TWI558589B - Combining smart phone and voice recognition car anti - theft system and its method - Google Patents

Combining smart phone and voice recognition car anti - theft system and its method Download PDF

Info

Publication number
TWI558589B
TWI558589B TW103122723A TW103122723A TWI558589B TW I558589 B TWI558589 B TW I558589B TW 103122723 A TW103122723 A TW 103122723A TW 103122723 A TW103122723 A TW 103122723A TW I558589 B TWI558589 B TW I558589B
Authority
TW
Taiwan
Prior art keywords
module
voice
theft
user
car
Prior art date
Application number
TW103122723A
Other languages
Chinese (zh)
Other versions
TW201601954A (en
Inventor
Wei Ho Tsai
Yen Hung Lin
Original Assignee
Univ Nat Taipei Technology
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Univ Nat Taipei Technology filed Critical Univ Nat Taipei Technology
Priority to TW103122723A priority Critical patent/TWI558589B/en
Publication of TW201601954A publication Critical patent/TW201601954A/en
Application granted granted Critical
Publication of TWI558589B publication Critical patent/TWI558589B/en

Links

Landscapes

  • Burglar Alarm Systems (AREA)

Description

結合智慧型手機及語音辨識之汽車防盜系統及其方法 Automobile anti-theft system combining smart phone and voice recognition and method thereof

本發明係關於一種結合智慧型手機及語音辨識之汽車防盜系統及其方法,特別是指一種能夠以智慧型手機做為汽車遙控器使用、並配合語音身份確認的機制,以達到更安全的防護。 The invention relates to a car anti-theft system and a method thereof combining smart phone and voice recognition, in particular to a mechanism capable of using a smart phone as a car remote controller and supporting voice identity confirmation, so as to achieve more secure protection. .

目前市面上大多數的汽車防盜器皆是透過手持式遙控器來進行控制,但此方式不僅容易因忘記攜帶遙控器而造成不便,更可能因遺失遙控器而造成重大損失。 At present, most car alarm devices on the market are controlled by a hand-held remote controller, but this method is not only inconvenient due to forgetting to carry the remote controller, but also may cause significant loss due to loss of the remote controller.

因此,現在有廠商嘗試以手機來做為遙控汽車防盜器的產品,其中主要是使用3G的行動通信技術做為車輛訊息傳遞媒介。雖然沒有遙控距離的限制,但是在沒有3G訊號的環境下,此防盜系統將無法工作,例如在地下停車場,或是一些偏僻的郊區或山區。此外,使用3G需在車上的防盜器主機內擺放一張行動電話用戶識別(Subscriber Identity Module;SIM)卡,因而衍生出上網費和月租費的問題。 Therefore, some manufacturers are now trying to use mobile phones as products for remote-controlled car alarms, mainly using 3G mobile communication technology as a vehicle message delivery medium. Although there is no remote distance limitation, in the absence of 3G signals, the anti-theft system will not work, such as in underground parking lots, or in some remote suburbs or mountains. In addition, the use of 3G requires the placement of a Subscriber Identity Module (SIM) card in the car's anti-theft device, which leads to the problem of Internet access and monthly fees.

為了解決以上問題,本發明亦是嘗試以智慧型手機來取代傳統車用防盜系統的遙控器,而本發明則是使用藍芽作為汽車防盜器與使用者溝通的媒介,藉以取代傳統遙控器;另外,為了避免因手機遺失而造成進一步車輛遭竊的重大損失,我們更透過語音身分確認來達成更安全的防 護,如此應為一最佳解決方案。 In order to solve the above problems, the present invention also attempts to replace the remote controller of the conventional vehicle anti-theft system with a smart phone, and the present invention uses the Bluetooth as a medium for communication between the car alarm and the user, thereby replacing the conventional remote controller; In addition, in order to avoid the loss of further vehicle theft caused by the loss of the mobile phone, we also use voice identity confirmation to achieve a more secure defense. This should be an optimal solution.

本發明係關於一種結合智慧型手機及語音辨識之汽車防盜系統及其方法,係能夠以智慧型手機做為汽車遙控器使用、並配合語音身份確認的機制,以達到更安全的汽車遙控器使用與防護。 The invention relates to a car anti-theft system and a method thereof combining smart phone and voice recognition, which can use a smart phone as a car remote controller and cooperate with a voice identity confirmation mechanism to achieve a safer car remote control use. And protection.

可達成上述結合智慧型手機及語音辨識之汽車防盜系統及其方法,其中該結合智慧型手機及語音辨識之汽車防盜系統,係包含:一智慧型手機,內部係至少包含有一藍芽模組、一收音模組、一儲存模組及一與該藍芽模組、收音模組、儲存模組相連接之汽車防盜軟體,其中該汽車防盜軟體能夠控制該收音模組對一使用者發出的語音進行收音為一錄音檔,並儲存於該儲存模組內,而該汽車防盜軟體更能夠對與該智慧型手機之藍芽模組所配對之裝置,藉由該藍芽模組傳送該錄音檔;一汽車本體,內部係至少裝設有一汽車防盜主機及至少一個警報器,其中該汽車防盜主機內係包含有一中央處理模組、一指令執行模組、一藍芽模組、一語音特徵參數擷取模組、一語者相似度判斷模組、一語音辨識模組及一儲存模組,其中該指令執行模組、藍芽模組、語音特徵參數擷取模組、語者相似度判斷模組、語音辨識模組及儲存模組係與該中央處理模組相連接;其中,該汽車防盜主機之藍芽模組與該智慧型手機之藍芽模組配對後,該汽車防盜主機能夠將接收之錄音檔藉由該語音特徵參數擷取模組進行擷取語音特徵參數,並透過該語者相似度判斷模組進行比對後,若判斷為使用者本人,則能夠進一步以該語音辨識模組進行語音辨識,並將辨識所得之指令內容輸入該指令執行模組中執行動作;反之,若判斷為非使用者本人,則透過 該中央處理模組傳送一訊號至該警報器,以啟動該警報器。 The above-mentioned automobile anti-theft system and method for combining the smart phone and the voice recognition can be achieved, wherein the car anti-theft system combining the smart phone and the voice recognition includes: a smart phone, the internal system includes at least one Bluetooth module, a radio module, a storage module and a car anti-theft software connected to the Bluetooth module, the radio module and the storage module, wherein the car anti-theft software can control the voice of the radio module to a user The audio is recorded as a recording file and stored in the storage module, and the car anti-theft software is capable of transmitting the recording file to the device paired with the Bluetooth module of the smart phone by the Bluetooth module. a car body, the interior is equipped with at least one car anti-theft host and at least one alarm, wherein the car anti-theft host includes a central processing module, an instruction execution module, a Bluetooth module, and a voice feature parameter The module, the speaker similarity judgment module, a voice recognition module and a storage module, wherein the instruction execution module, the Bluetooth module, and the voice feature The digital capture module, the speaker similarity judgment module, the voice recognition module and the storage module are connected to the central processing module; wherein the Bluetooth module of the automobile security host and the blue of the smart phone After the bud module is paired, the car anti-theft host can capture the voice feature parameters by using the voice feature parameter capture module, and compare the parameters through the speaker similarity judgment module. For the user himself, the speech recognition module can further perform speech recognition, and input the recognized instruction content into the instruction execution module to perform an action; otherwise, if it is determined that the user is not the user, then The central processing module transmits a signal to the alarm to activate the alarm.

更具體的說,所述儲存模組係用以儲存使用者語音模型、聲學模型及門檻值。 More specifically, the storage module is used to store user speech models, acoustic models, and threshold values.

更具體的說,所述汽車防盜主機能夠將接收之錄音檔藉由該語音特徵參數擷取模組進行擷取語音特徵參數,並透過該語者相似度判斷模組將擷取之語音特徵參數與使用者語音模型進行比對,即可取得相似度分數,若相似度分數大於門檻值,則判斷為授權使用者本人。 More specifically, the car anti-theft host can capture the voice feature parameters by using the voice feature parameter capture module, and use the speaker similarity judgment module to capture the voice feature parameters. The similarity score can be obtained by comparing with the user voice model, and if the similarity score is greater than the threshold value, it is determined that the user is authorized.

更具體的說,所述判斷為使用者本人後,則將所擷取之語音特徵參數進行語音辨識,該語音辨識是透過該語音辨識模組將所擷取之語音特徵參數與聲學模型進行比對,再將比對所得之音節轉換為所對應的指令內容。 More specifically, after the user is determined as the user, the voice feature parameters captured are voice-recognized, and the voice recognition module compares the captured voice feature parameters with the acoustic model through the voice recognition module. Yes, the syllables obtained by the comparison are converted into the corresponding instruction contents.

本發明之結合智慧型手機及語音辨識之汽車防盜方法,其方法為:1.開啟一智慧型手機內之汽車防盜軟體,該軟體使手機與汽車本體之防盜主機進行藍芽配對,並對使用者發出的語音進行收音,之後,再透過藍芽將使用者發出的語音傳送至該汽車防盜主機;2.再由該汽車防盜主機對使用者發出的語音進行擷取語音特徵參數,並進行語者相似度比對、以判斷是否為授權使用者本人;以及3.若判斷結果為授權使用者本人,則進一步將所擷取之語音特徵參數進行語音辨識,並將辨識所得之指令內容輸入該汽車防盜主機中執行動作;反之,若判斷為非授權使用者本人,則進行警報警示。 The invention relates to a smart phone and a voice recognition automobile anti-theft method, the method is as follows: 1. Open a car anti-theft software in a smart phone, the software enables the mobile phone to be blue-paired with the anti-theft host of the car body, and is used The voice is transmitted by the voice, and then the voice sent by the user is transmitted to the car anti-theft host through the Bluetooth; 2. The car anti-theft host then captures the voice feature parameters of the voice sent by the user, and speaks Comparing the similarity to determine whether the authorized user is the user; and 3. if the judgment result is the authorized user himself, further performing the voice recognition on the captured voice feature parameter, and inputting the recognized instruction content into the The car anti-theft host performs the action; otherwise, if it is determined that the unauthorized user is himself, the alarm is displayed.

更具體的說,所述汽車防盜主機內能夠儲存至少一組的使用 者語音模型、至少一組的聲學模型及至少一組的門檻值。 More specifically, the car anti-theft host can store at least one group of uses. A speech model, at least one set of acoustic models, and at least one set of threshold values.

更具體的說,所述汽車防盜主機能夠將擷取之語音特徵參數與使用者語音模型進行比對,即可取得相似度分數,若相似度分數大於門檻值,則判斷為授權使用者本人。 More specifically, the car anti-theft host can compare the captured speech feature parameters with the user speech model to obtain a similarity score, and if the similarity score is greater than the threshold, it is determined to authorize the user himself.

更具體的說,所述判斷為授權使用者本人後,則將所擷取之語音特徵參數進行語音辨識,該語音辨識是將所擷取之語音特徵參數與聲學模型進行比對,再將比對所得之音節轉換為所對應的指令內容。 More specifically, after the determination is to authorize the user himself, the voice feature parameters captured are voice-recognized, and the voice recognition is to compare the captured voice feature parameters with the acoustic model, and then compare Convert the resulting syllable to the corresponding instruction content.

〔本發明〕 〔this invention〕

1‧‧‧智慧型手機 1‧‧‧Smart mobile phone

11‧‧‧藍芽模組 11‧‧‧Bluetooth Module

12‧‧‧收音模組 12‧‧‧ Radio Module

13‧‧‧儲存模組 13‧‧‧ Storage Module

14‧‧‧汽車防盜軟體 14‧‧‧Automobile anti-theft software

2‧‧‧汽車本體 2‧‧‧Automobile body

21‧‧‧汽車防盜主機 21‧‧‧Car Guard

211‧‧‧中央處理模組 211‧‧‧Central Processing Module

212‧‧‧指令執行模組 212‧‧‧Command Execution Module

213‧‧‧藍芽模組 213‧‧‧Bluetooth Module

214‧‧‧語音特徵參數擷取模組 214‧‧‧Voice feature parameter acquisition module

215‧‧‧語者相似度判斷模組 215‧‧‧Speaker similarity judgment module

216‧‧‧語音辨識模組 216‧‧‧Voice recognition module

217‧‧‧儲存模組 217‧‧‧ storage module

22‧‧‧警報器 22‧‧‧Announcer

201‧‧‧方法1 201‧‧‧Method 1

202‧‧‧方法2 202‧‧‧Method 2

203‧‧‧方法3 203‧‧‧Method 3

第1A圖係本發明結合智慧型手機及語音辨識之汽車防盜系統及其方法之整體架構示意圖。 FIG. 1A is a schematic diagram showing the overall structure of a car anti-theft system and a method thereof combined with a smart phone and a voice recognition.

第1B圖係本發明結合智慧型手機及語音辨識之汽車防盜系統及其方法之汽車防盜主機內部架構示意圖。 FIG. 1B is a schematic diagram of the internal structure of the automobile anti-theft host combining the smart phone and the voice recognition automobile anti-theft system and the method thereof.

第2圖係本發明結合智慧型手機及語音辨識之汽車防盜系統及其方法之流程示意圖。 FIG. 2 is a schematic flow chart of the automobile anti-theft system and the method thereof combined with the smart phone and the voice recognition.

有關於本發明之前述及其他技術內容、特點與功效,在以下配合參考圖式之較佳實施例的詳細說明中,將可清楚的呈現。 The above and other technical contents, features and advantages of the present invention will be apparent from the following detailed description of the preferred embodiments.

請參閱第1A圖及第1B圖,為本發明結合智慧型手機及語音辨識之汽車防盜系統及其方法之整體架構示意圖,由圖中可知,該結合智慧型手機及語音辨識之汽車防盜系統,係包含一智慧型手機1及一汽車本體2,其中該智慧型手機1內部係至少包含有一藍芽模組11、一收音模 組12、一儲存模組13及一汽車防盜軟體14,該汽車防盜軟體14會與該藍芽模組11、收音模組12、儲存模組13相連接,而當開啟該汽車防盜軟體14時,透過該藍芽模組11則會列出所偵測到的藍芽裝置,當使用者點選該汽車本體2上所設置的藍芽模組時,則會與該汽車本體2上所設置的藍芽模組進行配對,而通常該手機第一次完成配對後,於下次使用者開啟防盜軟體則自動配對。另外該汽車防盜軟體14能夠控制該收音模組12對一使用者所發出的語音進行收音為一錄音檔,因此使用者能夠操作該汽車防盜軟體14(例如按下螢幕中間的按鈕說出欲操作防盜系統的指令)進行收音,而當放開按鈕後,則會將使用者所說的語音存成錄音檔,並再儲存於該儲存模組13內,而該汽車防盜軟體14則透過藍芽將錄音檔傳送給該汽車本體2之汽車防盜主機21。 Please refer to FIG. 1A and FIG. 1B , which are schematic diagrams showing the overall architecture of a vehicle anti-theft system and a method for combining the smart phone and the voice recognition according to the present invention. As can be seen from the figure, the car anti-theft system combining the smart phone and the voice recognition, The system includes a smart phone 1 and a car body 2, wherein the smart phone 1 has at least one Bluetooth module 11 and a radio module. The group 12, a storage module 13 and a car anti-theft software 14, the car anti-theft software 14 will be connected with the Bluetooth module 11, the radio module 12, the storage module 13, and when the car anti-theft software 14 is turned on Through the Bluetooth module 11, the detected Bluetooth device will be listed. When the user clicks on the Bluetooth module set on the car body 2, it will be set with the car body 2. The Bluetooth module is paired, and usually the phone is automatically paired the next time the user turns on the anti-theft software after the first pairing is completed. In addition, the car anti-theft software 14 can control the sound module 12 to collect the voice emitted by a user into a recording file, so the user can operate the car anti-theft software 14 (for example, pressing a button in the middle of the screen to say that the user wants to operate The anti-theft system commands) to receive the radio, and when the button is released, the user's voice is stored in the recording file and stored in the storage module 13, and the car anti-theft software 14 transmits the blueprint. The recording file is transmitted to the car anti-theft host 21 of the car body 2.

而該汽車本體2內部係至少裝設有一汽車防盜主機21及至少一個警報器22,其中該汽車防盜主機21內係包含有一中央處理模組211、一指令執行模組212、一藍芽模組213、一語音特徵參數擷取模組214、一語者相似度判斷模組215、一語音辨識模組216及一儲存模組217,而當該汽車防盜主機21接收之錄音檔後,能夠藉由該語音特徵參數擷取模組214進行擷取語音特徵參數,而本實施例中則是使用標準的梅爾頻率倒頻譜係數(Mel-Frequency Cepstrum Coefficients;MFCC)做為特徵參數,以擷取梅爾頻率倒頻譜參數(Mel-Frequency Cepstrum Coefficients;MFCC)。 The car body 2 is equipped with at least one car anti-theft host 21 and at least one alarm device 22, wherein the car anti-theft host 21 includes a central processing module 211, an instruction execution module 212, and a Bluetooth module. 213, a voice feature parameter capture module 214, a speaker similarity determination module 215, a voice recognition module 216 and a storage module 217, and when the car anti-theft host 21 receives the recording file, can borrow The speech feature parameter capture module 214 captures the speech feature parameters, and in this embodiment, the standard Mel-Frequency Cepstrum Coefficients (MFCC) is used as the feature parameter to capture Mel-Frequency Cepstrum Coefficients (MFCC).

擷取語音特徵參數完成後,則透過該語者相似度判斷模組215將擷取之語音特徵參數與使用者語音模型進行比對,而在辨識之前必須先建立使用者的語音模型,本實施例中則是使用標準的高斯混合模型 (Gaussian Mixture Model;GMM),也就是說儲存模組217中必須要有使用者的高斯混合模型,並於擷取語音特徵參數完成後,則計算其輸入的語音與被宣稱的語者模型(授權使用者的語音模型)是否有夠大的相似度,也就是分析語音中的語音特徵是否足夠證明。若宣稱人的模型為λ,計算輸入音檔的語音特徵序列X對此模型的相似度,Pr(X|λ),再用聯合背景模型(Universal Background Model;UBM)計算「非宣稱人」的相似度,Pr(X|λ UBM),其中聯合背景模型λ UBM是由大量非該授權使用者的語音所建立出的語者模型。當求出log Pr(X|λ)和log Pr(X|λ UBM)的差值後,再設一個門檻值(Threshold)θ,來做為判斷的標準,如下1.1所示: After the speech feature parameter is completed, the speech similarity judgment module 215 compares the captured speech feature parameter with the user speech model, and the user's speech model must be established before the identification. In the example, a standard Gaussian Mixture Model (GMM) is used, that is, the storage module 217 must have a Gaussian mixture model of the user, and after the speech feature parameters are completed, the input is calculated. Whether the voice has a sufficiently large similarity to the claimed speaker model (the voice model of the authorized user), that is, whether the voice features in the voice are sufficient to prove. If the model of the person is declared as λ , the similarity of the speech feature sequence X of the input sound file is calculated, Pr( X | λ ), and then the "Unclaimed Person" is calculated by the Universal Background Model (UBM). Similarity, Pr( X | λ UBM ), where the joint background model λ UBM is a speaker model established by a large number of voices other than the authorized user. When the difference between log Pr( X | λ ) and log Pr( X | λ UBM ) is found, a threshold value (Threshold) θ is set as the criterion for judgment, as shown in the following 1.1:

而當判斷為使用者本人後,則將所擷取之語音特徵參數進行語音辨識,本實施例中語音辨識模組216是基於隱藏式馬可夫模型(Hidden Markov Model;HMM)方法來進行辨識,辨認單位為音節。 After the user is judged as the user, the voice feature parameters captured are voice-recognized. In this embodiment, the voice recognition module 216 is based on the Hidden Markov Model (HMM) method for identification and identification. The unit is a syllable.

因此經由該語音辨識模組216將擷取之語音特徵參數與聲學模型進行辨識後,再將比對所得之音節轉換為所對應的指令內容,則能夠輸入該指令執行模組212中,以執行使用者以語音所下達的指令(例如開鎖、上鎖、啟動、開後行李廂、警戒)。 Therefore, after the voice recognition module 216 recognizes the captured voice feature parameter and the acoustic model, and converts the obtained syllable into the corresponding instruction content, the voice execution module 212 can be input into the instruction execution module 212 to execute. The instructions issued by the user in voice (such as unlocking, locking, starting, opening the trunk, alerting).

但當語音相似度判斷模組215判斷為非授權使用者本人,之後,則透過該中央處理模組211傳送一訊號至該警報器22,以啟動該警報器22,以警示有非授權使用者本人欲操作該汽車防盜軟體14。 However, when the voice similarity determination module 215 determines that the user is an unauthorized user, a signal is transmitted to the alarm device 22 through the central processing module 211 to activate the alarm device 22 to alert the unauthorized user. I want to operate the car anti-theft software 14.

而本發明之結合智慧型手機及語音辨識之汽車防盜方法,如 第2圖所示,其方法為:1.開啟一智慧型手機內之汽車防盜軟體,與一汽車本體之汽車防盜主機進行藍芽配對,並對使用者發出的語音進行收音,之後,再將使用者發出的語音傳送至該汽車防盜主機201;2.再由該汽車防盜主機對使用者發出的語音進行擷取語音特徵參數,並進行語者相似度比對、以進行判斷是否為授權使用者本人202;以及3.若判斷結果為授權使用者本人,則進一步將擷取之語音特徵參數進行語音辨識後,將辨識所得之指令內容輸入該汽車防盜主機中執行動作;反之,若判斷為非授權使用者本人,則進行警報警示203。 The car anti-theft method combining the smart phone and the voice recognition of the invention, such as As shown in Fig. 2, the method is as follows: 1. Turn on the car anti-theft software in a smart phone, pair with the car anti-theft host of a car body, and listen to the voice sent by the user, and then, The voice sent by the user is transmitted to the car anti-theft host 201; 2. The car anti-theft host captures the voice feature parameters of the voice sent by the user, and compares the similarity of the speakers to determine whether the authorization is authorized. 202; and 3. If the result of the judgment is the authorized user, the speech feature of the captured speech feature is further subjected to speech recognition, and the recognized instruction content is input into the car anti-theft host to perform an action; If the unauthorized user himself/herself, the police will give a warning 203.

本發明所提供之結合智慧型手機及語音辨識之汽車防盜系統及其方法,與其他習用技術相互比較時,其優點如下: The automobile anti-theft system and the method thereof combined with the smart phone and the voice recognition provided by the invention are compared with other conventional technologies, and the advantages thereof are as follows:

1.本發明能夠以智慧型手機做為汽車遙控器使用、並配合語音身份確認的機制,以達到更安全的汽車遙控器使用與防護。 1. The invention can use the smart phone as a car remote controller and cooperate with the voice identity confirmation mechanism to achieve safer use and protection of the car remote controller.

2.本發明可以避免因手機遺失而造成進一步車輛遭竊的重大損失,我們透過語音身分確認來達成更安全的防護。此系統結合了身分確認及語音辨識的功能,讓使用者在使用上多了一份安全的保護機制,不會因為手機不慎遺失或遭人竊取,而造成車上財物損失或車子失竊,並且透過指令操作操控防盜器,增加其使用的便利性。 2. The invention can avoid the great loss of further vehicle theft caused by the loss of the mobile phone, and we can achieve safer protection through voice identity confirmation. This system combines the functions of identity confirmation and voice recognition, allowing users to have a safe protection mechanism in use, without causing loss of property or car theft due to accidental loss or theft of the mobile phone. Operate the anti-theft device through command operations to increase the convenience of its use.

藉由以上較佳具體實施例之詳述,係希望能更加清楚描述本發明之特徵與精神,而並非以上述所揭露的較佳具體實施例來對本發明之 範疇加以限制。相反地,其目的是希望能涵蓋各種改變及具相等性的安排於本發明所欲申請之專利範圍的範疇內。 The features and spirit of the present invention are more clearly described in the above detailed description of the preferred embodiments of the present invention. The scope is limited. On the contrary, the intention is to cover various modifications and equivalents within the scope of the invention as claimed.

1‧‧‧智慧型手機 1‧‧‧Smart mobile phone

11‧‧‧藍芽模組 11‧‧‧Bluetooth Module

12‧‧‧收音模組 12‧‧‧ Radio Module

13‧‧‧儲存模組 13‧‧‧ Storage Module

14‧‧‧汽車防盜軟體 14‧‧‧Automobile anti-theft software

2‧‧‧汽車本體 2‧‧‧Automobile body

21‧‧‧汽車防盜主機 21‧‧‧Car Guard

22‧‧‧警報器 22‧‧‧Announcer

Claims (8)

一種結合智慧型手機及語音辨識之汽車防盜系統,係包含:一智慧型手機,內部係至少包含有一藍芽模組、一收音模組、一儲存模組及一與該藍芽模組、收音模組、儲存模組相連接之汽車防盜軟體,其中該汽車防盜軟體能夠控制該收音模組對一使用者發出的語音進行收音為一錄音檔,並儲存於該儲存模組內,而該汽車防盜軟體更能夠對與該智慧型手機之藍芽模組所配對之裝置,藉由該藍芽模組傳送該錄音檔;一汽車本體,內部係至少裝設有一汽車防盜主機及至少一個警報器,其中該汽車防盜主機內係包含有一中央處理模組、一指令執行模組、一藍芽模組、一語音特徵參數擷取模組、一語者相似度判斷模組、一語音辨識模組及一儲存模組,其中該指令執行模組、藍芽模組、語音特徵參數擷取模組、語者相似度判斷模組、語音辨識模組及儲存模組係與該中央處理模組相連接;其中,該汽車防盜主機之藍芽模組與該智慧型手機之藍芽模組配對後,該汽車防盜主機能夠將接收之錄音檔藉由該語音特徵參數擷取模組進行擷取語音特徵參數,並透過該語者相似度判斷模組進行比對後,若判斷為授權使用者本人,則能夠進一步以該語音辨識模組進行語音辨識,並將辨識所得之指令內容輸入該指令執行模組中執行動作;反之,若判斷為非授權使用者本人,之後,則透過該中央處理模組傳送一訊號至該警報器,以啟動該警報器。 A car anti-theft system combining smart phone and voice recognition includes: a smart phone, the internal system includes at least one Bluetooth module, a radio module, a storage module, and a Bluetooth module, and a radio The car anti-theft software connected to the module and the storage module, wherein the car anti-theft software can control the sound module to receive a voice recorded by a user into a recording file and store in the storage module, and the car The anti-theft software can further transmit the recording file to the device paired with the Bluetooth module of the smart phone; the car body has at least one car anti-theft host and at least one alarm device in the interior The car anti-theft host includes a central processing module, an instruction execution module, a Bluetooth module, a speech feature parameter capture module, a speaker similarity determination module, and a speech recognition module. And a storage module, wherein the instruction execution module, the Bluetooth module, the voice feature parameter capture module, the speaker similarity judgment module, the voice recognition module, and the storage module are The central processing module is connected; wherein the Bluetooth module of the automobile anti-theft host is paired with the Bluetooth module of the smart phone, the car anti-theft host can capture the received recording file by the voice feature parameter The group performs the speech feature parameter, and after comparing the similarity judgment module of the language, if it is determined that the user is authorized, the speech recognition module can further perform speech recognition, and the obtained instruction is recognized. The content input is executed in the instruction execution module; otherwise, if it is determined that the unauthorized user is himself, then a signal is transmitted to the alarm through the central processing module to activate the alarm. 如申請專利範圍第1項所述之結合智慧型手機及語音辨識之汽車防盜 系統,其中該汽車防盜主機之儲存模組係用以儲存使用者語音模型、聲學模型及門檻值。 The scope of the patent application of paragraph 1 in combination with the smart phone and voice recognition of car alarm system, wherein the storage of the host vehicle anti-theft module is used for storing user speech model, the acoustic model and the threshold value. 如申請專利範圍第2項所述之結合智慧型手機及語音辨識之汽車防盜系統,其中該汽車防盜主機能夠將接收之錄音檔藉由該語音特徵參數擷取模組進行擷取語音特徵參數,並透過該語音相似度判斷模組將擷取之語音特徵參數與使用者語音模型進行比對,即可取得相似度分數,若相似度分數大於門檻值,則判斷為授權使用者本人。 The automobile anti-theft system combining the smart phone and the voice recognition according to the second aspect of the patent application, wherein the car anti-theft host can capture the voice feature parameters by using the voice feature parameter capturing module. And the voice similarity judgment module compares the captured voice feature parameter with the user voice model to obtain a similarity score, and if the similarity score is greater than the threshold value, it is determined to authorize the user himself. 如申請專利範圍第2項所述之結合智慧型手機及語音辨識之汽車防盜系統,其中判斷為授權使用者本人後,則將擷取之語音特徵參數進行語音辨識,該語音辨識是透過該語音辨識模組將擷取之語音特徵參數與聲學模型進行比對後,再將比對所得之音節轉換為所對應的指令內容。 For example, in the automobile anti-theft system combining the smart phone and the voice recognition described in the second aspect of the patent application, wherein the user is authorized to perform the voice recognition, the voice recognition parameter is voice-transmitted through the voice. The identification module compares the captured speech feature parameters with the acoustic model, and then converts the syllable obtained by the comparison into the corresponding instruction content. 一種結合智慧型手機及語音辨識之汽車防盜方法,其方法為:開啟一智慧型手機內之汽車防盜軟體,並對使用者發出的語音進行收音,之後,再與一汽車本體之汽車防盜主機進行藍芽配對後,將使用者發出的語音傳送至該汽車防盜主機;再由該汽車防盜主機對使用者發出的語音進行擷取語音特徵參數,並進行語者相似度比對、以進行判斷是否為授權使用者本人;以及若判斷結果為授權使用者本人,則進一步將擷取之語音特徵參數進行語音辨識後,將辨識所得之指令內容輸入該汽車防盜主機中執行指令;反之,若判斷為非授權使用者本人,則進行警報警示。 The invention relates to a car anti-theft method combining smart phone and voice recognition, which comprises: opening a car anti-theft software in a smart phone, and collecting the voice sent by the user, and then performing a car anti-theft host with a car body. After the Bluetooth pairing, the voice sent by the user is transmitted to the car anti-theft host; and the car anti-theft host captures the voice feature parameters of the voice sent by the user, and compares the similarity of the speakers to determine whether In order to authorize the user himself; and if the result of the judgment is the authorized user himself, after the voice feature parameter is further subjected to voice recognition, the recognized instruction content is input into the car anti-theft host to execute the instruction; If the unauthorized user is himself, a warning will be given. 如申請專利範圍第5項所述之結合智慧型手機及語音辨識之汽車防盜方法,其中該汽車防盜主機內能夠儲存至少一組的使用者語音模型、至 少一組的聲學模型及至少一組的門檻值。 The automobile anti-theft method combining the smart phone and the voice recognition according to claim 5 , wherein the car anti-theft host can store at least one set of user voice models, at least one set of acoustic models, and at least one group Threshold value. 如申請專利範圍第6項所述之結合智慧型手機及語音辨識之汽車防盜方法,其中該汽車防盜主機能夠將擷取之語音特徵參數與使用者語音模型進行比對,即可取得相似度分數,若相似度分數大於門檻值,則判斷為使用者本人。 For example, the automobile anti-theft method combining the smart phone and the voice recognition according to the sixth aspect of the patent application, wherein the car anti-theft host can compare the captured voice feature parameters with the user voice model to obtain the similarity score. If the similarity score is greater than the threshold value, it is determined as the user himself. 如申請專利範圍第7項所述之結合智慧型手機及語音辨識之汽車防盜方法,其中判斷為授權使用者本人後,則將擷取之語音特徵參數進行語音辨識,該語音辨識是將擷取之語音特徵參數與聲學模型進行比對後,並再將比對所得之音節轉換為所對應的指令內容。 For example, in the vehicle anti-theft method combining the smart phone and the voice recognition described in claim 7 , wherein the user is authorized to perform the voice recognition, the voice recognition parameter is captured. After the speech feature parameters are compared with the acoustic model, the syllables obtained by the comparison are converted into corresponding instruction contents.
TW103122723A 2014-07-01 2014-07-01 Combining smart phone and voice recognition car anti - theft system and its method TWI558589B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
TW103122723A TWI558589B (en) 2014-07-01 2014-07-01 Combining smart phone and voice recognition car anti - theft system and its method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
TW103122723A TWI558589B (en) 2014-07-01 2014-07-01 Combining smart phone and voice recognition car anti - theft system and its method

Publications (2)

Publication Number Publication Date
TW201601954A TW201601954A (en) 2016-01-16
TWI558589B true TWI558589B (en) 2016-11-21

Family

ID=55641470

Family Applications (1)

Application Number Title Priority Date Filing Date
TW103122723A TWI558589B (en) 2014-07-01 2014-07-01 Combining smart phone and voice recognition car anti - theft system and its method

Country Status (1)

Country Link
TW (1) TWI558589B (en)

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
TW200610361A (en) * 2003-11-07 2006-03-16 F S W Image module and display in electronic device, system, and method of recognition and control and else apparatus and else method
TW200722316A (en) * 2005-12-06 2007-06-16 Ind Tech Res Inst Intelligent monitoring system for vehicle using wireless network
TW200922818A (en) * 2007-11-16 2009-06-01 Inventec Appliances Corp A car with a burglarproof system
CN202180801U (en) * 2011-07-29 2012-04-04 广东亚太天能信息识别技术有限公司 Automobile anti-theft system with phonetic recognition function
TWI409179B (en) * 2010-07-19 2013-09-21 Geee Creations Inc Vehicle burglarproof system using bluetooth handheld device for control
CN103684773A (en) * 2012-09-24 2014-03-26 国民技术股份有限公司 Secret lock device, portable terminal and audio unlocking method based on portable terminal

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
TW200610361A (en) * 2003-11-07 2006-03-16 F S W Image module and display in electronic device, system, and method of recognition and control and else apparatus and else method
TW200722316A (en) * 2005-12-06 2007-06-16 Ind Tech Res Inst Intelligent monitoring system for vehicle using wireless network
TW200922818A (en) * 2007-11-16 2009-06-01 Inventec Appliances Corp A car with a burglarproof system
TWI409179B (en) * 2010-07-19 2013-09-21 Geee Creations Inc Vehicle burglarproof system using bluetooth handheld device for control
CN202180801U (en) * 2011-07-29 2012-04-04 广东亚太天能信息识别技术有限公司 Automobile anti-theft system with phonetic recognition function
CN103684773A (en) * 2012-09-24 2014-03-26 国民技术股份有限公司 Secret lock device, portable terminal and audio unlocking method based on portable terminal

Also Published As

Publication number Publication date
TW201601954A (en) 2016-01-16

Similar Documents

Publication Publication Date Title
US10586534B1 (en) Voice-controlled device control using acoustic echo cancellation statistics
US10915614B2 (en) Biometric authentication
US11475899B2 (en) Speaker identification
CN106463112B (en) Voice recognition method, voice awakening device, voice recognition device and terminal
US20130332165A1 (en) Method and systems having improved speech recognition
US9646610B2 (en) Method and apparatus for activating a particular wireless communication device to accept speech and/or voice commands using identification data consisting of speech, voice, image recognition
JP2000080828A (en) Vehicle control device
CN111131601B (en) Audio control method, electronic equipment, chip and computer storage medium
JP6011584B2 (en) Speech recognition apparatus and speech recognition system
CN105448294A (en) Intelligent voice recognition system for vehicle equipment
CN108463369B (en) Vehicle infotainment and connectivity system
KR102417524B1 (en) Speech recognition based vehicle control method
US10997971B2 (en) Wakeword detection using a secondary microphone
US20210312911A1 (en) Wakeword detection using a secondary microphone
US20190244612A1 (en) Method of processing a speech signal for speaker recognition and electronic apparatus implementing same
CN109830234A (en) A kind of intelligent vehicle-carried information interaction device and exchange method
KR20140067687A (en) Car system for interactive voice recognition
JP2016133378A (en) Car navigation device
TWI558589B (en) Combining smart phone and voice recognition car anti - theft system and its method
JP2006025079A (en) Head set and wireless communication system
KR102406115B1 (en) System and method for automatic control of vehicle audio
US20240071392A1 (en) Upgrade method, upgrade apparatus, and electronic device
WO2020167385A1 (en) Wakeword detection using a secondary microphone
JP7310594B2 (en) Vehicle communication system
TW202046061A (en) Method of starting voice assistant and electronic device with voice assistant

Legal Events

Date Code Title Description
MM4A Annulment or lapse of patent due to non-payment of fees