CN110047465A - A kind of accounting language identification information input device - Google Patents
A kind of accounting language identification information input device Download PDFInfo
- Publication number
- CN110047465A CN110047465A CN201910355212.4A CN201910355212A CN110047465A CN 110047465 A CN110047465 A CN 110047465A CN 201910355212 A CN201910355212 A CN 201910355212A CN 110047465 A CN110047465 A CN 110047465A
- Authority
- CN
- China
- Prior art keywords
- module
- voice
- accounting
- central processing
- identification information
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/005—Language recognition
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/26—Speech to text systems
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
- G10L21/00—Processing of the speech or voice signal to produce another audible or non-audible signal, e.g. visual or tactile, in order to modify its quality or its intelligibility
- G10L21/02—Speech enhancement, e.g. noise reduction or echo cancellation
- G10L21/0208—Noise filtering
Abstract
The invention belongs to technical field of language recognition, a kind of accounting language identification information input device is disclosed, the accounting language identification information input device includes: voice input module, template data module, monitoring inspection module, central processing module, display module, storage module.The sound that the present invention passes through the sending of voice input module typing accounting, and analog signal is converted for voice by central processing module, digital signal is converted by analog signal, digital signal and waveform database are contrasted, then it is split into single phoneme, finally according to the raw new word language in phoneme queries dictionary library or sentence, it is finally completed the work of words input.
Description
Technical field
The invention belongs to technical field of language recognition more particularly to a kind of accounting language identification information input devices.
Background technique
Currently, needing largely to carry out the statistics of number when accounting is worked, traditional data inputting uses computer
Words input, it is prolonged to work, not only make the body of accounting feel tired, and it is tired out in the state of be easy to produce mistake
Accidentally.Accounting language identification information input device plays a very important role statistics, improves work efficiency,
Reduce the workload of staff.But traditional computer is logged in by hand under prolonged working condition, is easy to produce mistake
Accidentally;Existing relevant language identification information input system simultaneously, since the sound of acquisition is mingled with the noise of many types, drop
The clarity and resolution of low acquisition voice;It is selected in existing relevant language identification information input system according to voice prompting
It selects in suitable template procedure, system will appear the slow phenomenon of reaction speed;Existing relevant language identification information typing system
When system storage Various types of data, Various types of data occupies a large amount of space of storage module, leads to the reduction of system response rate.
In conclusion problem of the existing technology is:
(1) traditional computer is logged in by hand under prolonged working condition, is easy to produce mistake.
(2) existing relevant language identification information input system, since the sound of acquisition is mingled with making an uproar for many types
Sound reduces the clarity and resolution of acquisition voice.
(3) suitable template procedure is selected according to voice prompting in existing relevant language identification information input system
In, system will appear the slow phenomenon of reaction speed.
(4) when existing relevant language identification information input system storage Various types of data, Various types of data occupies storage mould
The a large amount of space of block leads to the reduction of system response rate.
Summary of the invention
In view of the problems of the existing technology, the present invention provides a kind of accounting language identification information input devices.
The invention is realized in this way a kind of accounting language identification information input device includes: voice input module, template
Data module, monitoring check module, central processing module, display module, storage module.
Voice input module;It is connect with central control module, the voice of user is acquired by microphone apparatus, and will
The voice transfer of acquisition is to central control module.
Template data module;It is connect with central processing module, relevant number or text template is selected according to voice prompting,
The data storage of the template data module is in storage module.
Monitoring checks module;It is connect with central processing module;All voices of the accounting in voice input are recorded, are convenient for
The data of later period inquiry storage in the past.
Central processing module;Module, display module, storage mould are checked with voice input module, template data module, monitoring
Block connection, converts analog signal for voice, converts digital signal for analog signal, digital signal and waveform database are formed
Comparison, is then split into single phoneme, finally according to the raw new word language in phoneme queries dictionary library or sentence, is finally completed words input
Work.
Display module;It connect with central processing module, is shown the content after speech recognition by display screen.
Storage module is attached with central processing module, the number such as audio, text, number for storing modules
According to.
Further, sound is carried out framing, uses move function by the voice input module when carrying out the typing of sound
Window is realized that the length of every frame is 25ms, there is overlapping for 15ms between every two frame.
Advantages of the present invention and good effect are as follows:
The present invention carries out the conversion of text by the sound of speech recognition accounting, thus by relevant data transmission to correlation
File in, reduce the workload of accounting;To reduce the generation of mistake.Template data module is according to voice in the present invention
In the relevant number of prompt selection or text template procedure, using the Feature Correspondence Algorithm progress detected based on salient region
Match, improves the operational efficiency and accuracy rate of system.Voice input module acquires user by microphone apparatus in the present invention
Voice, and by the voice transfer of acquisition to central control module, many noises containing miscellaneous person during voice collecting are adopted
With noise classification voice enhancement algorithm, the clarity of accounting voice can be improved, convenient for the identification of system.Mould is stored in the present invention
The data such as the audio of modules, text, number are carried out effective classification storage by block, and raising system is to the effect for establishing report
Rate;The storage space of storage module is saved simultaneously.
Detailed description of the invention
Fig. 1 is accounting language identification information input device identification information input method process provided in an embodiment of the present invention
Figure.
Fig. 2 is accounting language identification information input device structural schematic diagram provided in an embodiment of the present invention.
In figure: 1, voice input module;2, template data module;3, monitoring checks module;4, central processing module;5, it shows
Show module;6, storage module.
Specific embodiment
In order to further understand the content, features and effects of the present invention, the following examples are hereby given, and cooperate attached drawing
Detailed description are as follows.
Structure of the invention is explained in detail with reference to the accompanying drawing.
As shown in Figure 1, accounting language identification information input device provided in an embodiment of the present invention identifies input method, specifically
The following steps are included:
S101: firstly, the voice of acquisition user, and by the voice transfer of acquisition to data processing system;
S102: during acquisition, recording all voices of the accounting in voice input, convenient for the inquiry past in later period
The data of storage;
S103: according to the relevant number of voice prompting selection or text template, analog signal is converted by voice, will be simulated
Signal is converted into digital signal, and digital signal and waveform database are contrasted;Then it is split into single phoneme, finally according to sound
The raw new word language in position queries dictionary library or sentence, are finally completed the work of words input;
S104: according to the data of typing, system production static state and dynamic financial report;
Content after speech recognition: being shown by S105 using display screen, and the audio to modules, text,
The data such as number are stored.
As shown in Figure 1, a kind of accounting language identification information input device is provided with voice input module 1, template data mould
Block 2, monitoring check module 3, central processing module 4, display module 5, storage module 6.
Voice input module 1;It is connect with central control module 4, the voice of user is acquired by microphone apparatus, and
By the voice transfer of acquisition to central control module.
Template data module 2;It is connect with central processing module 4, according to the relevant number of voice prompting selection or text mould
Plate, the data storage of template data module 2 is in storage module 6.
Monitoring checks module 3;It is connect with central processing module 4;All voices of the accounting in voice input are recorded, just
The data of storage in the past are inquired in the later period.
Central processing module 4;With voice input module 1, template data module 2, monitoring check module 3, display module 5,
Storage module connection 6, converts analog signal for voice, converts digital signal, digital signal and waveform number for analog signal
It is contrasted according to library, is then split into single phoneme, finally according to the raw new word language in phoneme queries dictionary library or sentence, be finally completed
The work of words input.
Display module 5;It connect with central processing module 4, is shown the content after speech recognition by display screen.
Storage module 6 is attached with central processing module 4, audio, text, the number etc. for storing modules
Data.
The voice input module 1 acquires the voice of user by microphone apparatus, and by the voice transfer of acquisition
To central control module, many noises containing miscellaneous person during voice collecting, using noise classification voice enhancement algorithm, specifically
The following steps are included:
Step 1 acquires noisy accounting voice, obtains the preceding 120ms of noisy speech as noise;
Step 2 carries out feature extraction to the noise of acquisition, carries out noise classification, example: Traffic, Machine to noise
Noise;
Step 3 is chosen voice enhancement algorithm according to noise type, is enhanced accounting voice.
The template data module 2 selects in relevant number or text template procedure according to voice prompting, in order to improve
The relevant number of voice prompting selection or text template are specific to wrap using the Feature Correspondence Algorithm detected based on salient region
Include following steps:
Step 1 establishes corresponding set to number or text template and voice prompting data respectively;
Step 2 carries out region detection according to corresponding set is established, obtains characteristic, feature is described;
Step 3 carries out characteristic matching according to the characteristic of acquisition, rejects error hiding.
The data such as the audio of modules, text, number are carried out effective classification storage by the storage module 6, are improved
To the efficiency for establishing report, storage module 6 is provided with data acquisition system and establishes module, data combining classification module and data knot system
Close spider module;
Data acquisition system establishes module, inputs the raw data sets such as audio, text, the number of modules;To initial data
Collection establishes set D, isolates most classes and minority class sample, if several classes of sample sets are L, minority class sample set is S, calculates
The number of samples Num for being included in conjunction with L1And the number of samples Num that set S is included2;
Data combining classification module separates Category Attributes set X and connection attribute set Y in D;
Data combine traversal categorization module, traverse respectively to Category Attributes set X and connection attribute set Y, calculate
Respective index filters out Category Attributes set;The Category Attributes set respectively filtered out is merged, classification data is obtained
Set.
The working principle of the invention: the sound of accounting is carried out by typing by voice input module 1, and sound is carried out
Framing realized using move function window, and the length of every frame is 25ms, there is that 15ms's is overlapping between every two frame, by sound into
The conversion of traveling wave shape extracts MFCC feature according to the physiological property of human ear and each frame waveform is become a multi-C vector, then
Analog signal is converted to by central processing module 4, digital signal and waveform database are contrasted, and are then split into single sound
Position is finally completed the work of words input, passes through display module 5 finally according to the raw new word language in phoneme queries dictionary library or sentence
The conversion of number, text in the process is shown, by storage module 6 to the process of entire speech recognition carry out text into
Row storage.
Application principle of the present invention is further described below with reference to specific experiment.
The present invention enhances Traffic noise using in noise classification voice enhancement algorithm, is subtracted using selection spectrum
Method;Machine noise is enhanced, and using selection Vienna filtering algorithm, experimental data matching result is as follows:
1 Traffic matching result of table
2 Machine matching result of table
The above is only the preferred embodiments of the present invention, and is not intended to limit the present invention in any form,
Any simple modification made to the above embodiment according to the technical essence of the invention, equivalent variations and modification, belong to
In the range of technical solution of the present invention.
Claims (2)
1. a kind of accounting language identification information input device, which is characterized in that the accounting language identification information input device packet
It includes:
Voice input module, template data module, monitoring check module, central processing module, display module, storage module.
Voice input module;It is connect with central control module, the voice of user is acquired by microphone apparatus, and will acquisition
Voice transfer to central control module.
Template data module;It is connect with central processing module, it is described according to the relevant number of voice prompting selection or text template
The data storage of template data module is in storage module.
Monitoring checks module;It is connect with central processing module;All voices of the accounting in voice input are recorded, the later period is convenient for
The data of inquiry storage in the past.
Central processing module;Check that module, display module, storage module connect with voice input module, template data module, monitoring
It connects, converts analog signal for voice, convert digital signal for analog signal, digital signal and waveform database are formed pair
Than being then split into single phoneme, finally according to the raw new word language in phoneme queries dictionary library or sentence, being finally completed words input
Work.
Display module;It connect with central processing module, is shown the content after speech recognition by display screen.
Storage module is attached with central processing module, the data such as audio, text, number for storing modules.
2. accounting language identification information input device as described in claim 1, which is characterized in that the voice input module exists
When carrying out the typing of sound, sound being subjected to framing, is realized using move function window, the length of every frame is 25ms, every two
There is overlapping for 15ms between frame.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910355212.4A CN110047465A (en) | 2019-04-29 | 2019-04-29 | A kind of accounting language identification information input device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910355212.4A CN110047465A (en) | 2019-04-29 | 2019-04-29 | A kind of accounting language identification information input device |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110047465A true CN110047465A (en) | 2019-07-23 |
Family
ID=67280115
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910355212.4A Pending CN110047465A (en) | 2019-04-29 | 2019-04-29 | A kind of accounting language identification information input device |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110047465A (en) |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1342967A (en) * | 2000-09-13 | 2002-04-03 | 中国科学院自动化研究所 | Unified recognizing method for multi-speed working pattern |
US20130110511A1 (en) * | 2011-10-31 | 2013-05-02 | Telcordia Technologies, Inc. | System, Method and Program for Customized Voice Communication |
CN103578464A (en) * | 2013-10-18 | 2014-02-12 | 威盛电子股份有限公司 | Language model establishing method, speech recognition method and electronic device |
CN106557971A (en) * | 2016-11-18 | 2017-04-05 | 畅捷通信息技术股份有限公司 | Based on the financial data processing method of speech recognition technology, system and terminal |
CN108984510A (en) * | 2018-09-12 | 2018-12-11 | 山西中电科新能源技术有限公司 | By voice by the system of data input table |
-
2019
- 2019-04-29 CN CN201910355212.4A patent/CN110047465A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1342967A (en) * | 2000-09-13 | 2002-04-03 | 中国科学院自动化研究所 | Unified recognizing method for multi-speed working pattern |
US20130110511A1 (en) * | 2011-10-31 | 2013-05-02 | Telcordia Technologies, Inc. | System, Method and Program for Customized Voice Communication |
CN103578464A (en) * | 2013-10-18 | 2014-02-12 | 威盛电子股份有限公司 | Language model establishing method, speech recognition method and electronic device |
CN106557971A (en) * | 2016-11-18 | 2017-04-05 | 畅捷通信息技术股份有限公司 | Based on the financial data processing method of speech recognition technology, system and terminal |
CN108984510A (en) * | 2018-09-12 | 2018-12-11 | 山西中电科新能源技术有限公司 | By voice by the system of data input table |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108564942B (en) | Voice emotion recognition method and system based on adjustable sensitivity | |
CN102723078B (en) | Emotion speech recognition method based on natural language comprehension | |
CN110364143A (en) | Voice awakening method, device and its intelligent electronic device | |
CN105845127A (en) | Voice recognition method and system | |
CN105938399B (en) | The text input recognition methods of smart machine based on acoustics | |
CN110570873B (en) | Voiceprint wake-up method and device, computer equipment and storage medium | |
CN106504754A (en) | A kind of real-time method for generating captions according to audio output | |
CN102298694A (en) | Man-machine interaction identification system applied to remote information service | |
CN107393554A (en) | In a kind of sound scene classification merge class between standard deviation feature extracting method | |
CN107644643A (en) | A kind of voice interactive system and method | |
CN110379441B (en) | Voice service method and system based on countermeasure type artificial intelligence network | |
CN109036467A (en) | CFFD extracting method, speech-emotion recognition method and system based on TF-LSTM | |
CN103531206B (en) | A kind of local speech emotional characteristic extraction method with global information of combination | |
Deshmukh et al. | Speech based emotion recognition using machine learning | |
Wang et al. | Research on speech emotion recognition technology based on deep and shallow neural network | |
CN111326178A (en) | Multi-mode speech emotion recognition system and method based on convolutional neural network | |
CN111312292A (en) | Emotion recognition method and device based on voice, electronic equipment and storage medium | |
CN109935226A (en) | A kind of far field speech recognition enhancing system and method based on deep neural network | |
CN103871417A (en) | Specific continuous voice filtering method and device of mobile phone | |
CN105869636A (en) | Speech recognition apparatus and method thereof, smart television set and control method thereof | |
CN106710591A (en) | Voice customer service system for power terminal | |
CN104679733B (en) | A kind of voice dialogue interpretation method, apparatus and system | |
CN110390937A (en) | A kind of across channel method for recognizing sound-groove based on ArcFace loss algorithm | |
CN109300478A (en) | A kind of auxiliary Interface of person hard of hearing | |
CN106778443A (en) | A kind of method that assisting blind understands expression |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190723 |