201010402 九、發明說明： 【發明所屬之技術領域】 本發明涉及一 作控制方法。 【先前技術】 種具有彳功能之電子設備及其操201010402 IX. Description of the invention: [Technical field to which the invention pertains] The present invention relates to a control method. [Prior Art] An electronic device with a cymbal function and its operation
目前對於產品的旋轉動作_#許多應用，包括許多 =士數位相機或掌上電腦，都内建了旋轉動㈣測技術， :藉由旋轉動作的偵測，達到變更操作功能的目的。例如， 變換歌曲曲目或流覽下一張照片等操作。 然而這些技術的應用，都是透過電子產品内建硬體動 =測器，如加速度感測器、三軸感測器及多軸感測器達 到偵測的目的。從而會增加產品成本。 而目則大多數多媒體產品都内建了影像感測器，如數 =目機+機及掌上電腦等，因此若能利用現有内建的影 像感測器，實現相同功能，科、可 „ n^ 就了以減少採用硬體動作感測 器的使用，從而達到降低成本之目的。 【發明内容】 有雲於此’有必要提供一種利用影像感測進行動作賴 測進仃操作控制之電子設備及其操作控制方法。 、種電子°又備，其包括影像感測器及監視器。所述影 像感測器用於感測圖像，監視器用於顯示所述影像感測器 感測的圖像及操作界面。所述監視器具有-基準方向，所 述電子《X備還包括操作控制系統，所述操作控制系統包 括：獲取模塊，用於獲取影像感測11感測之圖像；特徵提 201010402 取模塊，用於從拍攝之圖像中提取 用於判斷圖像中豎直向上之 M ，方向判斷模塊， 角；存所述鱗方向之間夹 塊，根據料方向騎模塊^ /之操作指令；操作模 叙^執行對應操作指令。 -_控制方法，該操作控制 ： ί之準方向的夾角進行控制操作，其包括二 圖像，k取圖像中之特徵量；獲得 ❹ =Γ向之央角;根據得到之夹角執行相應操作。 该麵作控制系統可以通過獲取圖像得出監視器之 :向相對拍攝圖像中豐直向上之方向轉過之角度，從而調 整操作界面之顯不方向，便於使用者觀察。由於無需使用 硬體動作感測器’所以可以大大降低旋轉偵測之成本。 【實施方式】 下面將結合附圖對本發明實施方式作進一步之詳細描 述0 • 請參閱圖1、圖2及圖3,本發明實施方式提供之一種 電子設備1。該電子設備i可以為數位相機、數位攝像機或 帶有攝像頭的其他電子設備，在本實施方式中，所述電子 設備1為數位相機。 所述電子設備1包括影像感測器20、操作控制系統3〇 及監視器40。 所述影像感測器20用於感測圖像。所述監視器40具 有底部41及頂部42。所述監視器40用於顯示所述影像感 測器20感測的圖像及操作界面。本實施方式中，所述操作 201010402 界面為菜單及圖標顯示。所述監視器4〇具有一基準方向 N’本實施方式中，所述基準方向N垂直所述監視器4〇的 底部41指向所述監視器4〇頂部42。 所述操作控制系統30包括：獲取模塊31、特徵提取模 塊32、方向判斷模塊33、操作模塊34及存儲模塊35。 e 所述獲取模塊31，用於獲取影像感測器20感測之圖 ，。所述特徵提取模塊32，用於從拍攝之圖像中提取特徵 =。所述特徵量至少由圖像之顏色、亮度、紋理、形狀或 空間關係中的一種轉化而成。本實施方式中，特徵提取模 塊=分別提取拍攝圖像的顏色及亮度特徵量，顏色特徵量 的提取係先將拍攝的圖像劃分為複數個區域，將各區域的 階矩與二階矩作為顏色特徵。亮度特徵量的提取 的==像劃分為複數個區域，將各區域利用現有 Si::::圖方法求得邊緣圖像⑽ 算法：：判斷模塊33，用於將特徵量通過圖像方向識別、, 鼻法异出圖像中彆亩Α μ μ + t ^ ^^ 夾角。本實^ ㈣料鲜方向N的 部仏所述基準方向n垂直底部41指向頂 斤相像方向朗演算法可 該學習機利用大量影像的顏 元成， =轉化成的特徵量作為樣本進行;練:實= 所述圖像方向識別演算 只轭方式中， 為樣本利用支援向量里機==轉化 里、機制為以大量圖片朝上的方向相對基準方向 -201010402 ❹ 不同夾角作為不同類別的樣本進行训練。訓練時，首先將 每個圖片劃分為複數個區域，將各區域的顏色的一階矩與 二階矩作為顏色特徵，從而提取出顏色特徵量。在提取亮 度特徵量時，依然將各圖片劃分為複數個區域，將各區域 =用邊緣方向直㈣求得邊緣圖像攜帶的亮度f訊以此 獲取亮度的特徵量。然後再利用支持向量機進行分類學 二:’訓練出的支援向量機可以將圖片的各種角度分入不同 ’該類別可以係每隔1度—個類別，或每隔1〇度-個 施方^於類別越多’訓練時間越長，計算越複雜。本實 於控制電子設備1的幾個控制動作，並且 片採用每隔90度為-個類別，將大量圖 -度，。度及27。度作;時:夾角r度、 訓練完成後，-旦輸入一㈣:^進订訓練。當學習機 學習她H 幅圖片的顏色及亮度的特徵量， 入的特：量：根大量圖片剌練時所形成的經驗對輸 實施方式中，所述方向到相應的類別結果。本 取拍攝圖像的顏色及亮提取模塊32提 出相應的類別，即圖像：Γ直：f入訓練好的學習機’得 方向N的央角。 2直向上的方向Μ相對所述基準 存儲模塊35，用於在锉+ Α 操作指令操作指令。所述 電子設備進行選項的』=不同對具有複數個選項的 播放功能的手機進行多媒體播放 201010402 的圖形進行控制。例 式中，斛、+、# 系統圖私位置的調整。太 ： 述操作控制系統30包括複數個符梦 實包方 .標或文字。本實施方式中，所::二，符號可 細作指令用於嘴敕馬圖。所述 7用於調整所述符號相對於 Μ 角。本實施方式中，當所述監視器4〇上^的方=:的爽 f向上的方向μ相對於基準方向ν逆時針m圖像登 所述操作指令為將所述符號相對於所述基準方W90度時， ，針夹角9。度;當所述監視器：上方以^ 像暨直向上的方向M 顯不的拍攝圖 度時，所述摔作於八兵技卓向Nl£時針夾角為⑽ 夾角調整為=:=符r對於所述基準方向n 拍攝圖像丄當所述藍視器4〇上顯示的 為270产睹β M相對於基準方向1"逆時針夾角 _ ^時，戶斤述操作指令為將所述符號相對於所述某準 方向N爽角調整為逆時針夾角,度。 所述基準 ❿ 塊A根據所述方向判斷模塊33得出的夾角執 仃麵射日^請參閱圖4,當所述電子設備i順時針旋轉 9〇度時，所述基準方向N隨著監視器40順時針旋轉9〇度， 2監視器40上顯示的拍攝圖像豎直向上的方向μ相對又於 基準方向Ν逆時針夾角為9〇度，所述操作模塊34執行將 所述1號相對於所述基準方向Ν夾角調整為逆時針爽角9〇 度的操作指令，其餘角度的操作以此類推。從而無論數位 相機在拍攝圖像時旋轉多少度，所述符號的方向總係相應 的反向旋轉相應的角度，始終與圖像豎直向上的方向Μ相 同，便於操作者的觀察及操作。 11 201010402 請參閱圖5’係本發明實施方式提供的-種操作控制方 法的流程圖。 $ • 步驟SU〇:所述獲取模塊獲取影像感測器2〇感測 的圖像。 步驟S115:所述特徵提取模塊32提取圖像中的特徵 量，所述特徵量至少由圖像的顏色、亮度、紋理、形 空間關係令的-種轉化而成。本實施方式中，特徵提取模 ❷ 塊32分別提取拍攝圖像的顏色及亮度特徵量，顏色特 的提取係先將拍攝的圖像劃分為複數個區域，將各區域的 顏色的-階矩與二階矩作為顏色特徵。亮度特徵量的提取 系先將拍攝的圖像劃分為複數個區域’將各區域利用邊緣 方向直方囷求得邊緣圖像攜帶的亮度資訊，以此獲取亮声 的特徵量。 叉％儿度 步驟S120 :獲得感測的影像豐直向上的方向與基準方 2的夾角。湘圖像方向酬演算法對獲取的特徵量進行 ^析從而獲得感測的影像豎直向上的方向與基準方向的夹 4/本實施方式中，所述基準方向N垂直底部41指向頂部 與習^圖像方向識別演算法可以係通過學習機完成，該 i轉化成Li量影像的顏色、亮度、紋理、形狀或空間關 _ 、、徵量作為樣本進行訓練。本實施方式中，所 的 識別演算法採用大量圖像的顏色及亮度轉化成 、$量作為樣本利用支持向量機進行訓練。 準方=支持向量機的機制為以大量圖片朝上的方向相對基 ° ^旋轉不同夾角作為不同類別的樣本進行訓練。首 12 201010402 先，將每個圖片劃分為複數個 .階矩與二階矩作為顏色特徵，從而描域的顏色的- 然將各圖片劃分為複數個區域，將 =域利用邊緣方向直方圖求得邊緣圖像攜帶的 度的特徵量。然後再利用支持向量機進行 入不同類別，該類別可以係每隔！度-個類別：f:。 =個類別。由於類別越多，訓練時間越長，計算越複雜。 中：於控制電子設備1的幾個控制動作， ί圖所以採用每隔9〇度為-個類別，將大 圖片&直向上的方向Μ相對基準方向 度，度、⑽度及270度作為四個類別十失角^ ，機訓練完成後一旦輸人—幅圖片的顏色及亮I的二 篁，學習機就可以根據先前大量圖片訓練時所形成的經驗 對輸入的特徵量進行歸類處 φ 旲。太眚竑+ ϋ 攸而侍到相應的類別結 果本實施方式中，所述方向判斷模塊3 32提取拍攝圖像的顏色及 :从取模塊 機，得出相應的類別，即圖像;==:好的學習 述基準方向Ν的夾角。 直向上的方向Μ相對所 步驟S125:判斷圖像中&直向上的方向μ相對基準方 向Ν的夾角相應的操作。本實施方式中 塊33判斷圖傻中私吉* μ沾士人χ 4々Π Α辦模 角相應的操7乍 的方向M相對基準方向N的央 步驟S130:根據得到的夾角執行相應操作。本實施方 13 201010402 式2，當獲得不同的角度時，操作模塊34根據對應的角度 ··將符號相對於所述基準方向N做旋轉操作。所述符號為^ 統圖標。 ‘’' 該操作控制系統可以通過獲取圖像得出監視器的基準 方向相對拍攝圖像中豎直向上的方向轉過的角度，從^調 整操作界面的顯示方向，便於使用者觀察。由於無需使用 硬體，作感測器’所以可以大大降低旋轉㈣的成本。 φ θ綜上所述，本發明確已符合發明專利之要件，遂依法 提出專利申請。惟，以上所述者僅為本發明之較佳實施方 式，自不能以此限制本案之申請專利範圍。舉凡熟悉本案 技藝之人士援依本發明之精神所作之等效修飾或變化，皆 應涵蓋於以下申請專利範圍内。 白 【圖式簡單說明】 圖1為本發明實施方式提供的一種電子設備的硬體架 構圖； /' 鲁圖2為圖1的電子設備的操作控制系統的功能模塊圖； 圓3為圖1的電子設備的監視器的操作界面的示意圖； 圖4為圖1的電子設備的監視器旋轉後的操作界面示 圖5為圖1的電子設備的操作控制方法流程 【主要元件符號說明】 電子設備 1 操作控制系統 30 特徵提取模塊 32 影像感測器 20 獲取模塊 31 方向判斷模塊 33 201010402 操作模塊 34 存儲模塊 35 監視器 40 監視器底部 41 監視器頂部 42At present, for the rotating action of the product _# many applications, including many = digital camera or PDA, have built-in rotary (four) measurement technology: through the detection of rotating motion, to achieve the purpose of changing the operation function. For example, changing the song track or browsing the next photo. However, the application of these technologies is achieved through the built-in hardware of electronic products such as acceleration sensors, three-axis sensors and multi-axis sensors. This will increase the cost of the product. However, most multimedia products have built-in image sensors, such as number = camera + machine and handheld computer, so if you can use the existing built-in image sensor to achieve the same function, you can In order to reduce the use of hardware motion sensors, the purpose of reducing the cost is achieved. [Summary of the Invention] There is a need to provide an electronic device that uses image sensing to perform motion control operations. The operation control method, the electronic device further includes an image sensor and a monitor, the image sensor is used for sensing an image, and the monitor is used for displaying the image and operation sensed by the image sensor. The interface has a reference direction, and the electronic device further includes an operation control system, and the operation control system includes: an acquisition module, configured to acquire an image sensed by the image sensing 11; and the feature is taken in 201010402 a module for extracting M from the captured image for judging the vertical upward direction of the image, the direction judging module, the corner; storing the block between the scale directions, riding the module according to the material direction ^ / The instruction is executed; the operation mode describes the execution of the corresponding operation instruction. The -_ control method controls the operation of: ί the angle of the quasi-direction control operation, which includes two images, k takes the feature quantity in the image; obtains ❹ =Γ To the corner of the corner; perform the corresponding operation according to the angle obtained. The surface control system can obtain the image by acquiring the image: adjusting the operation interface by turning the angle in the direction of the straight upward direction relative to the captured image. The display direction is convenient for the user to observe. Since the hardware motion sensor is not required, the cost of the rotation detection can be greatly reduced. [Embodiment] Hereinafter, embodiments of the present invention will be further described in detail with reference to the accompanying drawings. Referring to FIG. 1 , FIG. 2 and FIG. 3 , an electronic device 1 provided by an embodiment of the present invention may be a digital camera, a digital camera or other electronic device with a camera. In this embodiment, the The electronic device 1 is a digital camera. The electronic device 1 includes an image sensor 20, an operation control system 3A, and a monitor 40. The image sensor 20 is used for The monitor 40 has a bottom portion 41 and a top portion 42. The monitor 40 is used to display an image and an operation interface sensed by the image sensor 20. In the embodiment, the operation 201010402 interface is used. Displayed for menus and icons. The monitor 4 has a reference direction N'. In the present embodiment, the reference direction N is perpendicular to the bottom 41 of the monitor 4A and is directed toward the top 42 of the monitor 4. The operation control system 30 includes an acquisition module 31, a feature extraction module 32, a direction determination module 33, an operation module 34, and a storage module 35. The acquisition module 31 is configured to acquire a map sensed by the image sensor 20. The feature extraction module 32 is configured to extract feature= from the captured image. The feature quantity is converted by at least one of color, brightness, texture, shape or spatial relationship of the image. In this embodiment, the feature extraction module=extracts the color and the brightness feature quantity of the captured image respectively, and the extraction of the color feature quantity first divides the captured image into a plurality of regions, and uses the moment and the second moment of each region as the color. feature. The extracted == image of the luminance feature quantity is divided into a plurality of regions, and each region is obtained by using the existing Si:::: map method to obtain an edge image (10). Algorithm: The determination module 33 is used to identify the feature amount through the image direction. ,, in the nasal method, the angle of μ μ + t ^ ^^ is different. The actual ^ (4) the direction of the fresh direction N, the reference direction n the vertical bottom 41 points to the direction of the image of the jack-up image. The learning machine can use a large number of images of Yan Yuancheng, = the characteristic quantity converted into a sample; practice: = The image direction recognition algorithm is only in the yoke mode, and the sample is used to support the vector in the machine == conversion, and the mechanism is to train the sample with different angles in the direction of the large number of pictures upwards relative to the reference direction -201010402 ❹ different angles . During training, each picture is first divided into a plurality of regions, and the first moment and the second moment of the color of each region are used as color features, thereby extracting the color feature amount. When extracting the luminance feature quantity, each picture is still divided into a plurality of areas, and each area = the edge direction is straight (four) to obtain the brightness information carried by the edge image to obtain the feature quantity of the brightness. Then use the support vector machine for taxonomy 2: 'The trained support vector machine can divide the various angles of the picture into different 'this category can be every 1 degree - one category, or every 1 degree - one prescription ^ The more the category, the longer the training time, the more complicated the calculation. This is to control several control actions of the electronic device 1, and the slice adopts a plurality of graphs every 90 degrees. Degree and 27. Degree; time: angle r degree, after training is completed, enter one (four): ^ book training. When the learning machine learns the feature quantity of the color and brightness of her H picture, the special quantity: the experience formed by the large number of pictures during the practice, the direction to the corresponding category result. The color and light extraction module 32 of the captured image is presented with the corresponding category, that is, the image: straight: f enters the training machine's direction to obtain the central angle of the direction N. 2 Straight upward direction Μ relative to the reference storage module 35 for operating the command operation command at 锉+ 。. The electronic device performs the control of the multimedia video playback 201010402 for the mobile phone with the option of the playback function of the plurality of options. In the example, the adjustment of the private position of the system diagrams of 斛, +, #. Too: The operation control system 30 includes a plurality of characters, a standard or a text. In the present embodiment, the :: two, the symbol can be used as a command for the mouth and horse. The 7 is used to adjust the symbol relative to the corner. In the present embodiment, when the direction of the monitor 4 is lower than the direction μ of the upward direction, the image is inverted counterclockwise with respect to the reference direction ν, and the operation command is to compare the symbol with respect to the reference. When the square is 90 degrees, the needle is at an angle of 9. Degree; when the monitor is above the image in the direction of the cum straight upward M, the angle of the fall is equal to the angle of the needle (N) and the angle is adjusted to === Taking an image for the reference direction n, when the blue vision device 4 显示 displays 270 睹β M with respect to the reference direction 1 " counterclockwise angle _ ^, the operation instruction is to use the symbol The angle S is adjusted counterclockwise with respect to the certain quasi-direction. The reference block A is based on the angle obtained by the direction determining module 33. Referring to FIG. 4, when the electronic device i rotates 9 degrees clockwise, the reference direction N follows the monitoring. The device 40 rotates 9 degrees clockwise, and the direction of the vertical upward direction μ of the captured image displayed on the monitor 40 is opposite to the reference direction by an angle of 9 degrees, and the operation module 34 executes the number 1 The operation angle of the counterclockwise refresh angle is adjusted to 9 degrees with respect to the reference direction, and the operation of the remaining angles is similar. Thus, no matter how many degrees the digital camera rotates when the image is taken, the direction of the symbol is always the corresponding reverse rotation angle, always the same as the vertical direction of the image, which is convenient for the operator to observe and operate. 11 201010402 Please refer to FIG. 5' for a flowchart of an operation control method provided by an embodiment of the present invention. $ • Step SU〇: The acquisition module acquires an image sensed by the image sensor 2〇. Step S115: The feature extraction module 32 extracts feature quantities in the image, and the feature quantity is converted from at least the color, brightness, texture, and spatial relationship of the image. In the present embodiment, the feature extraction module block 32 extracts the color and the brightness feature quantity of the captured image, respectively, and the color special extraction system first divides the captured image into a plurality of regions, and the color-order moment of each region is The second moment is used as the color feature. The extraction of the luminance feature quantity is performed by dividing the captured image into a plurality of regions. The luminance information carried by the edge image is obtained by using the edge direction of each region to obtain the luminance information. Fork % Child Step S120: Obtain the angle between the direction in which the sensed image is straight and upward and the reference square 2. The Xiang image direction reward algorithm analyzes the acquired feature quantity to obtain a clip of the sensed image in the vertical direction and the reference direction. 4 In this embodiment, the reference direction N is perpendicular to the bottom 41 and points to the top and the habit. ^ Image direction recognition algorithm can be completed by the learning machine, and the i is converted into the color, brightness, texture, shape or space of the Li image, and the levy is trained as a sample. In the present embodiment, the recognition algorithm converts the color and brightness of a large number of images into a quantity and uses the amount as a sample to perform training using a support vector machine. The mechanism of the quasi-square = support vector machine is to train different samples in different directions with a large number of pictures facing upwards. First 12 201010402 First, each picture is divided into a plurality of moments and second moments as color features, so that the color of the field - then the pictures are divided into a plurality of regions, the = domain is obtained by using the edge direction histogram The feature quantity of the degree carried by the edge image. Then use the support vector machine to enter different categories, the category can be every! Degree - Category: f:. = a category. The more the categories, the longer the training time, the more complicated the calculation. In the control of several control actions of the electronic device 1, the map is used for every 9 degrees, and the direction of the large picture & straight direction is relative to the reference direction, degree, (10) degree and 270 degrees. The four categories are ten out of corners ^. Once the machine is trained, once the input is made - the color of the picture and the brightness of the picture I, the learning machine can classify the input feature according to the experience formed by the previous large number of pictures. φ 旲. In the present embodiment, the direction judging module 3 32 extracts the color of the captured image and: from the module taking machine, the corresponding category, that is, the image; : Good study of the angle between the reference directions. The direction of the straight upward direction is relative to the step S125: determining the operation corresponding to the angle between the direction μ of the straight direction and the reference direction 图像 in the image. In the present embodiment, the block 33 judges that the direction M of the corresponding operation angle is relative to the reference direction N. Step S130: Perform a corresponding operation according to the obtained included angle. The present embodiment 13 201010402 Equation 2, when different angles are obtained, the operation module 34 performs a rotation operation with respect to the reference direction N according to the corresponding angle. The symbol is a ^ icon. ‘’' The operation control system can obtain the angle at which the reference direction of the monitor is rotated in the direction of the vertical upward direction of the captured image by taking the image, and adjust the display direction of the operation interface to facilitate the user to observe. Since the sensor is not required to be used as the sensor, the cost of the rotation (four) can be greatly reduced. φ θ In summary, the present invention has indeed met the requirements of the invention patent, and has filed a patent application according to law. However, the above description is only a preferred embodiment of the present invention, and it is not possible to limit the scope of the patent application of the present invention. Equivalent modifications or variations made by persons skilled in the art in light of the present invention are intended to be included within the scope of the following claims. BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a hardware structural diagram of an electronic device according to an embodiment of the present invention; FIG. 2 is a functional block diagram of an operation control system of the electronic device of FIG. 1; FIG. 4 is a schematic diagram of an operation interface after the monitor of the electronic device of FIG. 1 is rotated. FIG. 5 is a flow chart of the operation control method of the electronic device of FIG. 1 [Description of main component symbols] 1 Operation Control System 30 Feature Extraction Module 32 Image Sensor 20 Acquisition Module 31 Direction Determination Module 33 201010402 Operation Module 34 Storage Module 35 Monitor 40 Monitor Bottom 41 Monitor Top 42
1515