M426829 五、新型說明： 【新型所屬之技術領域】 本創作係指-種影像處理裝置及影像擁取裝置，尤指一種可比 對一影_取裝置產生之-職影像+ _特定區軸之—人臉數量 是否屬於-預設數量範圍，以指示使用者進行拍照或調整該預覽影 像之影像處理裝置及影像擷取裝置。 _【先前技術】 隨著數位影像擷取技術的不斷發展，手機絲記型電腦等行動 裝置普遍轉触械的魏整合於射。_，辭人人都能隨 時隨地拿著相機或手機自減與朋友們—_照，也使得相機自拍 的=⑽财。⑽，瓣自输财_不到相機顯 =幕上的觀畫面，因此可能於自拍完後發現自己的臉沒有拍到M426829 V. New description: [New technical field] This creation refers to an image processing device and image capturing device, especially a comparable image-taking image generated by a film-taking device. Whether the number of faces belongs to the preset number range, and the image processing device and the image capturing device for instructing the user to take a picture or adjust the preview image. _[Prior Art] With the continuous development of digital image capture technology, mobile devices such as mobile phone silk-screen computers have generally turned into mechanical integration. _, the resignation can take the camera or mobile phone at any time and anywhere with friends - _ photos, also make the camera self-timer = (10). (10), the flap from the money _ less than the camera display = the screen on the screen, so it may be found after the self-timer that his face is not photographed
整地出現在相片t，而需要反覆地重拍，直到拍到想要的 目=止。在此情況下，不健f時間，也雜數位相機及行 置的電池電力，因而減少數位相機及行動裝置的使用時間。 因此’為了解決上述的問題’於習知技射，常見的解 係提供兩個影像麻裝置及單—顯示勞幕，其中—取晋 設置於顯示螢幕的同-側，.以供使_行自拍；另_=== 裝置设置於顯示營幕後方，以供使用者進行—般拍照。或者 習知技術储解-影像擷取裝置及兩麵村幕，該兩個顯示營 3 影像擷轉置的前方及後方，使得使用者位於影像_， 圍:1或後方時’皆能即時確認所有參與者的臉都在取景範 除此之外’亦可設計可轉的旋轉機構於顯示螢幕或者提 八自拍鏡，同樣能讓使用者即時觀看相機的影像。 然而，習知技術提供的解決方法仍然有缺點尚待改進。舉例來. 習知技術皆以新增裝置的方式來達到即時觀看自拍時的預覽影 ，’不論是增加域·、顯錢幕、設計可翻轉的旋轉機構或是 、拍鏡，皆增加設計行喊置的複雜度以及生產成本。除此之外，# 增加相機模組及顯示螢幕聘也增加行動裝置的耗電量，因而降低 仃=裝置的使用時間。至於設計可翻轉的機構則會因為機構轉轴於 4使用之下磨損，增加後續維修的成本。另一方面，由於目前市 見的自拍鏡係經由表面電鑛製成的凸面鏡，當使用者隨身揭 帶订動裝置並且無特別保護措施時，一段時間後往往會發現自拍鏡 已磨損，使得預覽影像之功能失效。 因此’基於上述之缺點，習知技術實有改進之必要。在提供使_ =者自拍雜的同時，也騎慮降低行動裝置的生產成切及故障 率。 【新型内容】 因此’本創作之主要目的在於提供一種可比對一影像揭取裝置 生之-預覽影像卜特雜_之—人臉數量是否屬於—預設數 M426829 置範圍，以指示使用者進行拍照或調整該預覽影像之影像處理裝置 及影像擷取裝置。 本創作提供一種影像處理裝置，用於一影像擷取裝置，包含有 -影像純單元’帛來減該f彡像触裝置難生之—預覽影像； .一人，辨識單元，用來辨識及計算該籠影像卜肢區域内之人 臉數！，-比對單元’用來比對該預覽影像中該特定區域内之人臉 «是否屬於-職數量細；从—飾單元，絲根據該比對 早70之—比對結果，輸出一指示訊號。 預覽二種影_取裝置’包含有一鏡頭’用來產生一 單元，用來理裳置，峰於該鏡頭，其包含有—影像接收 單元，用來辨1及2細取裝置所產生之一預覽影像；一人臉辨識 比對單元，用來比對該預覽特疋區域内之人臉數量；一 於—預設數量範圍；以及中該狀區域内之人臉數量是否屬 對結果，輸出-指示訊號。曰不早元’用來根據該比對單元之一比 【實施方式】 請參考第1圖，第i 圖。影像處理裝置1〇用於^本創作之一影像處理裝置10之示意 是獨立的電子產品如數位I影像擷取裝置中’該影像擷取裝置可以 影像處理裝置10包含有一 1機，或是裝置於手機及筆記型電腦中。 /像接收單元102、一人臉辨識單元104、 M426829 —比對單元106及一指示單元108。影像接收單元1〇2用來接收影 像擷取裝置之一鏡頭100所產生的一預覽影像，亦即影像擷取 裝置於拍照前所擷取到的預覽影像。而人臉辨識單元1〇4則於預覽 衫像IMG之一特定區域A (未繪出）中，進行人臉辨識以及計算預 覽影像IMG中出現的人臉數量FN。比對單元1〇6則用來比對人臉 數1FN是否等於一人臉預設數量SN，以輪出一比對結果res至 指示單元108。接著，指示單元1.08根據比對結果，輸出一指 示讯號IND，以通知使用者進行拍照動作；或是移動鏡頭1〇〇，進 而調整預覽影像IMG。 具體來說，請參考第2A圖及第2B圖，第2Λ圖及第2B圖為 第1圖之影像處理裝置10與使用者互動之示意圖。如第2A圖所示， 假設被攝者UA、UB共兩人，當啟動影像擷取裝置之自拍提示功能 時’根據本發明實施例，使用者（即被攝者UA、⑽其中之一）可 設定人臉預設數量SN為2以及設定特定區域A之範圍。在進行拍 照則，影像處理裝置10可先將預覽影像對應到一座標，並於 。玄跳中建立特定區域A相對應之一子座標。待人臉預設數量sn 及特定區域A的設定完成之後，以第2A圖為例，人臉辨識單元104 於預覽影像IMG之特定區域a中辨識出人臉數量FN為卜並且於 預覽如像IMG中總共辨識出2張人臉。接著，比對單元1〇6判斷人 臉辨識單元104輸出的人臉數量FN小於人臉預設數量SN，且有一 人臉位於特&區域人相對應之子座標的左側，即使用者⑽的人臉 位於特定區域A的左側，因此輸出對應的比對結果至指示單元 M426829 ⑽。指不單元108根據比對結果哪，輸出指示訊號_，通知 使用者UB調整所在的位置。假設指示單元⑽係 一味jn八，則可輸 出「左邊使用者請向右移動」的語音指示，通知使用者咖往特定 區域A的範圍内移動，如第2B圖所示。在此情況下，當使用者欲 / ^行自動拍照時’歧無法即時看·像擷取裝置之職影像⑽ ^伽者不僅可以預先奴邮巾欲&現的人臉預設數量艰，也 ^以攻定人臉要出現在相片巾的特定區域Α，而影像處理裝置川 #可根據上述設定’比對預覽影像觸是否符合設定，以通知使用者 妨拍照或是輕職影像IMG。如此—來，使用者可於拍照前， 依知衫像處理裝置10所發出的指示訊號励調整預覽影像聰， 以減少因拍照失敗而重複自拍的次數。 ^三之’本瓣之影像處理裝置1G可根據使用者設定的人臉數 ,卜及认疋範圍’判斷是否通知使用者進行拍照或是調整動作，以減 鲁乂使用者重覆自拍的次數。需注意的是，影像處理裝置1〇之影像接 .收單元i02、人臉辨識單元U)4、比對單元106及指示單元刚係用 $不只現相關功效之功能方塊，其可能由軟體、動體或硬體等方 貫現’且可適當將影像處理裝置1G包含的功能加以分配，以適應 ^類型之電子產品的設計需求。換句話說，當將影像處理裝置1〇 麵置時，可將影像處置财包含的功能單元重新配 符5行動裝置的設計需求。 +例來說’請參考第3圖及第4圖’第3圖及第4圖為本創作 7 實施例具影像擷取功能之行動裝置3〇及奶 包含有-縣#裝4 30G、— #㈣ 〜圖。仃練置30 衫像處理裝置3〇2及— 。β _ 鄕。影像擷取裝置300可為— 及&不早兀 device >CCD), (charge-coupled ccd) m〇 〇 一人臉辨識單元304及一比對單 "02已含有 人臉辨識，比對預覽影像IMG是否符人使用^預覽域觸執行 應的比對結果。指示單元3〇8 ^用者之設定，以及輸出對 08則根據比對結果RES，輪出相料 應的指示動作，通知使用者進 輸出相對 嘖史老望相—Μ 一飞者調整預覽影像。另-方面， Γ 圖，仃動裝置3G及4G不同之處在於，行練置40之 影像擷取裝置彻包含—人臉辨識單元彻。行鮮=旦之— 處理裝置402僅包含有一比斟显_ 、置之衫像 ^ 。在此結構下，可將人臉辨 2 ^ 置伽中，而影像處理I置搬則僅負責 办像比對功能。如此一來，雖麩符 、、 料摇古^甘 I、間化了衫像處理裝置402之功能， 部也“增加其他影像處理功能的空間。 〜因此，由於彈性化的影像處理功能配置，景燦處理裝置10之功 本創作之自拍提示功能，亦可增加或執行其他的影像處理The whole scene appears in the photo t, and it needs to be retaken repeatedly until the desired one is taken. In this case, the battery power of the camera and the device is also reduced, which reduces the use time of the digital camera and the mobile device. Therefore, in order to solve the above problems, in the conventional technique, the common solution system provides two image hemp devices and a single-display screen, wherein - the setting is set on the same side of the display screen, for the _ line Self-timer; another _=== The device is placed behind the display screen for the user to take a picture. Or the conventional technology storage-image capturing device and the two-sided village screen, the two display camp 3 image 撷 transposed front and rear, so that the user can be instantly confirmed when the image is located in the image _, circumference: 1 or rear All participants' faces are in the framing range. In addition, the rotating mechanism can be designed to rotate on the display screen or the self-portrait mirror. It also allows the user to instantly view the camera image. However, the solutions provided by the prior art still have shortcomings to be improved. For example, the conventional technology uses the newly added device to achieve the preview image when watching the self-timer in real time. 'Whether it is adding a field, displaying a money screen, designing a reversible rotating mechanism, or taking a picture, all the design lines are added. The complexity of the call and the cost of production. In addition, the addition of camera modules and display screens also increases the power consumption of mobile devices, thus reducing the use time of 仃=devices. As for the design of the reversible mechanism, the wear of the mechanism shaft is used under 4, which increases the cost of subsequent maintenance. On the other hand, since the self-portrait mirrors currently seen in the market are convex mirrors made of surface electric ore, when the user carries out the strapping device with no special protection measures, the self-portrait mirror is often worn out after a period of time, making the preview The function of the image is invalid. Therefore, based on the above disadvantages, the prior art is in need of improvement. While providing _ = self-timer, it also rides down to reduce the production cut and failure rate of the mobile device. [New content] Therefore, the main purpose of this creation is to provide a comparison with the image-receiving device - the preview image is the number of faces - the preset number M426829 is set to indicate the user An image processing device and an image capturing device that take a picture or adjust the preview image. The present invention provides an image processing device for an image capturing device, including a - image pure unit '帛 to reduce the difficulty of the image sensor device - preview image; one person, identification unit for identification and calculation The number of faces in the cage image area! , - the comparison unit 'is used to compare the face in the specific area of the preview image to whether the number of jobs is fine; from the - decoration unit, the silk is 70 according to the comparison - the result is output, and an indication is output Signal. Preview two kinds of shadows _ take-up device 'includes a lens' used to generate a unit for arranging, peaking at the lens, which includes an image receiving unit for distinguishing one of the 1 and 2 fine picking devices Preview image; a face recognition comparison unit is used to compare the number of faces in the preview feature area; one in the preset number range; and whether the number of faces in the area is the correct result, output - Indication signal.曰 Not early ’ is used to compare one of the comparison units. [Embodiment] Please refer to Figure 1, i. The image processing device 1 is used in an image processing device 10, which is an independent electronic product such as a digital I image capturing device. The image capturing device 10 can include a camera or a device. For mobile phones and laptops. / Image receiving unit 102, a face recognition unit 104, M426829 - comparison unit 106 and an indication unit 108. The image receiving unit 1〇2 is configured to receive a preview image generated by the lens 100 of the image capturing device, that is, the preview image captured by the image capturing device before the camera is taken. The face recognition unit 1〇4 performs face recognition and calculates the number of faces FN appearing in the preview image IMG in a specific area A (not shown) of the preview shirt image IMG. The comparison unit 1〇6 is used to compare whether the face number 1FN is equal to a face preset number SN to rotate a comparison result res to the indication unit 108. Then, the indicating unit 1.08 outputs an indication signal IND according to the comparison result to notify the user to take a photographing action; or moves the lens 1〇〇 to adjust the preview image IMG. Specifically, please refer to FIG. 2A and FIG. 2B. FIG. 2 and FIG. 2B are schematic diagrams of the image processing apparatus 10 of FIG. 1 interacting with the user. As shown in FIG. 2A, it is assumed that two subjects, UA and UB, when the self-timer cueing function of the image capturing device is activated, 'in accordance with an embodiment of the present invention, the user (ie, one of the subjects UA, (10)) The face preset number SN can be set to 2 and the range of the specific area A can be set. When the photographing is performed, the image processing apparatus 10 may first map the preview image to a label, and then. A certain area A corresponds to one of the sub-coordinates. After the setting of the face preset number sn and the specific area A is completed, taking the 2A picture as an example, the face recognition unit 104 recognizes the number of faces FN in the specific area a of the preview image IMG and previews it like IMG. A total of 2 faces were identified. Next, the comparison unit 1〇6 determines that the number of faces FN output by the face recognition unit 104 is smaller than the face preset number SN, and one face is located on the left side of the corresponding sub-coordinate of the special & area person, that is, the user (10) The face is located to the left of the specific area A, so the corresponding comparison result is output to the indication unit M426829 (10). The pointing unit 108 outputs an indication signal _ according to the comparison result, and notifies the user that the UB adjusts the position. Assuming that the pointing unit (10) is blindly jn eight, a voice indication "Please move the left user to the right" may be output to notify the user to move within the range of the specific area A, as shown in Fig. 2B. In this case, when the user wants to take a picture automatically, the image cannot be viewed immediately. (10) ^The gambler can not only pre-empt the mail towel to want & the current face preset number is difficult, Also, the target face should appear in a specific area of the photo towel, and the image processing device can use the above setting to compare whether the preview image touches the setting to notify the user to take a photo or a light job image IMG. In this way, before the camera is taken, the user can adjust the preview image by using the indication signal sent by the processing device 10 to reduce the number of times the self-timer is repeated due to the failure of the camera. ^三之's image processing device 1G can determine whether to notify the user to take a photo or adjust the motion according to the number of faces set by the user, and the range of the vocabulary to reduce the number of times the user repeats the self-timer. . It should be noted that the image receiving device i02, the face recognition unit U) 4, the comparison unit 106, and the pointing unit of the image processing device 1 are only used for function blocks not only related functions, but may be software, The moving body or the hardware is contiguous and the functions included in the image processing apparatus 1G can be appropriately allocated to suit the design requirements of the electronic product of the type. In other words, when the image processing apparatus 1 is placed, the functional units included in the image processing can be reconfigured to meet the design requirements of the mobile device. + For example, 'Please refer to Figure 3 and Figure 4'. Figure 3 and Figure 4 are the mobile devices with image capture function of the creation of the 7th embodiment and the milk contains - County #装4 30G,- #(四) ~ Figure. I practiced 30 shirts like the processing device 3〇2 and —. β _ 鄕. The image capturing device 300 can be - and & not earlier than device > CCD), (charge-coupled ccd) m, a face recognition unit 304 and a comparison list "02 already contain face recognition, Whether the preview image IMG is used to perform the comparison result by using the ^ preview field. The instruction unit 3〇8^user setting, and the output pair 08 are based on the comparison result RES, and the instruction action of the turn-out phase is notified, and the user is notified that the output is relatively old-fashioned—the one adjusts the preview image . On the other hand, the Γ diagram, the swaying device 3G and the 4G differ in that the image capturing device of the riging 40 includes the face recognition unit. The processing device 402 only contains a comparison image, a shirt image ^. Under this structure, the face can be distinguished by 2 ^, while the image processing I is only responsible for the image comparison function. In this way, although the bran, the material shakes the ancient, and the function of the shirt image processing device 402, the part also "adds space for other image processing functions. - Therefore, due to the flexible image processing function configuration, The self-timer reminder function of the work of the Jingcan processing device 10 can also add or execute other image processing.
魏及微笑_等舰，柯限概。㈣影像處理 ^夕樣化的影像處理舰，這些縣處理舰也可絲進一步斑本 合’則乍為使用者控制該影像擷取裂置進行拍照的方式。舉 A兄〜像處理震置1〇可另結合微笑偵測功能，侧預覽影像 中的人像疋，否极夫’畲彳貞測到的微笑數量不符合人臉數量FN M426829 像掏取裝置可藉由指示單元108，如剩口八，輸出語音「請 =笑士」，來it知伽者微笑；#伽__量符合场預設數= ^ ’該影賴轉置可輸出「請拍照」，或是自純行拍照。 队、^ _方面’指示單元⑽可以是任意形式的提示裝置，例如剩 置皆屬於而=是能夠發出訊號以提示使用者的裝 _ $不限於上述裝置。舉例來說，假設指示 域中4&讀’當比對單元106判斷預覽影像IMG的特定區 =中，辨識到的人臉數量FN等於人臉預設數量观時，指示單 Γ臉數^出、綠色指秘號，以通知使用者進行拍照；若辨識到的 红⑼里…不等於人臉預設數量SN時，指示單元108則可輸出 、’工，叫知使肖者輕賊影像細 ^了執行_作，也爛輪其鳩 ^用來播放音樂、指示燈及震動器也可以用來作為手機簡訊提示動 ^辦，細域Α _細_，瓣可任音 特4域A的範圍大小纽於_雜M ^ 舉例來說，請參考第5A圖至第 的U立置 作容鮮W - 圖’第5A圖至第5D圖為本創Wei and smile _ waiting for the ship, Ke limited. (4) Image processing The image processing ship of the eve of the evening, these county processing ships can also be further spliced, and then the user controls the image to take a picture and take a picture. Lift A brother ~ like the processing shake 1 〇 can be combined with the smile detection function, the side preview image in the image 疋, no 极夫 '畲彳贞 The number of smiles does not meet the number of faces FN M426829 like a capture device can By instructing unit 108, such as the remaining port eight, the output voice "please = laugh", to know that the gambler smiles; #伽__ quantity matches the field preset number = ^ 'The shadow can be transposed to output "Please take a photo Or take a photo from a pure line. The team, the _ aspect indicator unit (10) may be any type of prompting means, for example, the remaining ones belong to and = the device capable of emitting a signal to prompt the user is not limited to the above device. For example, suppose the indication field 4& read 'when the comparison unit 106 determines the specific area of the preview image IMG=, the number of recognized faces FN is equal to the number of face presets, indicating the number of faces The green refers to the secret number to inform the user to take a picture; if the recognized red (9) is not equal to the preset number SN of the face, the indication unit 108 can output, 'work, call the knowledge to make the image of the thief light ^ The implementation of _, but also the rotten round 鸠 ^ used to play music, lights and vibrators can also be used as a mobile phone newsletter tips, fine domain Α _ fine _, the valve can be 4 Range size is new to _ miscellaneous M ^ For example, please refer to Figure 5A to the first U stand for the fresh W - Figure '5A to 5D
作只苑例刀別表不不同特定區域A 線區域之蚊_ A可以是Ί ® ° Μ 5 A ®所示，斜 Q心 疋預L轉1MG全部涵蓋的範圍。於第 5B圖中，亦可將預覽影像觸等分為複數個子區域，並H中 一個子區域為特定區域A，。再者 扎崎私疋其中 :弟5C圖中，使用者也可以 〇Ly 同時選擇多個特定區域六 ^ 1 ’像疋在其中一個子區域設定出現兩 二 而在另個子區域中設定出現-張人臉諸如此類的設定。 蓋，2由於目㈣份料機可同時喊數位相機魏與觸控螢 觸如第5D圖所示，特定區域a也可以由使用者在手機的 工上，任意晝出所需的特定區域A。如此不僅增添自拍時的 趣味性’也可増加影像構_準雜。 程ϋ關於影像處縣置10之運作方式，可歸納為一影像處理流、 王i δΗ參考第6圖，第6圖為本創作實施例之影像處理流程60 · 之示思圖。景》像處理流程60包含以下步驟： 步驟600 :開始。 步驟602 :接收鏡頭1〇〇擷取到的預覽影像IMG。 步驟604 :於特定區域A中，判斷預覽影像IMG中出現之人臉 數量FN。 步驟606 :判斷是否人臉數量FN等於人臉預設數量SN。 步驟608 :當人臉數量FN等於人臉預設數量SN時，輸出指示鲁_ 訊號IND，通知使用者進行拍照。 ’步驟610 :當人臉數量FN不等於人臉預設數量sn時，輪出指 示訊號IND，通知使用者調整預覽影像iMG。 步驟612:結束。 影像處理流裎60之詳細說明及變化可參考前敘，於此不贅述。For example, the mosquitoes _ A can be Ί ® ° Μ 5 A ® , and the range of the oblique Q-heart 疋 pre-L to 1MG is covered. In Fig. 5B, the preview image can also be divided into a plurality of sub-areas, and one sub-area of H is a specific area A. In addition, in the 5C picture, the user can also select multiple specific areas in the 5C picture. The image is displayed in one of the sub-areas and in the other sub-area. Settings such as faces. Cover, 2 because the eye (four) dispenser can simultaneously call the digital camera Wei and the touch flash as shown in Figure 5D, the specific area a can also be arbitrarily pulled out by the user on the mobile phone, the desired specific area A . This not only adds to the fun of self-portraits, but also adds image structure. Cheng Hao's operation mode of the county is 10, which can be summarized as an image processing stream, and Wang δ Η refers to FIG. 6 , and FIG. 6 is a schematic diagram of the image processing flow 60 of the present embodiment. The scene processing process 60 includes the following steps: Step 600: Start. Step 602: Receive the preview image IMG captured by the lens 1. Step 604: In the specific area A, determine the number of faces FN appearing in the preview image IMG. Step 606: Determine whether the number of faces FN is equal to the preset number SN of faces. Step 608: When the number of faces FN is equal to the preset number of faces SN, the output indicates the Lu signal IND, and the user is notified to take a photo. Step 610: When the number of faces FN is not equal to the preset number sn of faces, the indication signal IND is rotated to notify the user to adjust the preview image iMG. Step 612: End. The detailed description and changes of the image processing flow 60 can be referred to the foregoing description, and will not be described herein.