TWI544358B - Rfid acess control method and system combining face recognition based on secret splitting and rfid tag used therein - Google Patents

Rfid acess control method and system combining face recognition based on secret splitting and rfid tag used therein Download PDF

Info

Publication number
TWI544358B
TWI544358B TW104101867A TW104101867A TWI544358B TW I544358 B TWI544358 B TW I544358B TW 104101867 A TW104101867 A TW 104101867A TW 104101867 A TW104101867 A TW 104101867A TW I544358 B TWI544358 B TW I544358B
Authority
TW
Taiwan
Prior art keywords
registered
information
face
frequency face
low frequency
Prior art date
Application number
TW104101867A
Other languages
Chinese (zh)
Other versions
TW201627901A (en
Inventor
林文暉
王平
黃千華
劉克祥
吳保樺
Original Assignee
崑山科技大學
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 崑山科技大學 filed Critical 崑山科技大學
Priority to TW104101867A priority Critical patent/TWI544358B/en
Application granted granted Critical
Publication of TWI544358B publication Critical patent/TWI544358B/en
Publication of TW201627901A publication Critical patent/TW201627901A/en

Links

Description

結合基於秘密分持臉部辨識的射頻辨識門禁方法及系統以及 適用於其之射頻辨識標籤 Radio frequency identification access control method and system based on secret separation facial recognition and RF identification tag for it

本發明是有關於一種射頻辨識門禁方法及系統以及適用於其之射頻辨識標籤,特別是有關於一種結合基於秘密分持臉部辨識的射頻辨識門禁方法及系統以及適用於其之射頻辨識標籤。 The invention relates to a radio frequency identification access control method and system and a radio frequency identification tag suitable for the same, in particular to a radio frequency identification access control method and system based on secret separation facial recognition and a radio frequency identification tag suitable for the same.

近年來,愈來愈多的人們採用自動感應式門禁系統,取代原本既有之鑰匙孔門禁設備,藉以增加使用之方便性及安全性。其中,射頻辨識(radio frequency identification,RFID)技術為一廣泛使用之技術,其係於門禁設備中配置有射頻辨識讀取器(RFID reader),且令使用者攜帶具有射頻辨識標籤(RFID tag)之物件,當使用者讓射頻辨識標籤靠近或對準射頻辨識讀取器時,射頻辨識讀取器會驗證此射頻辨識標籤是否合法,藉以決定是否讓使用者進入。 In recent years, more and more people have adopted automatic inductive access control systems to replace the existing keyhole access control devices, thereby increasing the convenience and safety of use. Among them, radio frequency identification (RFID) technology is a widely used technology, which is equipped with an RFID reader in the access control device, and allows the user to carry an RFID tag. The object, when the user brings the RFID tag close to or aligned with the RFID reader, the RFID reader verifies whether the RFID tag is legal, thereby determining whether to let the user enter.

然而,目前習知之射頻辨識技術大多具有認卡不認人之驗證問題。因此,一些學者提出了於射頻辨識門禁系統中整合臉部辨識之技術來解決此驗證問題。但是,此些整合臉部辨識之射頻辨識門禁系統多半會將臉部辨識 資料全部儲存於後端資料庫中,當待辨識者靠近射頻辨識門禁系統後再將擷取到之影像與後端資料庫中之多個臉部辨識資料做依序辨識,藉以判斷待辨識者是否為合格之使用者。如此一來,射頻辨識門禁系統需要花費一段時間進行比對,且若後端資料庫遭到竊取,則會造成個人隱私問題。 However, most of the conventional radio frequency identification technologies have the problem of verifying that the card is not recognized. Therefore, some scholars have proposed to integrate face recognition technology in the RFID access control system to solve this verification problem. However, most of these RF recognition access control systems that integrate face recognition will recognize the face. The data is stored in the back-end database. When the person to be identified is close to the RFID access control system, the captured image and the plurality of face recognition data in the back-end database are sequentially identified to determine the identifier to be identified. Is it a qualified user? As a result, the RFID access control system takes a while to compare, and if the back-end database is stolen, it will cause personal privacy problems.

有鑑於上述習知技藝之問題,本發明之目的就是在提供一種結合基於秘密分持臉部辨識的射頻辨識門禁方法及系統以及適用於其之射頻辨識標籤,藉以進一步更增加使用上之方便性及安全性。 In view of the above problems of the prior art, the object of the present invention is to provide a radio frequency identification access control method and system based on secret separation face recognition and a radio frequency identification tag suitable for the same, thereby further increasing the convenience of use. And security.

為達上述目的,本發明提供一種結合基於秘密分持臉部辨識的射頻辨識門禁系統,至少包含:射頻辨識標籤,儲存有至少一註冊低頻臉資訊;資料庫,儲存有複數個註冊高頻臉編碼資訊;以及門禁辨識裝置,電性連接資料庫,此門禁辨識裝置至少包含:射頻辨識讀取器,電性通訊射頻辨識標籤,此射頻辨識讀取器驗證射頻辨識標籤是否合法;影像擷取元件,擷取待辨識者之人臉影像;以及處理元件,電性連接射頻辨識讀取器及影像擷取元件,此處理元件藉由小波轉換技術將人臉影像分解成較低頻分量之待辨識低頻臉資訊及較高頻分量之待辨識高頻臉資訊,且透過編碼技術編碼待辨識高頻臉資訊以產生待辨識高頻臉編碼資訊,其中處理元件透過射頻辨識讀取器從通過驗證之射頻辨識標籤處取得註冊低頻臉資訊,並比對待辨識低頻臉資訊是否符合註冊低頻臉資訊,若“否”,則處理元件判定待辨識者不合格,若“是”,則處理元件從資料庫中取出對應於註冊低頻臉資訊之註冊高頻臉編碼資訊,並比對待辨識高頻 臉編碼資訊是否符合取出之註冊高頻臉編碼資訊,若“否”,則處理元件判定待辨識者不合格,若“是”,則處理元件判定待辨識者合格。 To achieve the above objective, the present invention provides a radio frequency identification access control system combining secret recognition based facial recognition, comprising at least: an RFID tag storing at least one registered low frequency face information; a database storing a plurality of registered high frequency faces The coded information; and the access control device, the electrical connection database, the access control device comprises at least: a radio frequency identification reader, an electrical communication radio frequency identification tag, the radio frequency identification reader verifies whether the radio frequency identification tag is legal; the image capture The component captures the face image of the person to be identified; and the processing component is electrically connected to the RFID reader and the image capturing component, and the processing component decomposes the face image into lower frequency components by using wavelet transform technology Identifying the low frequency face information and the higher frequency component of the high frequency face information to be recognized, and encoding the high frequency face information to be identified by the coding technique to generate the high frequency face coded information to be recognized, wherein the processing component passes the verification through the RFID reader Obtain the registered low frequency face information at the RFID tag, and compare with the recognition of the low frequency face information. Frequency face information, if "No", the processing component determines that the person to be identified is unqualified. If "Yes", the processing component retrieves the registered high frequency face code information corresponding to the registered low frequency face information from the database, and compares the identification high frequency Whether the face coding information meets the registered high frequency face coding information, if "No", the processing component determines that the person to be identified is unqualified. If "Yes", the processing component determines that the person to be identified is qualified.

其中,門禁辨識裝置更包含近端儲存元件,電性連接處理元件。 當處理元件判定待辨識低頻臉資訊符合註冊低頻臉資訊後,先搜索對應於此註冊低頻臉資訊之註冊高頻臉編碼資訊是否儲存於近端儲存元件中,若“是”,則處理元件從近端儲存元件中取出此註冊高頻臉編碼資訊,若“否”,則處理元件再從資料庫中取出對應於此註冊低頻臉資訊之註冊高頻臉編碼資訊。 The access control device further includes a near-end storage component and is electrically connected to the processing component. After the processing component determines that the low-frequency face information to be recognized meets the registered low-frequency face information, first searching whether the registered high-frequency face code information corresponding to the registered low-frequency face information is stored in the near-end storage component, and if “yes”, the processing component is The registered high frequency face code information is taken out from the near end storage component. If "No", the processing component then retrieves the registered high frequency face code information corresponding to the registered low frequency face information from the database.

其中,註冊高頻臉編碼資訊及待辨識高頻臉編碼資訊係各別為區域二元化圖型特徵直方圖頻譜或各別為區域三元化圖型特徵直方圖頻譜。 The registered high frequency face coding information and the high frequency face coding information to be identified are respectively a histogram spectrum of the regional dualization pattern feature or a histogram spectrum of the regional ternary pattern characteristic.

此外,本發明更提供一種結合基於秘密分持臉部辨識的射頻辨識門禁方法,至少包含下列步驟:利用射頻辨識讀取器驗證射頻辨識標籤是否合法,其中射頻辨識標籤儲存有至少一註冊低頻臉資訊;利用影像擷取元件擷取待辨識者之人臉影像;當判定射頻辨識標籤合法後,利用處理元件接收此人臉影像,並藉由小波轉換技術將此人臉影像分解成較低頻分量之待辨識低頻臉資訊及較高頻分量之待辨識高頻臉資訊;利用處理元件透過編碼技術編碼待辨識高頻臉資訊以產生待辨識高頻臉編碼資訊;利用處理元件透過射頻辨識讀取器從通過驗證之射頻辨識標籤處取得其所儲存之註冊低頻臉資訊;利用處理元件比對待辨識低頻臉資訊是否符合此註冊低頻臉資訊,若“否”,則處理元件判定待辨識者不合格;若處理元件判定待辨識低頻臉資訊符合此註冊低頻臉資訊,則利用處理元件從資料庫中取出對應於此註冊低頻臉資訊之一註冊高頻臉編碼資訊;以及利用處理元件比對待辨識高頻臉編碼資訊是否符合取出之註冊高頻臉 編碼資訊,若“否”,則處理元件判定待辨識者不合格,若“是”,則處理元件判定待辨識者合格。 In addition, the present invention further provides a radio frequency identification access control method based on secret separation facial recognition, comprising at least the following steps: verifying whether the radio frequency identification tag is legal by using a radio frequency identification reader, wherein the radio frequency identification tag stores at least one registered low frequency face. Information; using the image capture component to capture the face image of the person to be identified; after determining that the RFID tag is legal, the processing component is used to receive the face image, and the face image is decomposed into lower frequencies by wavelet transform technology. The component of the low frequency face information to be identified and the higher frequency component to be identified by the high frequency face information; the processing component is used to encode the high frequency face information to be recognized by the coding technique to generate the high frequency face coded information to be recognized; and the processing component is read by the radio frequency identification The device obtains the registered low-frequency face information stored from the verified RFID tag; and uses the processing component to identify whether the low-frequency face information meets the registered low-frequency face information. If “No”, the processing component determines that the identifier is not recognized. Qualified; if the processing component determines that the low frequency face information to be recognized meets the registered low frequency face , By the processing element is removed from the library corresponding to this registered face information registered one low-frequency coding information face; and using a high frequency processing element than the treated face identification whether the encoding information extracted high frequency registration of the face If the information is "No", the processing component determines that the person to be identified is unqualified. If "Yes", the processing component determines that the person to be identified is qualified.

其中,於判定待辨識低頻臉資訊符合射頻辨識標籤中之註冊低頻臉資訊後,先搜索對應於此註冊低頻臉資訊之註冊高頻臉編碼資訊是否儲存於近端儲存元件中,若“是”,則處理元件從近端儲存元件中取出此註冊高頻臉編碼資訊,若“否”,則處理元件再從資料庫中取出對應於此註冊低頻臉資訊之註冊高頻臉編碼資訊。 After determining that the low frequency face information to be recognized meets the registered low frequency face information in the radio frequency identification tag, first searching whether the registered high frequency face coding information corresponding to the registered low frequency face information is stored in the near end storage component, if “yes” The processing component extracts the registered high frequency face code information from the near end storage component. If “No”, the processing component retrieves the registered high frequency face code information corresponding to the registered low frequency face information from the database.

其中,本發明之結合基於秘密分持臉部辨識的射頻辨識門禁方法,更包含下列步驟:利用處理元件藉由小波轉換技術將註冊人臉影像分解成較低頻分量之註冊低頻臉資訊及較高頻分量之註冊高頻臉資訊;利用處理元件透過編碼技術編碼註冊高頻臉資訊以產生註冊高頻臉編碼資訊,其中註冊高頻臉編碼資訊對應於註冊低頻臉資訊;儲存註冊低頻臉資訊於射頻辨識標籤中;以及儲存註冊高頻臉編碼資訊於資料庫中。 The combination of the present invention is based on the secret identification method for identifying the face recognition, and further includes the following steps: using the processing component to decompose the registered face image into the lower frequency component of the lower frequency component by using the wavelet transform technology and comparing Registering high-frequency face information of high-frequency components; using the processing component to encode high-frequency face information by encoding technology to generate registered high-frequency face code information, wherein registering high-frequency face coded information corresponds to registering low-frequency face information; storing registered low-frequency face information In the RFID tag; and store the registered high frequency face code information in the database.

其中,註冊高頻臉編碼資訊及待辨識高頻臉編碼資訊係各別為區域二元化圖型特徵直方圖頻譜或各別為區域三元化圖型特徵直方圖頻譜。 The registered high frequency face coding information and the high frequency face coding information to be identified are respectively a histogram spectrum of the regional dualization pattern feature or a histogram spectrum of the regional ternary pattern characteristic.

其中,於比對待辨識低頻臉資訊是否符合註冊低頻臉資訊之步驟中,係使用最近鄰演算法計算待辨識低頻臉資訊與註冊低頻臉資訊之間之尺度空間距離,並判斷此尺度空間距離是否小於低頻階段門檻值,若“是”,則判定此待辨識低頻臉資訊符合此註冊低頻臉資訊。 In the step of comparing whether the low-frequency face information meets the registered low-frequency face information, the closest neighbor algorithm is used to calculate the scale space distance between the low-frequency face information to be recognized and the registered low-frequency face information, and determine whether the scale space distance is It is less than the threshold value of the low frequency stage. If “Yes”, it is determined that the low frequency face information to be recognized meets the registered low frequency face information.

其中,於比對待辨識高頻臉編碼資訊是否符合取出之註冊高頻臉編碼資訊之步驟中,係使用最近鄰演算法計算待辨識高頻臉編碼資訊與註冊高 頻臉編碼資訊之間之距離值,並判斷此距離值是否小於高頻階段門檻值,若“是”,則判定此待辨識高頻臉編碼資訊符合取出之註冊高頻臉編碼資訊。 Wherein, in the step of comparing whether the high-frequency face coded information is in conformity with the registered high-frequency face coded information, the nearest neighbor algorithm is used to calculate the high-frequency face coded information to be recognized and the registration is high. The distance value between the frequency-coded information is determined, and whether the distance value is less than the high-frequency threshold value, and if YES, it is determined that the high-frequency face coded information to be recognized conforms to the registered high-frequency face coded information.

另外,本發明更提供一種用於結合基於秘密分持臉部辨識的射頻辨識門禁系統中之射頻辨識標籤,至少包含:標籤元件,利用射頻辨識技術與門禁辨識裝置進行通訊,使得門禁辨識裝置驗證射頻辨識標籤是否合法;以及儲存元件,電性連接標籤元件並儲存有至少一註冊低頻臉資訊,當門禁辨識裝置驗證射頻辨識標籤合法後,儲存元件將註冊低頻臉資訊傳送至門禁辨識裝置,以使得門禁辨識裝置比對註冊低頻臉資訊與使用射頻辨識標籤之待辨識者之待辨識低頻臉資訊是否相符。 In addition, the present invention further provides an RFID tag for use in a radio frequency identification access control system based on secret separation face recognition, comprising at least: a tag component, which communicates with the access control device by using radio frequency identification technology, so that the access control device is verified. Whether the RFID tag is legal; and the storage component is electrically connected to the tag component and stores at least one registered low frequency face information. When the access control device verifies that the RFID tag is legal, the storage component transmits the registered low frequency face information to the access control device, The access control device compares whether the registered low frequency face information matches the to-be-identified low frequency face information of the to-be-identified person using the RFID tag.

因此,依本發明之結合基於秘密分持臉部辨識的射頻辨識門禁方法及系統以及適用於其之射頻辨識標籤,可具有一或多個下述優點: Therefore, the radio frequency identification access control method and system based on secret separation facial recognition and the radio frequency identification tag suitable for the same according to the present invention may have one or more of the following advantages:

(1)藉由將比對用之註冊人臉影像利用小波轉換技術及編碼技術分解並編碼成較低頻分量之註冊低頻臉資訊及較高頻分量之註冊高頻臉編碼資訊,並將註冊低頻臉資訊儲存於射頻辨識標籤中,以及將對應於此註冊低頻臉資訊之註冊高頻臉編碼資訊儲存於資料庫中,藉以使得當射頻辨識標籤或資料庫中之任一者遺失或被竊取資料時,其均無法取得完整之個人身分資料,因此可避免個人隱私權遭到侵犯之問題以及提供更安全的門禁安全機制。 (1) Decomposing and encoding the registered face image for comparison using the wavelet transform technique and coding technique into the registered low frequency face information of the lower frequency component and the registered high frequency face coded information of the higher frequency component, and registering The low frequency face information is stored in the RFID tag, and the registered high frequency face code information corresponding to the registered low frequency face information is stored in the database, so that any one of the RFID tag or the database is lost or stolen. When the information is available, it is unable to obtain complete personal identity information, thus avoiding the violation of personal privacy rights and providing a safer access control security mechanism.

(2)藉由將註冊低頻臉資訊儲存於射頻辨識標籤中,藉以使得本發明之門禁辨識裝置可比對待辨識低頻臉資訊及射頻辨識標籤中之註冊低頻臉資訊,而無須一一比對資料庫中之多個低頻臉資訊,因此可加速辨識速率。此外,藉由先比對射頻辨識標籤中之註冊低頻臉資訊,再比對資料庫中之註冊高頻臉編碼資訊,以及註冊高頻臉編碼資訊係對應於註冊低頻臉資訊,藉以使得當比 對待辨識高頻臉編碼資訊及註冊高頻臉編碼資訊時,可直接取得並比對對應於註冊低頻臉資訊之註冊高頻臉編碼資訊,而無須一一比對資料庫中之多個註冊高頻臉編碼資訊,因此可進一步加速辨識速率。 (2) By storing the registered low frequency face information in the RFID tag, the access control device of the present invention can compare the registered low frequency face information in the low frequency face information and the RFID tag without having to compare the database. Multiple low frequency face information in the medium, thus speeding up the recognition rate. In addition, by comparing the registered low frequency face information in the RFID tag, the registered high frequency face code information in the database, and the registered high frequency face code information corresponding to the registered low frequency face information, so that the ratio is When the high frequency face coding information and the registered high frequency face coding information are to be recognized, the registered high frequency face coding information corresponding to the registered low frequency face information can be directly obtained and compared without having to register multiple registrations in the database. Frequency face coding information, so the recognition rate can be further accelerated.

100‧‧‧射頻辨識標籤 100‧‧‧RF tags

110‧‧‧標籤元件 110‧‧‧Label components

120‧‧‧儲存元件 120‧‧‧Storage components

200‧‧‧資料庫 200‧‧‧Database

300‧‧‧門禁辨識裝置 300‧‧‧Access identification device

310‧‧‧射頻辨識讀取器 310‧‧‧RF reader

320‧‧‧影像擷取元件 320‧‧‧Image capture component

330‧‧‧處理元件 330‧‧‧Processing components

340‧‧‧近端儲存元件 340‧‧‧ Near-end storage components

400‧‧‧出入口 400‧‧‧ entrances and exits

810-992‧‧‧步驟 810-992‧‧‧Steps

第1圖係為本發明之結合基於秘密分持臉部辨識的射頻辨識門禁系統之系統方塊圖。 FIG. 1 is a system block diagram of a radio frequency identification access control system based on secret separation facial recognition according to the present invention.

第2圖係為本發明之結合基於秘密分持臉部辨識的射頻辨識門禁系統之實施樣態示意圖。 FIG. 2 is a schematic diagram showing the implementation of the radio frequency identification access control system based on secret separation facial recognition according to the present invention.

第3圖係為本發明之結合基於秘密分持臉部辨識的射頻辨識門禁方法於註冊階段之步驟流程圖。 Figure 3 is a flow chart showing the steps of the radio frequency identification access control method based on secret separation face recognition in the registration phase of the present invention.

第4圖係為本發明之結合基於秘密分持臉部辨識的射頻辨識門禁方法於驗證階段之步驟流程圖。 Figure 4 is a flow chart showing the steps of the RF identification access control method based on secret separation face recognition in the verification phase of the present invention.

請參閱第1圖至第4圖,其中第1圖係為本發明之結合基於秘密分持臉部辨識的射頻辨識門禁系統之系統方塊圖,第2圖係為本發明之結合基於秘密分持臉部辨識的射頻辨識門禁系統之實施樣態示意圖,第3圖係為本發明之結合基於秘密分持臉部辨識的射頻辨識門禁方法於註冊階段之步驟流程圖,第4圖係為本發明之結合基於秘密分持臉部辨識的射頻辨識門禁方法於驗證階段之步驟流程圖。 Please refer to FIG. 1 to FIG. 4 , wherein FIG. 1 is a system block diagram of a radio frequency identification access control system based on secret separation face recognition according to the present invention, and FIG. 2 is a combination based on secret separation according to the present invention. A schematic diagram of the implementation of the radio frequency identification access control system for face recognition, and FIG. 3 is a flow chart of the steps of the radio frequency identification access control method based on secret separation facial recognition in the registration phase, and FIG. 4 is a flowchart of the present invention. The combination of the radio frequency identification access control method based on secret separation facial recognition is a flow chart of steps in the verification phase.

如第1圖至第4圖所示,本發明之結合基於秘密分持臉部辨識的射頻辨識門禁系統至少包含射頻辨識標籤(radio frequency identification tag,RFID tag)100、資料庫200以及門禁辨識裝置300。其中,射頻辨識標籤100儲存有一或多個註冊低頻臉資訊,資料庫200則儲存有複數個註冊高頻臉編碼資訊,而門禁辨識裝置300可例如裝設於出入口400上或出入口400之側邊,其中出入口400可例如為住家門口、工廠門口、廠房門口、車門口、校門口、房門口或其他需要人員出入管制之場所出入口,其中此門禁辨識裝置300電性連接資料庫200且至少包含射頻辨識讀取器(RFID reader)310、影像擷取元件320以及處理元件330。 其中,資料庫200可例如為一大容量儲存元件或雲端資料庫。此外,射頻辨識讀取器310電性通訊射頻辨識標籤100並驗證此射頻辨識標籤100是否合法。而影像擷取元件320擷取待辨識者之人臉影像,其中影像擷取元件320可例如為照相設備、攝影設備或攝像設備等可以拍攝影像之裝置。另外,處理元件330則電性連接射頻辨識讀取器310、影像擷取元件320以及資料庫200,並進行註冊及驗證之處理動作。 As shown in FIG. 1 to FIG. 4, the radio frequency identification access control system based on secret separation facial recognition includes at least a radio frequency identification tag (RFID tag) 100, a data library 200, and an access control device. 300. The radio frequency identification tag 100 stores one or more registered low frequency face information, and the database 200 stores a plurality of registered high frequency face coding information, and the access control device 300 can be installed on the side of the entrance or exit 400 or the side of the entrance and exit 400, for example. The access port 400 can be, for example, a doorway of a home, a door of a factory, a door of a factory, a door of a door, a door of a school, a door of a door, or other entrances and exits that require access control by a person. The access control device 300 is electrically connected to the database 200 and includes at least a radio frequency. An RFID reader 310, an image capture component 320, and a processing component 330 are provided. The database 200 can be, for example, a large-capacity storage element or a cloud database. In addition, the RFID reader 310 electrically communicates with the RFID tag 100 and verifies whether the RFID tag 100 is legitimate. The image capturing component 320 captures a human face image of the person to be recognized. The image capturing component 320 can be, for example, a device capable of capturing images, such as a camera device, a photographic device, or an imaging device. In addition, the processing component 330 is electrically connected to the RFID reader 310, the image capturing component 320, and the database 200, and performs processing operations of registration and verification.

舉例來說,射頻辨識標籤100可至少包含標籤元件110以及儲存元件120,其中標籤元件110可例如建置有射頻辨識通訊元件,藉以利用射頻辨識(RFID)技術與門禁辨識裝置300進行通訊,使得門禁辨識裝置300可藉由此進行之通訊而驗證射頻辨識標籤100是否合法。而儲存元件120則電性連接標籤元件110並儲存有一或多個註冊低頻臉資訊,當門禁辨識裝置300驗證射頻辨識標籤100合法後,儲存元件120可將儲存之註冊低頻臉資訊傳送至門禁辨識裝置300,以使得門禁辨識裝置300可將使用影像擷取元件320所擷取到之待辨識者之待辨 識低頻臉資訊,與從射頻辨識標籤100之儲存元件120中所取得之註冊低頻臉資訊進行比對其是否相符,藉以第一步驗證此待辨識者是否合法。 For example, the RFID tag 100 can include at least a tag component 110 and a storage component 120. The tag component 110 can be configured, for example, with a radio frequency identification communication component to communicate with the access control device 300 using radio frequency identification (RFID) technology. The access identification device 300 can verify whether the RFID tag 100 is legal by the communication performed thereby. The storage component 120 is electrically connected to the tag component 110 and stores one or more registered low frequency face information. When the access control device 300 verifies that the RFID tag 100 is legal, the storage component 120 can transmit the stored registered low frequency face information to the access control identifier. The device 300 is configured such that the access control device 300 can identify the person to be identified using the image capturing component 320 The low frequency face information is compared with the registered low frequency face information obtained from the storage component 120 of the RFID tag 100 to determine whether the person to be identified is legal.

其中,本發明之結合基於秘密分持臉部辨識的射頻辨識門禁系統可透過一註冊階段以儲存註冊低頻臉資訊於射頻辨識標籤100中以及儲存註冊高頻臉編碼資訊於資料庫200中。舉例來說,門禁辨識裝置300可先取得至少一註冊人臉影像(步驟810),此註冊人臉影像可例如為從一外部裝置(未繪示)透過任一之通訊協定傳送至門禁辨識裝置300,亦或者,可於此註冊階段拍攝此註冊人以得到註冊人臉影像。其中,每一註冊人可例如具有一或多張註冊人臉影像以各別表示其不同之狀態(例如正面嚴肅臉、側臉或笑臉等),藉以增加之後驗證時之方便性。 The radio frequency identification access control system based on the secret separation facial recognition can store the registered low frequency face information in the radio frequency identification tag 100 and store the registered high frequency face coding information in the data base 200 through a registration stage. For example, the access control device 300 can first obtain at least one registered face image (step 810), and the registered face image can be transmitted to the access control device, for example, from an external device (not shown) through any communication protocol. 300. Alternatively, the registrant may be photographed at this registration stage to obtain a registered face image. Each of the registrants may, for example, have one or more registered face images to indicate their different states (for example, a serious face, a face or a smile), thereby increasing the convenience of subsequent verification.

而當門禁辨識裝置300取得註冊人臉影像後,可利用處理元件330藉由小波轉換技術將此註冊人臉影像分解成較低頻分量之註冊低頻臉資訊及較高頻分量之註冊高頻臉資訊(步驟820)。其中於步驟820中,處理元件330可先將此註冊人臉影像轉換成圖像頻譜,並根據圖像頻譜中之頻段將註冊人臉影像分解成較低頻分量之註冊低頻臉資訊及較高頻分量之註冊高頻臉資訊。其中,註冊高頻臉資訊可例如具有各別表示三個方向(橫向、直向及斜向)之三個高頻臉資訊。另外,處理元件330亦可依據註冊人臉影像之解析度,進行一或多階之小波轉換,藉以取得特定尺寸、辨識度或解析度之註冊低頻臉資訊,之後再儲存此註冊低頻臉資訊於射頻辨識標籤100中(步驟830)。其中,註冊低頻臉資訊可例如具有人臉之較多共同特徵,例如人臉輪廓、臉部器官之位置等基本圖像;而註冊高頻臉資訊則含有大多數之臉部局部變化,例如光照、表情或裝扮等。其中,由於註冊低頻臉資訊僅具有基本圖像,且射頻辨識標籤100僅儲存有註冊低頻臉 資訊,而未儲存有註冊高頻臉資訊,因此若射頻辨識標籤100發生遺失之狀況,其亦不會讓他人取得完整之個人身分資料,故可避免個人隱私權遭到侵犯之問題以及提供更安全的門禁安全機制。 After the access control device 300 obtains the registered face image, the processing component 330 can be used to decompose the registered face image into the registered low frequency face information of the lower frequency component and the registered high frequency face of the higher frequency component by using the wavelet transform technology. Information (step 820). In step 820, the processing component 330 may first convert the registered face image into an image spectrum, and decompose the registered face image into a lower frequency component of the lower frequency component according to the frequency band in the image spectrum. Registration of high frequency face information of frequency components. The registered high frequency face information may have, for example, three high frequency face information indicating three directions (horizontal, straight, and oblique). In addition, the processing component 330 can also perform one or more wavelet transforms according to the resolution of the registered face image, thereby obtaining the registered low frequency face information of a specific size, identification or resolution, and then storing the registered low frequency face information. The RFID tag 100 is in the process (step 830). The registered low frequency face information may have, for example, more common features of the face, such as a face contour, a position of a facial organ, and the like; and the registered high frequency face information contains most of the facial local changes, such as illumination. , expressions or dress up. Wherein, since the registered low frequency face information only has the basic image, and the radio frequency identification tag 100 only stores the registered low frequency face Information, but there is no registered high-frequency face information, so if the RFID tag 100 is lost, it will not allow others to obtain complete personal identity information, so it can avoid the problem of personal privacy violation and provide more Secure access control security mechanism.

另外,處理元件330可透過編碼技術編碼取得之註冊高頻臉資訊以產生註冊高頻臉編碼資訊(步驟840)。舉例來說,當處理元件330於步驟820中取得註冊高頻臉資訊後,可利用區域二元化圖型(local binary patterns,LBP)技術、區域三元化圖型(local ternary patterns,LTP)技術或其他適合之編碼技術對此註冊高頻臉資訊進行編碼處理,以產生區域二元化圖型特徵直方圖頻譜、區域三元化圖型特徵直方圖頻譜或其他不可逆之註冊高頻臉編碼資訊。以區域二元化圖型(LBP)技術為例,處理元件330可先將註冊高頻臉資訊中之每個像素點依序作為中心點,且以此中心點與其鄰近的像素形成n*n(例如3*3)之區塊,並以中心點之灰階值作為比較基值,若區塊內之鄰域像素灰階值大於或等於此比較基值時,則此區塊內之此鄰域像素點編碼為“1”,否則編碼為“0”。之後,給予此區塊內之每一鄰域像素點一個不同之權重值(例如1、2、4、8...或128之一者)且此些權重值不重複,再加總此區塊內之編碼與權重值相乘後之數值,以得到代表此中心點之數值。之後,組合此些數值以得到註冊高頻臉資訊所對應之註冊高頻臉編碼資訊。 In addition, the processing component 330 can encode the registered high frequency face information obtained by the encoding technique to generate the registered high frequency face coding information (step 840). For example, after the processing component 330 obtains the registered high frequency face information in step 820, the local binary patterns (LBP) technology and the local ternary patterns (LTP) may be utilized. The technology or other suitable coding technique encodes the registered high frequency face information to generate a region binarized pattern feature histogram spectrum, a regional ternary pattern feature histogram spectrum or other irreversible registered high frequency face coding. News. Taking the area binary pattern (LBP) technology as an example, the processing component 330 may firstly mark each pixel in the registered high frequency face information as a center point, and form a center point with its neighboring pixels to form n*n. a block (for example, 3*3), and the grayscale value of the center point is used as a comparison base value. If the grayscale value of the neighborhood pixel in the block is greater than or equal to the comparison base value, then the block is in the block. The neighborhood pixel is encoded as "1", otherwise the code is "0". Thereafter, each neighboring pixel in the block is given a different weight value (for example, one of 1, 2, 4, 8, ... or 128) and the weight values are not repeated, and the total area is added. The value of the code within the block multiplied by the weight value to obtain the value representing this center point. After that, the values are combined to obtain the registered high frequency face code information corresponding to the registered high frequency face information.

續言之,由於註冊高頻臉資訊係對應於註冊低頻臉資訊,因此可得到對應於註冊低頻臉資訊之註冊高頻臉編碼資訊。然而發明人要強調的是,以區域二元化圖型(LBP)技術編碼註冊高頻臉資訊僅為舉例性說明,並未限縮本發明之保護範圍。例如,本案亦可使用區域三元化圖型(LTP)技術以有效降維並保有抗噪能力。而於步驟840後,處理元件330將註冊高頻臉編碼資訊傳送至資 料庫200中並將此註冊高頻臉編碼資訊儲存於資料庫200中(步驟850)。其中,由於處理元件330於步驟840中係對註冊高頻臉資訊進行不可逆之編碼,因此當資料庫200被竊取資料時,其亦無法損失完整之個人身分資料,故本發明確實可避免個人隱私權遭到侵犯之問題以及提供更安全的門禁安全機制。也就是說,本發明係將註冊低頻臉資訊及註冊高頻臉編碼資訊各別儲存於射頻辨識標籤100及資料庫200中,藉以使得使用者不會損失完整之個人身分資料,因此可避免個人隱私權遭到侵犯之問題以及提供更安全的門禁安全機制。 In other words, since the registered high frequency face information corresponds to the registered low frequency face information, the registered high frequency face coded information corresponding to the registered low frequency face information can be obtained. However, the inventor would like to emphasize that the registration of high frequency face information by the regional binary pattern (LBP) technique is merely an illustrative example and does not limit the scope of protection of the present invention. For example, the case can also use regional ternary pattern (LTP) technology to effectively reduce dimensionality and maintain anti-noise ability. After step 840, the processing component 330 transmits the registered high frequency face code information to the capital. The registered high frequency face coded information is stored in the repository 200 in the repository 200 (step 850). Wherein, since the processing component 330 performs irreversible encoding on the registered high frequency face information in step 840, when the data repository 200 is stolen, the personal identity data cannot be lost, so the invention can avoid personal privacy. The issue of violation of rights and the provision of safer access control security mechanisms. In other words, the present invention stores the registered low frequency face information and the registered high frequency face code information in the RFID tag 100 and the database 200, so that the user does not lose the complete personal identity data, thereby avoiding the individual. The issue of privacy violations and the provision of safer access control security mechanisms.

而於註冊階段完成後,當一待辨識者攜帶射頻辨識標籤100而欲通過本發明之射頻辨識門禁系統時,則門禁辨識裝置300可先利用射頻辨識讀取器310驗證此待辨識者所攜帶之此射頻辨識標籤100是否合法(步驟910),若“否”,則門禁辨識裝置300判定此射頻辨識標籤100不合法(步驟911),因此將不進行後續之驗證動作,更甚者,門禁辨識裝置300亦可具有警示元件(未繪示)以警告此待辨識者或警示門禁辨識裝置300之建構者或保全人員。而若門禁辨識裝置300判定此射頻辨識標籤100合法,則會利用影像擷取元件320擷取此待辨識者之人臉影像(步驟920),且當處理元件330接收影像擷取元件320所取得之此人臉影像後,可藉由小波轉換技術將此人臉影像分解成較低頻分量之待辨識低頻臉資訊及較高頻分量之待辨識高頻臉資訊(步驟930)。其中,處理元件330分解此人臉影像所使用之小波轉換技術大致相同於前述步驟820中之分解註冊人臉影像所使用之小波轉換技術,故於此將不贅述。 After the registration phase is completed, when an identifier to be transmitted carries the RFID tag 100 and the RFID access control system of the present invention is to be used, the access control device 300 can first use the RFID reader 310 to verify that the identifier is carried by the identifier. Whether the RFID tag 100 is legal (step 910), if "No", the access control device 300 determines that the RFID tag 100 is illegal (step 911), so no subsequent verification action will be performed, and even more, the access control The identification device 300 can also have a warning component (not shown) to warn the to-be-identified person or the constructor or security personnel of the access control identification device 300. If the access control device 300 determines that the RFID tag 100 is legal, the image capturing component 320 captures the human face image of the to-be-identified person (step 920), and when the processing component 330 receives the image capturing component 320. After the face image, the face image can be decomposed into lower frequency components of the lower frequency component and the higher frequency component to be recognized by the wavelet transform technology (step 930). The wavelet transform technology used by the processing component 330 to decompose the facial image is substantially the same as the wavelet transform technique used in the decomposed registered facial image in the foregoing step 820, and thus will not be described herein.

而於步驟930後,處理元件330可透過一編碼技術編碼此待辨識高頻臉資訊以產生一待辨識高頻臉編碼資訊(步驟950)。其中,處理元件330於步驟950中對待辨識高頻臉資訊所進行之編碼動作大致相同於前述步驟840中對註冊 高頻臉資訊所進行之編碼動作,故於此將不贅述。並且,於步驟940中所使用之編碼技術應相同於前述步驟840中所使用之編碼技術。亦即,若註冊高頻臉編碼資訊例如為區域二元化圖型特徵直方圖頻譜,則待辨識高頻臉編碼資訊亦為區域二元化圖型特徵直方圖頻譜;若註冊高頻臉編碼資訊例如為區域三元化圖型特徵直方圖頻譜,則待辨識高頻臉編碼資訊亦為區域三元化圖型特徵直方圖頻譜。 After step 930, the processing component 330 can encode the to-be-identified high-frequency face information by an encoding technique to generate a high-frequency face encoding information to be recognized (step 950). The encoding operation performed by the processing component 330 in step 950 to identify the high-frequency face information is substantially the same as the registration in the foregoing step 840. The encoding operation performed by the high frequency face information will not be described here. Also, the encoding technique used in step 940 should be identical to the encoding technique used in step 840 above. That is, if the registered high frequency face coding information is, for example, a region binary pattern type histogram spectrum, the high frequency face coding information to be identified is also a region binary pattern type histogram spectrum; if the high frequency face coding is registered The information is, for example, a histogram spectrum of the ternary pattern of the region, and the high-frequency face coding information to be identified is also a histogram spectrum of the ternary pattern of the region.

另外,當處理元件330得到待辨識低頻臉資訊後,可透過射頻辨識讀取器310而從通過步驟910驗證後之射頻辨識標籤100處,取得射頻辨識標籤100內所儲存之註冊低頻臉資訊(步驟940)。之後,處理元件330比對此待辨識低頻臉資訊是否符合射頻辨識標籤100內所儲存之註冊低頻臉資訊(步驟960),若“否”,則處理元件330判定此待辨識者不合格(步驟991),因此將不進行後續之驗證動作,更甚者,門禁辨識裝置300亦可警告此待辨識者或警示門禁辨識裝置300之建構者或保全人員。 In addition, after the processing component 330 obtains the low frequency face information to be recognized, the registered low frequency face information stored in the RFID tag 100 can be obtained from the RFID tag 100 verified by the step 910 through the RFID reader 310. Step 940). Thereafter, the processing component 330 determines whether the low frequency face information to be recognized meets the registered low frequency face information stored in the RFID tag 100 (step 960). If "No", the processing component 330 determines that the unidentified person is unqualified (step 991), therefore, no subsequent verification actions will be performed. Moreover, the access control device 300 can also alert the constructor or security personnel of the to-be-identified person or the warning access control device 300.

其中於步驟960中,處理元件330可使用最近鄰(nearest neighbor,NN)演算法進行比對之動作。舉例來說,處理元件330可先定義出低頻臉資訊中之比對點,並計算待辨識低頻臉資訊中之比對點與註冊低頻臉資訊中對應之比對點之間之尺度空間距離,再判斷此計算得到之尺度空間距離是否小於一設定之低頻階段門檻值,若“是”,則判定此待辨識低頻臉資訊符合射頻辨識標籤100內所儲存之註冊低頻臉資訊。 In step 960, processing component 330 can perform a comparison action using a nearest neighbor (NN) algorithm. For example, the processing component 330 may first define a comparison point in the low frequency face information, and calculate a scale space distance between the comparison point in the low frequency face information to be recognized and the corresponding comparison point in the registered low frequency face information. It is further determined whether the calculated scale space distance is less than a set low frequency stage threshold. If “Yes”, it is determined that the to-be-identified low frequency face information conforms to the registered low frequency face information stored in the RFID tag 100.

之後,若處理元件330判定待辨識低頻臉資訊符合射頻辨識標籤100內所儲存之註冊低頻臉資訊,則處理元件330取出對應於此註冊低頻臉資訊之註冊高頻臉編碼資訊(步驟970)。其中,處理元件330可例如利用射頻辨識標籤 100之序號或註冊低頻臉資訊之人臉編號,要求資料庫200傳送對應於此序號或人臉編號之註冊高頻臉編碼資訊至處理元件330。亦或者,門禁辨識裝置300可更包含近端儲存元件340,其電性連接處理元件330且可儲存有例如為常出入之註冊人之註冊高頻臉編碼資訊,使得處理元件330於步驟970中可搜索對應於此註冊低頻臉資訊之註冊高頻臉編碼資訊是否儲存於近端儲存元件340中,若“是”,則處理元件330從近端儲存元件340中取出此對應之註冊高頻臉編碼資訊,若“否”,則處理元件330再從資料庫200中取出對應於此註冊低頻臉資訊之註冊高頻臉編碼資訊。藉此,本發明可進一步減少資料傳送時之處理時間。其中,當資料庫200中之資訊進行更新時,處理元件330可取得更新後之資訊,並將此更新後之資訊傳送至近端儲存元件340中並取代被更新之資訊,藉以同步更新近端儲存元件340中之資訊。 Thereafter, if the processing component 330 determines that the low frequency face information to be recognized conforms to the registered low frequency face information stored in the RFID tag 100, the processing component 330 retrieves the registered high frequency face code information corresponding to the registered low frequency face information (step 970). Wherein, the processing component 330 can utilize, for example, a radio frequency identification tag The serial number of 100 or the face number of the registered low frequency face information requires the database 200 to transmit the registered high frequency face code information corresponding to the serial number or the face number to the processing component 330. Alternatively, the access control device 300 can further include a near-end storage component 340 electrically coupled to the processing component 330 and can store registered high-frequency face encoding information, such as a frequent registrant, such that the processing component 330 is in step 970. The registered high frequency face code information corresponding to the registered low frequency face information can be searched for stored in the near end storage component 340. If YES, the processing component 330 takes the corresponding registered high frequency face from the near end storage component 340. Encoding information, if "No", the processing component 330 then retrieves the registered high frequency face coding information corresponding to the registered low frequency face information from the database 200. Thereby, the present invention can further reduce the processing time when data is transmitted. Wherein, when the information in the database 200 is updated, the processing component 330 can obtain the updated information, and transmit the updated information to the near-end storage component 340 and replace the updated information, thereby synchronously updating the near-end. The information in element 340 is stored.

而當處理元件330取得對應於此註冊低頻臉資訊之註冊高頻臉編碼資訊後,會比對從步驟950中所產生之待辨識高頻臉編碼資訊是否符合於步驟970中所取出之此註冊高頻臉編碼資訊(步驟980),若“否”,則處理元件330判定此待辨識者不合格(步驟991);若“是”,則處理元件330判定此待辨識者合格(步驟992),以使得此待辨識者可進行出入之行為。其中於步驟980之比對中,處理元件330可使用最近鄰(NN)演算法進行比對之動作。舉例來說,處理元件330可計算待辨識高頻臉編碼資訊與註冊高頻臉編碼資訊之間之一距離值,並判斷此距離值是否小於設定之高頻階段門檻值,若“是”,則處理元件330即判定此待辨識高頻臉編碼資訊符合取出之此註冊高頻臉編碼資訊;若“否”,則處理元件330判定此待辨識高頻臉編碼資訊不符合取出之此註冊高頻臉編碼資訊。 When the processing component 330 obtains the registered high frequency face code information corresponding to the registered low frequency face information, it compares whether the high frequency face coded information to be recognized generated in step 950 meets the registration taken in step 970. The high frequency face coding information (step 980), if "no", the processing component 330 determines that the person to be identified is unsatisfactory (step 991); if "yes", the processing component 330 determines that the person to be identified is qualified (step 992) In order for the person to be identified to be able to enter and exit. In the alignment of step 980, processing component 330 can perform the alignment action using the nearest neighbor (NN) algorithm. For example, the processing component 330 can calculate a distance value between the high frequency face coded information to be recognized and the registered high frequency face coded information, and determine whether the distance value is less than a set high frequency stage threshold value, and if yes, Then, the processing component 330 determines that the high frequency face coded information to be recognized conforms to the registered high frequency face coded information; if no, the processing component 330 determines that the high frequency face coded information to be recognized does not meet the registered high value. Frequency face coding information.

也就是說,於步驟980中,處理元件330僅需將於步驟960比對過之此註冊低頻臉資訊所對應之註冊高頻臉編碼資訊,以及於步驟950中所取得之待辨識高頻臉編碼資訊二者之間進行比對,而無需比對資料庫200中之其他註冊高頻臉編碼資訊。因此,本發明可具有較快之辨識速率。 That is, in step 980, the processing component 330 only needs to register the registered high frequency face code information corresponding to the registered low frequency face information in step 960, and the high frequency face to be recognized obtained in step 950. The encoded information is compared between the two without the need to compare the other registered high frequency face coded information in the database 200. Therefore, the present invention can have a faster recognition rate.

以上所述僅為舉例性,而非為限制性者。任何未脫離本發明之精神與範疇,而對其進行之等效修改或變更,均應包含於後附之申請專利範圍中。 The above is intended to be illustrative only and not limiting. Any equivalent modifications or alterations to the spirit and scope of the invention are intended to be included in the scope of the appended claims.

910-992‧‧‧步驟 910-992‧‧‧Steps

Claims (10)

一種結合基於秘密分持臉部辨識的射頻辨識門禁系統,至少包含:一射頻辨識標籤,儲存有至少一註冊低頻臉資訊,其中該註冊低頻臉資訊係對一註冊人臉影像進行一小波轉換後所成;一資料庫,儲存有複數個註冊高頻臉編碼資訊,其中該註冊高頻臉資訊係對相同之該註冊人臉影像進行同一之該小波轉換後所成,其中該註冊低頻臉資訊不同於該註冊高頻臉資訊,且該註冊低頻臉資訊及該註冊高頻臉資訊各別儲存於該射頻辨識標籤及該資料庫中;以及一門禁辨識裝置,電性連接該資料庫,該門禁辨識裝置至少包含:一射頻辨識讀取器,電性通訊該射頻辨識標籤,該射頻辨識讀取器驗證該射頻辨識標籤是否合法;一影像擷取元件,擷取一待辨識者之一人臉影像;以及一處理元件,電性連接該射頻辨識讀取器及該影像擷取元件,當該射頻辨識讀取器驗證該射頻辨識標籤合法後,該處理元件藉由一小波轉換技術將該人臉影像分解成較低頻分量之一待辨識低頻臉資訊及較高頻分量之一待辨識高頻臉資訊,且透過一編碼技術編碼該待辨識高頻臉資訊以產生一待辨識高頻臉編碼資訊,其中該處理元件透過該射頻辨識讀取器從通過驗證之該射頻辨識標籤處取得該註冊低頻臉資訊,並比對從該射頻辨識標籤處所取得之該待辨識低頻臉資訊是否符合該註冊低頻臉資訊,若“否”,則該 處理元件判定該待辨識者不合格,若“是”,則該處理元件從該資料庫中取出對應於該註冊低頻臉資訊之一註冊高頻臉編碼資訊,並比對從該資料庫中所取得之該待辨識高頻臉編碼資訊是否符合取出之該註冊高頻臉編碼資訊,若“否”,則該處理元件判定該待辨識者不合格,若“是”,則該處理元件判定該待辨識者合格。 A radio frequency identification access control system combining secret recognition based face recognition includes at least one radio frequency identification tag storing at least one registered low frequency face information, wherein the registered low frequency face information is subjected to a wavelet transform on a registered face image a database storing a plurality of registered high frequency face code information, wherein the registered high frequency face information is formed by performing the same wavelet conversion on the same registered face image, wherein the registered low frequency face information Different from the registered high frequency face information, and the registered low frequency face information and the registered high frequency face information are respectively stored in the RFID tag and the database; and an access control device electrically connected to the database, The access control device includes at least: a radio frequency identification reader electrically communicating the radio frequency identification tag, the radio frequency identification reader verifying whether the radio frequency identification tag is legal; and an image capturing component capturing a face of a person to be recognized And a processing component electrically connected to the RFID reader and the image capturing component, when the RFID reader verifies After the RFID tag is legal, the processing component decomposes the face image into one of the lower frequency components and one of the higher frequency components to be recognized by the wavelet transform technology, and the high frequency face information to be recognized is transmitted through a wavelet component. The encoding technique encodes the high-frequency face information to be recognized to generate a high-frequency face-coded information to be recognized, wherein the processing component obtains the registered low-frequency face information from the verified RFID tag through the RFID reader, and compares Whether the low frequency face information to be recognized obtained from the RFID tag conforms to the registered low frequency face information, if "No", then The processing component determines that the unidentified person is unqualified. If YES, the processing component extracts, from the database, the registered high frequency face coding information corresponding to the registered low frequency face information, and compares the information from the database. Whether the obtained high-frequency face coded information is consistent with the registered high-frequency face coded information obtained, and if “No”, the processing component determines that the person to be identified is unqualified, and if “Yes”, the processing component determines that Those who are to be identified are qualified. 如申請專利範圍第1項所述之結合基於秘密分持臉部辨識的射頻辨識門禁系統,其中該門禁辨識裝置更包含一近端儲存元件電性連接該處理元件,當該處理元件判定該待辨識低頻臉資訊符合該註冊低頻臉資訊後,先搜索對應於該註冊低頻臉資訊之該註冊高頻臉編碼資訊是否儲存於該近端儲存元件中,若“是”,則該處理元件從該近端儲存元件中取出該註冊高頻臉編碼資訊,若“否”,則該處理元件再從該資料庫中取出對應於該註冊低頻臉資訊之該註冊高頻臉編碼資訊。 The radio frequency identification access control system according to the first aspect of the invention, wherein the access control device further comprises a near-end storage component electrically connected to the processing component, and the processing component determines that the After identifying the low frequency face information and the registered low frequency face information, first searching whether the registered high frequency face code information corresponding to the registered low frequency face information is stored in the near end storage element, and if yes, the processing component is from the The registered high frequency face code information is taken out from the near end storage component. If no, the processing component extracts the registered high frequency face code information corresponding to the registered low frequency face information from the database. 如申請專利範圍第1項所述之結合基於秘密分持臉部辨識的射頻辨識門禁系統,其中該註冊高頻臉編碼資訊及該待辨識高頻臉編碼資訊係各別為一區域二元化圖型特徵直方圖頻譜或各別為一區域三元化圖型特徵直方圖頻譜。 The radio frequency identification access control system based on secret facial recognition is described in claim 1, wherein the registered high frequency face coding information and the to-be-identified high frequency face coding information system are each a region dualization. The pattern feature histogram spectrum or each is a region ternary pattern feature histogram spectrum. 一種結合基於秘密分持臉部辨識的射頻辨識門禁方法,至少包含下列步驟:利用一射頻辨識讀取器驗證一射頻辨識標籤是否合法,其中該射頻辨識標籤儲存有至少一註冊低頻臉資訊,其中該註冊低頻臉資訊係對一註冊人臉影像進行一小波轉換後所成;利用一影像擷取元件擷取一待辨識者之一人臉影像; 當判定該射頻辨識標籤合法後,利用一處理元件接收該人臉影像,並藉由一小波轉換技術將該人臉影像分解成較低頻分量之一待辨識低頻臉資訊及較高頻分量之一待辨識高頻臉資訊;利用該處理元件透過一編碼技術編碼該待辨識高頻臉資訊以產生一待辨識高頻臉編碼資訊;利用該處理元件透過該射頻辨識讀取器從通過驗證之該射頻辨識標籤處取得該註冊低頻臉資訊;利用該處理元件比對從該射頻辨識標籤處所取得之該待辨識低頻臉資訊是否符合該註冊低頻臉資訊,若“否”,則該處理元件判定該待辨識者不合格;若該處理元件判定該待辨識低頻臉資訊符合該註冊低頻臉資訊,則利用該處理元件從一資料庫中取出對應於該註冊低頻臉資訊之一註冊高頻臉編碼資訊,其中該註冊高頻臉資訊係對相同之該註冊人臉影像進行同一之該小波轉換後所成,其中該註冊低頻臉資訊不同於該註冊高頻臉資訊,且該註冊低頻臉資訊及該註冊高頻臉資訊各別儲存於該射頻辨識標籤及該資料庫中;以及利用該處理元件比對該待辨識高頻臉編碼資訊是否符合從該資料庫中所取得之該註冊高頻臉編碼資訊,若“否”,則該處理元件判定該待辨識者不合格,若“是”,則該處理元件判定該待辨識者合格。 A radio frequency identification access control method based on secret separation facial recognition includes at least the following steps: verifying whether a radio frequency identification tag is legal by using a radio frequency identification tag, wherein the radio frequency identification tag stores at least one registered low frequency face information, wherein The registered low frequency face information is formed by performing a wavelet transform on a registered face image; and using an image capturing component to capture a face image of one of the persons to be recognized; After determining that the RFID tag is legal, the face image is received by a processing component, and the face image is decomposed into one of the lower frequency components to be recognized by the wavelet transform technology, and the low frequency face information and the higher frequency component are to be recognized. A high-frequency face information is to be recognized; the processing component is used to encode the high-frequency face information to be recognized by an encoding technique to generate a high-frequency face coded information to be recognized; and the processing component is used to pass the verification through the RFID reader. Obtaining the registered low frequency face information at the RFID tag; using the processing component to compare whether the low frequency face information to be recognized obtained from the RFID tag meets the registered low frequency face information, and if “No”, the processing component determines If the processing component determines that the low frequency face information to be recognized meets the registered low frequency face information, the processing component uses the processing component to retrieve a high frequency face code corresponding to the registered low frequency face information from a database. Information, wherein the registered high frequency face information is formed by performing the same wavelet transformation on the same registered face image, wherein The registered low frequency face information is different from the registered high frequency face information, and the registered low frequency face information and the registered high frequency face information are separately stored in the radio frequency identification tag and the database; and the processing component is used to be identified Whether the high frequency face coded information meets the registered high frequency face coded information obtained from the database, if "No", the processing component determines that the person to be identified is unqualified, and if "yes", the processing component determines The person to be identified is qualified. 如申請專利範圍第4項所述之結合基於秘密分持臉部辨識的射頻辨識門禁方法,其中於判定該待辨識低頻臉資訊符合該註冊 低頻臉資訊後,先搜索對應於該註冊低頻臉資訊之該註冊高頻臉編碼資訊是否儲存於一近端儲存元件中,若“是”,則該處理元件從該近端儲存元件中取出該註冊高頻臉編碼資訊,若“否”,則該處理元件再從該資料庫中取出對應於該註冊低頻臉資訊之該註冊高頻臉編碼資訊。 The radio frequency identification access control method based on secret separation facial recognition, as described in claim 4, wherein determining that the low frequency facial information to be recognized conforms to the registration After the low frequency face information, first searching whether the registered high frequency face code information corresponding to the registered low frequency face information is stored in a near end storage element, and if YES, the processing component is taken out from the near end storage element Registering the high frequency face code information, if "No", the processing component then retrieves the registered high frequency face code information corresponding to the registered low frequency face information from the database. 如申請專利範圍第4項所述之結合基於秘密分持臉部辨識的射頻辨識門禁方法,更包含下列步驟:利用該處理元件藉由該小波轉換技術將一註冊人臉影像分解成較低頻分量之該註冊低頻臉資訊及較高頻分量之一註冊高頻臉資訊;利用該處理元件透過該編碼技術編碼該註冊高頻臉資訊以產生該註冊高頻臉編碼資訊,其中該註冊高頻臉編碼資訊對應於該註冊低頻臉資訊;儲存該註冊低頻臉資訊於該射頻辨識標籤中;以及儲存該註冊高頻臉編碼資訊於該資料庫中。 The radio frequency identification access control method based on secret separation facial recognition according to claim 4 of the patent application scope further includes the following steps: using the processing component to decompose a registered facial image into a lower frequency by using the processing component; Registering the high frequency face information with one of the registered low frequency face information and the higher frequency component; using the processing component to encode the registered high frequency face information by the encoding technique to generate the registered high frequency face coded information, wherein the registered high frequency face information The face code information corresponds to the registered low frequency face information; the registered low frequency face information is stored in the RFID tag; and the registered high frequency face code information is stored in the database. 如申請專利範圍第4項所述之結合基於秘密分持臉部辨識的射頻辨識門禁方法,其中該註冊高頻臉編碼資訊及該待辨識高頻臉編碼資訊係各別為一區域二元化圖型特徵直方圖頻譜或各別為一區域三元化圖型特徵直方圖頻譜。 The radio frequency identification access control method based on the secret separation facial recognition method described in claim 4, wherein the registered high frequency face coding information and the to-be-identified high frequency face coding information system are each a region dualization The pattern feature histogram spectrum or each is a region ternary pattern feature histogram spectrum. 如申請專利範圍第4項所述之結合基於秘密分持臉部辨識的射頻辨識門禁方法,其中於比對該待辨識低頻臉資訊是否符合該註冊低頻臉資訊之步驟中,係使用一最近鄰演算法計算該待辨識低頻臉資訊與該註冊低頻臉資訊之間之一尺度空間距離,並判斷該尺度空間距離是否小於一低頻階段門檻值,若“是”,則 判定該待辨識低頻臉資訊符合該註冊低頻臉資訊。 The radio frequency identification access control method based on the secret separation facial recognition method according to claim 4, wherein in the step of comparing whether the low frequency face information to be recognized meets the registered low frequency face information, a nearest neighbor is used. The algorithm calculates a scale space distance between the low frequency face information to be recognized and the registered low frequency face information, and determines whether the scale space distance is less than a low frequency stage threshold value, and if yes, It is determined that the low frequency face information to be recognized meets the registered low frequency face information. 如申請專利範圍第4項所述之結合基於秘密分持臉部辨識的射頻辨識門禁方法,其中於比對該待辨識高頻臉編碼資訊是否符合取出之該註冊高頻臉編碼資訊之步驟中,係使用一最近鄰演算法計算該待辨識高頻臉編碼資訊與該註冊高頻臉編碼資訊之間之一距離值,並判斷該距離值是否小於一高頻階段門檻值,若“是”,則判定該待辨識高頻臉編碼資訊符合取出之該註冊高頻臉編碼資訊。 The method of combining the secret recognition facial recognition based radio frequency identification access method according to the fourth aspect of the patent application, wherein the step of comparing the high frequency face coded information to be recognized with the registered high frequency face coding information is taken out. Using a nearest neighbor algorithm to calculate a distance value between the to-be-identified high-frequency face coded information and the registered high-frequency face coded information, and determine whether the distance value is less than a high-frequency threshold value, if "yes" And determining that the high frequency face coded information to be identified conforms to the registered high frequency face coded information that is taken out. 一種用於結合基於秘密分持臉部辨識的射頻辨識門禁系統中之一射頻辨識標籤,至少包含:一標籤元件,利用一射頻辨識技術與一門禁辨識裝置進行通訊,使得該門禁辨識裝置驗證該射頻辨識標籤是否合法;以及一儲存元件,電性連接該標籤元件並儲存有至少一註冊低頻臉資訊,其中該註冊低頻臉資訊係對一註冊人臉影像進行一小波轉換後所成,且該註冊低頻臉資訊不同於一資料庫中所儲存之一註冊高頻臉資訊,當該門禁辨識裝置驗證該射頻辨識標籤合法後,該儲存元件將所儲存之該註冊低頻臉資訊傳送至該門禁辨識裝置,以使得該門禁辨識裝置比對從該儲存元件處所取得之該註冊低頻臉資訊與使用該射頻辨識標籤之一待辨識者之一待辨識低頻臉資訊是否相符。 An RFID tag for combining an RFID-based access control system based on secret separation facial recognition includes at least one tag component, and communicates with an access control device by using a radio frequency identification technology, so that the access control device verifies the Whether the RFID tag is legal; and a storage component electrically connected to the tag component and storing at least one registered low frequency face information, wherein the registered low frequency face information is formed by performing a wavelet transform on a registered face image, and the The registered low frequency face information is different from one of the registered high frequency face information stored in a database. When the access control device verifies that the RFID tag is legal, the storage component transmits the stored low frequency face information to the access control identification. And a device for causing the access control device to compare whether the registered low frequency face information obtained from the storage component matches the low frequency face information to be recognized by one of the identifiers to be identified.
TW104101867A 2015-01-20 2015-01-20 Rfid acess control method and system combining face recognition based on secret splitting and rfid tag used therein TWI544358B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
TW104101867A TWI544358B (en) 2015-01-20 2015-01-20 Rfid acess control method and system combining face recognition based on secret splitting and rfid tag used therein

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
TW104101867A TWI544358B (en) 2015-01-20 2015-01-20 Rfid acess control method and system combining face recognition based on secret splitting and rfid tag used therein

Publications (2)

Publication Number Publication Date
TWI544358B true TWI544358B (en) 2016-08-01
TW201627901A TW201627901A (en) 2016-08-01

Family

ID=57181804

Family Applications (1)

Application Number Title Priority Date Filing Date
TW104101867A TWI544358B (en) 2015-01-20 2015-01-20 Rfid acess control method and system combining face recognition based on secret splitting and rfid tag used therein

Country Status (1)

Country Link
TW (1) TWI544358B (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
TWI682329B (en) * 2018-12-14 2020-01-11 技嘉科技股份有限公司 Method, device and computer readable medium of facial recognition

Also Published As

Publication number Publication date
TW201627901A (en) 2016-08-01

Similar Documents

Publication Publication Date Title
Zhang et al. Robust palmprint verification using 2D and 3D features
CN108875341A (en) A kind of face unlocking method, device, system and computer storage medium
TW201528030A (en) System and method for biometric authentication in connection with camera-equipped devices
Rathgeb et al. Bit reliability-driven template matching in iris recognition
Elnasir et al. Proposed scheme for palm vein recognition based on linear discrimination analysis and nearest neighbour classifier
KR101774151B1 (en) Method and apparatus of enhancing iris recognition security using distributed iris template storage and matching
Li et al. Face liveness detection and recognition using shearlet based feature descriptors
Yadav et al. Fingerprint recognition based on minutiae information
Naveen et al. Face recognition and authentication using LBP and BSIF mask detection and elimination
KR20150112635A (en) Doorlock system
Wu et al. Comparative analysis and application of LBP face image recognition algorithms
Samai et al. 2d and 3d palmprint recognition using deep learning method
Dhavale DWT and DCT based robust iris feature extraction and recognition algorithm for biometric personal identification
Agarwal et al. An efficient back propagation neural network based face recognition system using haar wavelet transform and PCA
TWI544358B (en) Rfid acess control method and system combining face recognition based on secret splitting and rfid tag used therein
Sharma et al. A novel image compression based method for multispectral fingerprint biometric system
Oloyede et al. Evaluating the effect of occlusion in face recognition systems
Esan et al. Bimodal biometrics for financial infrastructure security
Chen An efficient palmprint recognition method based on block dominat orientation code
Javadtalab et al. Transparent non-intrusive multimodal biometric system for video conference using the fusion of face and ear recognition
Zuo et al. Adaptive quality-based performance prediction and boosting for iris authentication: methodology and its illustration
Vyas et al. Efficient features for smartphone-based iris recognition
Aggithaya et al. A multimodal biometric authentication system based on 2D and 3D palmprint features
Herbadji et al. Multimodal Biometric Verification using the Iris and Major Finger Knuckles
Abdelouahed et al. A comparative study of anti-spoofing detection based on deep learning