.

3B is a screen when used in a parking lot at night. Since the ambient brightness is dark at night, the discolored transparent lens 110 is discolored, and since the voice is more strongly detected at the left side of the user, the character display unit 160 of the left discolored transparent lens 110b has a letter corresponding to the recognized voice. Please pull parking ticket ”is displayed.

3C is a screen when used in a place where an alarm sounds during the day. Since the ambient brightness is a bright day, the discolored transparent lens 110 is colored, and since the voice is more strongly detected at the left side of the user, the character display unit 160 of the left discolored transparent lens 110b has a character corresponding to the recognized alarm sound. “[Alarm tone] beep- beep-” is displayed. According to an embodiment, an alarm sound that is not a human voice may be separately displayed.

4 shows the configuration of an embodiment of a transparent display panel used in the character display unit. As described above, the character display unit 160 may be implemented as a transparent display panel. Specifically, the transparent display panel includes a front plate 410 and a lower plate formed of transparent dielectric layers 414 and 422 having a high ratio of silicon dioxide (SiO 2), as shown in FIG. 4. (rear plate) 420, barrier ribs (not shown) formed of thin glass, transparent nanophosphors 424 having a size in nanometers (nm), and indium tin oxide It may be a transparent photoluminescence display panel including a bus electrode 414 formed of Indium Tin Oxide (ITO).

5 is a flowchart illustrating an embodiment of a guide broadcasting visualization method of the present invention.

The guide broadcasting visualization method according to an embodiment of the present invention includes a voice input step (S100) for receiving a voice from the outside, a voice preprocessing step (S200) for performing noise reduction and signal compensation on the input voice, and a preset Speech recognition step (S300) for recognizing the pre-processed voice using information on the speech pattern, post-text processing step (S400) for adjusting the output layout of the text corresponding to the recognized voice, and the brightness of the surroundings And a brightness adjustment step (S500) of adjusting a background color of the color fading transparent lens to which the text is output, and a character display step (S600) of outputting the post-processed text to the inner surface of the color fading transparent lens with brightness adjustment. do.

According to an embodiment, a voice pattern extracted from a sound source of a guide broadcast in a public place is used as a voice pattern, and a speaker dependent voice recognition technique may be used in the voice recognition step S300. According to another embodiment, the text post-processing step (S400) outputs the text in a light color when the color fading transparent lens is colored in a dark color, and brightens the text when the color fading transparent lens is colored in a dark color. It may include a process of outputting in color.

In the above-described guide broadcasting visualization apparatus and method of the present invention, the voice recognition target is specialized for guide broadcasting to increase the voice recognition rate and increase the readability of the recognition result using the color change lens, limited to the use for the hearing impaired It doesn't work. The guidance broadcasting visualization apparatus and method of the embodiment of the present invention is widely recognized as a technology for realizing efficient communication by voice and text to both the hearing impaired and the general public by recognizing the voice language and displaying it on the color lens on the color changing lens. It should be interpreted.

Modules, functional blocks or means of the present embodiment may be implemented in a variety of known elements, such as electronic circuits, integrated circuits, ASICs (Application Specific Integrated Circuit), each may be implemented separately, or two or more may be integrated into one Can be.

Although the embodiments have been described for the understanding of the present invention as described above, it will be understood by those skilled in the art, the present invention is not limited to the specific embodiments described herein, but variously without departing from the scope of the present invention. May be modified, changed and replaced. For example, the technique of the present invention may be applied to a picture, an image, etc., which may be displayed by a display such as an LCD instead of a character. Therefore, it is intended that the present invention cover all modifications and variations that fall within the true spirit and scope of the present invention.

1 is a block diagram of an embodiment of a guide broadcasting visualization apparatus according to the present invention.

2 is a block diagram illustrating a schematic configuration of an embodiment of a guide broadcast visualization unit.

3 illustrates a screen displayed by the announcement broadcasting visualization apparatus of the present invention in various environments.

4 shows the configuration of an embodiment of a transparent display panel used in the character display unit.

5 is a flowchart illustrating an embodiment of a guide broadcasting visualization method of the present invention.

Claims (9)

A voice input unit for receiving a voice from the outside; A speech preprocessor for performing noise reduction and signal compensation on the input speech; A voice pattern storage unit for storing information about a preset voice pattern; A speech recognition unit for recognizing the preprocessed speech using the speech pattern; A text post-processing unit for adjusting an output layout of a text corresponding to the recognized voice; Discolored transparent lens whose background color is adjusted according to the surrounding brightness, and Character display unit for outputting the post-processed character on the inner surface of the color fading transparent lens Glasses type announcement visualization device using a color changing lens including a. The method of claim 1, The voice pattern is a voice pattern extracted from an actual sound source of the announcement, Speaker-dependent speech recognition technique is used for the speech recognition Eyeglass visual broadcasting system using a color changing lens. The method of claim 1, The discolored transparent lens is formed by including a photochromic or electrochromic material, The character display unit is formed to include a transparent display panel Eyeglass visual broadcasting system using a color changing lens. The method of claim 3, wherein The transparent display panel is a transparent photonic light emitting display panel including an upper plate and a lower plate formed of a transparent dielectric, a partition formed of glass, a transparent nano phosphor, and an electrode formed of indium tin oxide (ITO). Eyeglass visual broadcasting system using a color changing lens. The method of claim 3, wherein The character post-processing unit outputs the character in a bright color when the color fading transparent lens is colored in a dark color, and outputs the character in a light color when the color fading transparent lens is colored in a dark color. Eyeglass visual broadcasting system using a color changing lens. The method of claim 1, The voice input unit is disposed on the left and right sides of the announcement broadcast apparatus as a plurality, The text post-processing unit outputs the text to an inner surface of the color fading transparent lens corresponding to a direction in which the input voice is most strongly detected among the voice input units. Eyeglass visual broadcasting system using a color changing lens. A voice input step of receiving a voice from the outside, A voice preprocessing step of performing noise reduction and signal compensation on the input voice; A voice recognition step of recognizing the preprocessed voice using information on a preset voice pattern; A text post-processing step of adjusting an output layout of a text corresponding to the recognized voice; A brightness adjustment step of adjusting a background color of the color-transparent transparent lens in which the text is output according to ambient brightness; Character display step of outputting the post-processed character to the inner surface of the color fading transparent lens of the brightness control Announcement visualization method comprising a. The method of claim 7, wherein The voice pattern is a voice pattern extracted from an actual sound source of the announcement, In the speech recognition step, a speaker-dependent speech recognition technique is used. How to visualize announcements. The method of claim 7, wherein The post-processing of the text may include outputting the text in a light color when the color fading transparent lens is colored in a dark color, and outputting the text in light color when the color fading transparent lens is colored in a dark color. How to visualize announcements.

Images