,M317045 ，.八、新型說明： 【新型所屬之技術領域】 本創作係關於行動運算裝置，尤其是有關於利用使用者臉部 的特徵辨識來控制存取的行動運算裝置。 【先前技術】 本說明書所稱「行動運算裝置」係指獨立具有運算功能、且 具有高度可移動性（mobility)的運算裝置。目前最常見的行動運 • 算裝置包括（但不限於）筆記型電腦、手機、個人數位助理（persc)nalM317045, .8, New Description: [Technical Fields of New Types] This creation relates to mobile computing devices, and more particularly to mobile computing devices that utilize feature recognition of a user's face to control access. [Prior Art] The term "motion computing device" as used in this specification refers to an arithmetic device that has a computing function independently and has high mobility. Currently the most common mobile computing devices include (but are not limited to) notebook computers, cell phones, personal digital assistants (persc) nal
Digital Assistant，PDA)等。 為了避免未授權（unauthorized)或未經驗證（unauthenticated) 的第三人使用或存取這些行動運算裝置所提供的功能或其所铸存 的資訊，這些行動運算裝置一般都可以設定成在開機時需要使用 者透過行動運算裝置所具備的人機界面（例如筆記型電腦的鍵盤 和螢幕、PDA的觸控式螢幕與光筆）輸入一個正確的密碼，才會 # 正常的開機供使用者只用；或者，這些行動運算裝置如果具有節 能功能的話，也會在進入睡眠節能模式後被「喚醒」（wake up) 時，要求輸正確的密碼後才會恢復正常的模式。這種要求確認使 用者身份才開放使用的功能，本說明書統稱為「存取控制」（ access control) 〇 除了前述以密碼控制存取外，隨著科技進步，各種新穎的存 取控制方式也愈來愈多。比如說，目前市場上利用晶片卡、指紋 5 ，M317045 、 辨識等方式來作存取控制㈣記型電腦已經相當常見。另外在控 •制特定人進出特定區域的安全監㈣技術領域裡，也有非常豐富 與類似的門料制減。除了常見_卡、感應卡外，利用生物 特徵的指紋、掌紋、瞳孔等辨識也並非罕見。還有—種已經證實 相當有實效的控制方式是則影像擷取裝置（例如攝影鏡頭）抓 取人臉的影像加以辨識後，作為是否放行的依據。 對照之下’目前多數行動運算裝置雖已逐漸將ccd或cm〇s 籲 攝影鏡頭作為必備的配件，但是尚未有行動運算裝置應用這些内 建（built-in)的攝影鏡頭來進行存取控制。 【新型内容】 一因此’本創作提出-種賴人__轉取㈣的行動運 算裝置（、下簡稱本裝置）。本裝置除具備—般行動運算裝置之軟 硬體結構與人機界面外，其主要特徵在於至少額外包含一攝影鏡 頭、與-辨識模組。這二者均内建於本裝置内其中辨識模組係 以韌體或軟體至少二者之一的形式安裝於本裝置内。 當本裝置需要崎存輪㈣，該觸模組於本裝置之人機 界面的螢幕上動態呈現攝影鏡頭所抓取的影像以及—參考圖形。 該參考圖形供❹者·自己相對於攝影鏡頭的位置、遠近、以 及角度等致使自己的臉部影像依據該參考圖形呈現於發幕上的 特定範圍内。然後使用者透過人機界面觸發辨識模組的辨識功 能、且在辨識模組完成和事先建立好的—或多筆特徵資料比對 6 1 M317045 ' 後，本震置依據比對的結果而開玫、或是拒絕使用者使用本裝置。 - @於本創作之優點與精神可以藉由以下的新型詳述及所附圖 式得到進一步的瞭解。 【實施方式】 本裝置與一般的行動冑算裝置，I是筆記型電腦 、平板電腦 (tablet PC)、手機、PDA等具備相同之運算軟硬體結構，例如包 鲁 有CPU、RAM、非揮發性的記憶體（Flash memory以及/或是 硬式磁碟等）、匯流排、輸出入控制器等。本裝置也與—般的行動 運算裝置-樣具備相同人機界面，例如内建的LCD顯示螢幕、觸 控式螢幕（以下統稱為螢幕）、鍵盤、滑鼠、光筆等。為了簡化起 見，以下本說明書是以一筆記型電腦為例說明本創作的實施例。 如第1圖所示，本裝置10的特徵在於至少額外包含一攝影鏡 碩20、與一辨識模組（未圖示）。攝影鏡頭2〇是内建於本裝置w 馨 的攝影鏡頭，而且和螢幕3〇是朝向同一方向，此外本創作對於攝 衫鏡頭2〇設置的位置、所採用技術（例如CCD或CM〇s)、解析 度（像素數目）等並不設限。 辨識模組係以韌體、軟體、或二者兼具的形式安裝於本裝置 10的非揮發性記憶體内’例如辨識模組可以是本裝χ 1〇的BI〇s 的-部份、或者是安裝於本裝置10的硬碟上而於本裝置1〇開機 過程中載入執行。此外，辨識模組還包含至少一筆特徵資料。 在本裝置10需要進行存取控制的適當時機，例如開機時、或 7 ，M317045 ，是在進人睡眠節能模式後被「倾」時、或是其他時候，本裝置 -1G會自_發辨識模組進人-認證模式。在此模式下，辨識模組 在螢幕3〇上動態呈現攝影鏡頭2〇所抓取的影像、以及一參考圖 形40。在本實施例中，參考圖形4〇是一個矩形框，使用者藉由 調整其相對於攝影鏡頭20的位置、遠近、以及角度等，使自^的 臉部影像呈現在參考圖形40限定的範圍内。接下來，使用者透過 域界面指㈣識模組進行比對。粒意到，視紐資料的特性， 辨識模組可以將整個特定範圍⑽影像和特徵㈣的影像作每一 像素（啦⑴的比對，或是先經過影像處理、絲特定範圍内重 要的人臉特徵後再和特徵資料比對，這裡的「特徵」有多種可能， 例如重要五官的位置及其相對位置、特別深逮或暗陳色塊的大 小、位置、範圍等。同樣地，本創作的標的並不在於所採用的比 對方式，許多相關的演算法已經有大量的揭露於文獻或先前技術 中’所以在此不作贅述。在辨識模組裡具有多筆特徵資料時，辨 • 識模組會逐一比對。如果找到一筆特徵資料符合使用者的人臉特 徵，辨識模組即會開放使用者使用本裝置1〇。反之，若是所有特 徵^料均不符合，則辨識模組即拒絕使用者使用本裝置1〇。 凊注思到在其他實施例中，其他適當的參考圖形也可以採 用。以第2a圖為例，參考圖形40是在螢幕30上所呈現的一組平 行的上、下水平標線。使用者藉由調整其相對於攝影鏡頭20的位 置、遠近、以及肖度等，使自己的臉部或頭顱的上下緣分別貼齊 上下標線’以將使用者臉部影像呈現在螢幕3〇 一定的範圍内。在 8 M317045 • 第2b圖所示的另一實施例，參考圖形40是在螢幕30上所呈現的 ‘ 一組平行的左、右垂直標線。或者，如第2(：、2(1圖所示的又一實 施例，參考圖形40是在螢幕30上所呈現的一組水平排列、適當 間隔的十字標線或圓形。請注意到，這些圖形定義了供使用者雙 眼注視用的參考點（十字或圓形的中心點），使用者使自己的雙眼 中心分別對準這些參考點，以將使用者臉部影像呈現在螢幕如 的一定範圍内。 ，藉由這個參考圖形40，各個使用者的臉部雖然五官、長寬、 胖痩不-，都可以㈣單的供賴模組在―致的條件（也就是 相對於攝影鏡頭20的距離、角度等）下進行臉部特徵的梅取與比 對，進而提高辨識的成功率。另外也請注意到，本裝置1〇的螢幕 30即為一光源，所以基本上擷取清晰人臉影像所需的照明並不成 問題，尤其參考圖形40更進一步供使用者調整其臉部相對於螢幕 30的距離、角度而使人臉可以在擷取特徵與比對時，受到亮度一 致的照明。如果螢幕30亮度有限時，可以於本裝置1〇的適當位 B 置（如攝影鏡頭20邊）設置照明元件21 (例如一個led)，由辨 識模組在線取人臉特徵影像時點亮，以提供額外的照明 除了前述的認證模式外，辨識模組還具備一種管理模式。在 辨識模組尚未進行過管理模式前，辨識模組並未包含任何特徵資 料（也因此不會也無法進行人臉特徵辨識的存取控制），而其理模 式裡的新增功能’即是在新增一筆可供未來辨識模式運行時比對 的一筆特徵資料。刪除功能則是透過本裝置1〇的人機界面刪^ 9 M317045 筆已經存在的特徵資料。管理模式還可以包含一個取代功能，取 代功能與新增功能類似，不過是取代已經存在的一筆特徵資料。 在本装置10正常完成開機後，使用者可以透過人機界面啟動 辨識模組的管理模式的新增、刪除、或取代魏。在進行新增功 月匕時辨識模組在螢幕30上動態呈現攝影鏡頭所抓取的影像、 、 考固$ 4〇。使用者猎由調整其相對於攝影鏡頭2〇的位 置遠近、以及角度等，使自己的臉部影像呈現在參考圖形4〇 所限定的螢幕3G的特定範_。接下來，使用者透過人機界面（例 如=下特定的按鍵組合）指示辨識模組處理與儲存該特定範圍内 々的影像作為—筆特徵資料。請注意到，辨識模組可以將整個特定 耗圍内的影像加㈣存、或是經過影像處理後⑽存重要的人臉 :徵以節省記憶空間與節省未來比對所需要的處理時間。也請注 思!^本創作的標的並不在於所採用的影像處理方式，許多相關 的决异法已_大量的揭露於讀或先前技術中，所以在此不作 贅述在辨識模組裡具有至少一筆特徵資料後，使用者即可以啟 動本裝置10的存取控制功能。 藉由以上較佳具體實_之詳述，料㈣更加清楚描述本 J Μ加以限制。相反地，其目的是希望能涵蓋各種改變 相等性的安排於本創作所欲切之專·__内。 ，M317045 i ’， 【圖式簡單說明】 . 第1圖係依據本創作一實施例之具有人臉特徵辨識功能的筆記型 電腦的示意圖。 第2a圖係依據本創作一實施例之參考圖形的示意圖。 第2b圖係依據本創作另一實施例之參考圖形的示意圖。 第2c圖係依據本創作又一實施例之參考圖形的示意圖。 第2d圖係依據本創作再一實施例之參考圖形的示意圖。 【主要元件符號說明】 10 本裝置 20 攝影鏡頭 21 照明元件 30 螢幕 40 參考圖形 11Digital Assistant, PDA), etc. In order to prevent unauthorized or unauthenticated third parties from using or accessing the functions provided by these mobile computing devices or the information they cast, these mobile computing devices can generally be set to be turned on. The user needs to input a correct password through the human-machine interface (such as the keyboard and screen of the notebook computer, the touch screen and the light pen of the PDA) of the mobile computing device, and then the normal booting is provided for the user to use only; Alternatively, if the mobile computing device has an energy saving function, it will be "wake up" after entering the sleep energy saving mode, and the normal mode will be restored after the correct password is requested. This kind of function is required to confirm the identity of the user and is open to use. This manual is collectively referred to as "access control". In addition to the aforementioned password-controlled access, with the advancement of technology, various novel access control methods are becoming more and more More and more. For example, it is quite common to use the chip card, fingerprint 5, M317045, identification, etc. for access control (4) on the market. In addition, in the field of safety supervision (4) technology for controlling specific people to enter and exit certain areas, there is also a very rich and similar door material reduction. In addition to common _ cards and proximity cards, it is not uncommon to use biometric fingerprints, palm prints, pupils, etc. There is also a control method that has proven to be quite effective. The image capturing device (such as a photographic lens) captures the image of the face and recognizes it as a basis for release. In contrast, most current mobile computing devices have gradually adopted ccd or cm〇s photographic lenses as a necessary accessory, but no mobile computing device has applied these built-in photographic lenses for access control. [New content] Therefore, the author of this creation proposes a mobile operation device (hereinafter referred to as the device). In addition to the soft hardware structure and human-machine interface of the mobile computing device, the device is characterized in that it additionally includes at least one photographic lens and an identification module. Both of these are built into the device, wherein the identification module is mounted in the device in the form of at least one of a firmware or a soft body. When the device requires a remnant wheel (4), the touch module dynamically displays the image captured by the photographic lens and the reference image on the screen of the human-machine interface of the device. The reference image of the viewer himself/herself is relative to the position, distance, angle, and the like of the photographic lens, so that his or her face image is presented in a specific range on the screen according to the reference pattern. Then, the user triggers the identification function of the identification module through the human-machine interface, and after the identification module completes and pre-establishes the - or more feature data comparison 6 1 M317045 ', the shock is opened according to the result of the comparison. Or refuse to use the device. - @ The advantages and spirit of this creation can be further understood by the following new details and drawings. [Embodiment] This device and the general action calculation device, I have the same computing hardware and software structure as a notebook computer, a tablet PC, a mobile phone, a PDA, etc., for example, a CPU, a RAM, and a non-volatile package. Sexual memory (Flash memory and / or hard disk, etc.), bus, input and output controller. The device also has the same human-machine interface as the general mobile computing device, such as a built-in LCD display screen, a touch-sensitive screen (hereinafter collectively referred to as a screen), a keyboard, a mouse, a light pen, and the like. For the sake of simplicity, the following description of the present invention is made by taking a notebook computer as an example. As shown in Fig. 1, the apparatus 10 is characterized by at least one additional photographic mirror 20 and an identification module (not shown). The photographic lens 2 is a photographic lens built into the device, and is oriented in the same direction as the screen 3 ,. In addition, the position of the photographic lens 2 〇 is set, and the technique used (for example, CCD or CM 〇 s). The resolution (number of pixels) and the like are not limited. The identification module is mounted in the non-volatile memory of the device 10 in the form of a firmware, a soft body, or both. For example, the identification module may be a part of the BI〇s of the device. Or it is installed on the hard disk of the device 10 and is loaded and executed during the booting process of the device. In addition, the identification module also contains at least one feature data. At the appropriate time when the device 10 needs to perform access control, for example, when the device is turned on, or 7 or M317045, when the device is "tilted" after entering the sleep energy-saving mode, or at other times, the device-1G will recognize itself. The module enters the person-authentication mode. In this mode, the recognition module dynamically displays the captured image of the photographic lens 2 on the screen 3, and a reference pattern 40. In the present embodiment, the reference figure 4A is a rectangular frame, and the user presents the face image from the ^ to the range defined by the reference figure 40 by adjusting the position, the distance, the angle, and the like with respect to the photographic lens 20. Inside. Next, the user compares the module through the domain interface (4). According to the characteristics of the data, the identification module can make the image of the entire specific range (10) image and the feature (4) for each pixel (ie (1) comparison, or the image processing, the important person within the specific range of the silk After the face features are compared with the feature data, there are many possibilities for the "features" here, such as the position of the important facial features and their relative positions, the size, position, and range of the special deep or dark color blocks. Similarly, this creation The target is not in the way of comparison. Many related algorithms have been extensively disclosed in the literature or prior art' so I will not go into details here. When there are multiple features in the identification module, the identification The module will be compared one by one. If a feature data is found to match the user's face feature, the recognition module will open the user to use the device. Otherwise, if all the features are not met, the identification module will be The user is denied access to the device. It is noted that in other embodiments, other suitable reference patterns may also be employed. Taking Figure 2a as an example, reference graphic 40 is in Firefly. A set of parallel upper and lower horizontal lines presented on the 30. The user aligns the upper and lower edges of his face or head by adjusting the position, distance, and diametry of the photographic lens 20, respectively. The upper and lower marking lines 'are to present the user's facial image within a certain range of the screen. In another embodiment shown in Fig. 2b, the reference figure 40 is presented on the screen 30. A set of parallel left and right vertical reticle. Or, as in the second (:, 2 (1), the reference figure 40 is a set of horizontally arranged, appropriately spaced crosses presented on the screen 30. Marking line or circle. Please note that these figures define the reference point (the center point of the cross or circle) for the user to look at with both eyes. The user points their center of the eye to these reference points respectively. The user's face image is displayed within a certain range of the screen. With this reference graphic 40, although the facial features of each user are five-in-one, long and wide, and fat, they are all capable of Condition (that is, relative to the photographic mirror) The distance, angle, etc. of 20) is used to compare and compare the facial features, thereby improving the success rate of recognition. Also, note that the screen 30 of the device is a light source, so it is basically clear. The illumination required for the face image is not a problem. In particular, the reference figure 40 is further provided for the user to adjust the distance and angle of the face relative to the screen 30 so that the face can be consistent in brightness when capturing features and comparisons. If the brightness of the screen 30 is limited, the lighting element 21 (for example, a led) may be disposed at the appropriate position B of the device (such as the side of the photographic lens 20), and the indicator module lights up when the face feature image is taken online. In addition to the aforementioned authentication mode, the identification module also has a management mode. The identification module does not contain any feature data before the identification module has been managed mode (and therefore cannot and cannot be performed) The access control of face feature recognition), and the new function in its mode is to add a feature data for comparison in the future identification mode. The delete function is to delete the feature data that already exists in the M319045 pen through the human-machine interface of the device. The management mode can also include a replacement function, which is similar to the new one, but replaces a feature that already exists. After the device 10 is normally turned on, the user can activate, delete, or replace the management mode of the identification module through the human-machine interface. When the new function is performed, the recognition module dynamically displays the image captured by the photographic lens on the screen 30, and the test is $4〇. The user hunts his or her face image to a specific range of the screen 3G defined by the reference pattern 4 by adjusting its position relative to the photographic lens 2 远, and the angle. Next, the user instructs the recognition module to process and store the image in the specific range as a pen feature data through a human-machine interface (for example, a specific key combination). Please note that the identification module can add (four) images of the entire specific cost or image processing (10) to save important faces: to save memory space and save the processing time required for future comparison. Please also note! ^ The subject of this creation is not the image processing method used. Many related different methods have been extensively disclosed in the reading or prior art, so there is no description in the identification module. After a feature data, the user can activate the access control function of the device 10. By the above detailed description, the material (4) is more clearly described in this J Μ. On the contrary, the purpose is to arbitrarily cover the various changes in the equality of the arrangements in this creation. M317045 i ', [Simple description of the drawing] Fig. 1 is a schematic diagram of a notebook computer having a face recognition function according to an embodiment of the present invention. Figure 2a is a schematic diagram of a reference pattern in accordance with an embodiment of the present invention. Figure 2b is a schematic diagram of a reference pattern in accordance with another embodiment of the present invention. Figure 2c is a schematic diagram of a reference pattern in accordance with yet another embodiment of the present invention. Fig. 2d is a schematic diagram of a reference pattern according to still another embodiment of the present creation. [Main component symbol description] 10 This unit 20 Photo lens 21 Lighting unit 30 Screen 40 Reference pattern 11