1321299 九、發明說明： 【發明所屬之技術領域】 示杰動態影像色彩偏移的檢測1321299 IX. Description of invention: [Technical field of invention] Detection of color shift of demonstration image
示器動態影像品質。 k供一種應用於檢測顯示器動態 系統，丨七、'丄 “ 像、 量化之巴偏狱碉以合理正確的分析顯 【先前技術】 /遺著平面顯示器（如液晶顯示器（LCD)或電漿顯示 器等）普遍應用於電視、電腦監視器、行動電話或各種家 電上，其顯示器面板的品質也愈需要重視，如色彩、對比、 反應間、7C度等。習知技術即已提出對於顯示器面板品 質的檢測方法，如檢測液晶顯示器在播放動態影像時會因 為液晶反應時間的延遲造成邊緣有殘影現象，或檢測於顯 示彩色晝面時在影像邊緣發生色彩偏移的現象。 習知用於上述顯示器之檢測技術有如國際組織VESA (Video Electronics Standards Association)所制定之平面 顯示器 量測標準（Flat Panel Display Measurements Standard， FPDM)，其中305-2章節中揭露一種顯示器殘餘晝面與預 熱時間的量測方法（305-2, Residual Image and Warm-Up-Time Measurement)僅針對顯示器最基本的液晶 反應時間來量測，其中量測殘影或色彩偏移現象的方式係 利用一光感測元件（photodiode )量測固定畫面轉變的液 晶反應，如以光感測元件量測晝面由全黑變化到全白的反 應時間，以及全白變到全黑的反應時間，但其缺點是無法 8 1321299 描述人眼觀看的感覺。 習知技術如 WIPO ( World Intellectual PropertyDisplay dynamic image quality. k for a dynamic system for detecting displays, 丨7, '丄' image, quantification of the prisoner with reasonable and correct analysis [previous technology] / legacy flat panel display (such as liquid crystal display (LCD) or plasma display Etc.) Generally used in televisions, computer monitors, mobile phones or various home appliances, the quality of the display panel needs to be paid more attention, such as color, contrast, reaction room, 7C degree, etc. The conventional technology has proposed the quality of the display panel. The detection method, such as detecting the liquid crystal display, may cause image sticking at the edge due to the delay of the liquid crystal reaction time, or detecting a color shift at the edge of the image when displaying the color surface. The detection technology of the display is like the Flat Panel Display Measurements Standard (FPDM) developed by the international organization VESA (Video Electronics Standards Association). The amount of residual back surface and warm-up time of the display is disclosed in Section 305-2. Measuring method (305-2, Residual Image and Warm-Up-Time Measurement) The most basic liquid crystal reaction time of the display is measured. The method of measuring the residual image or the color shift phenomenon is to measure the liquid crystal reaction of the fixed picture transition by using a photodiode, such as measuring with a light sensing component. The reaction time from full black to full white, and all white to black reaction time, but its shortcoming is that it can not describe the feeling of human eyes. Conventional technology such as WIPO (World Intellectual Property
Organization)第 W02004109634 號與第 W02005004498 號專利揭露一種評價動態影像顯示品質的方法與系統 (METHOD AND SYSTEM FOR EVALUATING MOVING IMAGE QUALITY OF DISPLAYS )，如第一圖所示之裝 • 置’其中有一 CCD攝影機3透過鏡子2拍攝螢幕5的晝面， • 並由一影像訊號產生器9產生測試訊號。鏡子2連接一旋 • 轉軸’藉由驅動控制器7控制旋轉，影像經鏡子2的旋轉 反射，再由CCD攝影機3拍攝，其影像訊號由電腦系統6 接收，電腦系統6更耦接至影像訊號產生器9，由此產生 的測試影像訊號與由CCD攝影機3拍攝的影像進行顯示品 質評價。此案利用鏡面旋轉來達到仿人眼追蹤，再計算螢 幕中影像邊界模糊的程度，並透過不斷的重複追蹤拍攝來 取得第一個追蹤影像及第二個靜態影像來校正追蹤的精確 度，但如此麻煩的步驟降低了準確度，此案描述在黑白影 • 像邊界模糊的狀態並未能實際描述在色彩部分偏移的狀 況。A method and system for evaluating dynamic image display quality (METHOD AND SYSTEM FOR EVALUATING MOVING IMAGE QUALITY OF DISPLAYS) is disclosed in the above-mentioned Patent No. WO2004109634 and No. WO2005004498, which has a CCD camera 3 as shown in the first figure. The face of the screen 5 is photographed through the mirror 2, and a test signal is generated by an image signal generator 9. The mirror 2 is connected to a rotating shaft. The rotating shaft is controlled by the driving controller 7. The image is reflected by the rotation of the mirror 2, and then captured by the CCD camera 3. The image signal is received by the computer system 6, and the computer system 6 is further coupled to the image signal. The generator 9, the resulting test image signal and the image taken by the CCD camera 3 are evaluated for display quality. In this case, the mirror rotation is used to achieve human eye tracking, and then the degree of blurring of the image boundary in the screen is calculated, and the first tracking image and the second still image are obtained through repeated repeated tracking to correct the tracking accuracy, but Such a cumbersome step reduces the accuracy, and the case is described in the state of black and white shadow image boundary blurring and does not actually describe the situation in which the color portion is shifted.
又一習知技術如WIPO第W02004075567號專利所揭 . 路评價動悲影像品質的量測系統/方法（MEASUREMENTAnother conventional technique is disclosed in WIPO Patent No. WO2004075567. The measurement system/method for evaluating the image quality of a sad image (MEASUREMENT)
SYSTEM FOR EVALUATING MOVING IMAGE QUALITY OF DISPLAYS ) ’ 此案係利用—光债測器（ph〇t〇detect〇r ) 偵測動,4影像位置’然仍侷限於黑白影像的邊界模糊判 斷，而且也未能評估彩色影像的品質，而在顏色及亮度轉 換方面準確性不足，且並無色度（chr_)的資訊。 9 1321299 再有 Yoshi Enami，Keiko Kitagishi 與 Koichi Oka (Otsuka Electronics Co· LTD )所提出的利用追蹤攝影系統 進行彩色影像的動態品質量測的方法，係以彩色攝影機才白 攝顯示器中的移動影像’並對其中因為移動造成的模糊狀 態參數化（parameterized)，予以分類，並轉化為一 xy色 度（Chromaticity)空間。 如第二A圖所示之實施例量測實驗結果，其為由白至 黑顏色變化時，從彩色攝影機拍攝的晝面在Xy色度空間之 亮度（Luminance)與時間曲線圖，其中先由拍攝的紅R、 綠G、藍B值轉化為三激值（Tristimulus Values ) X、γ與 Z，即反應由物體的反射率光源的能量，再轉換為色度座 標（x，y)。由第二A圖中所示，紅R、綠G、藍B在黑白 切換時亮度變化的升起時間（rise timing )相同，而色度座 標X與y值（水平線）亦無變化；而第二B圖則顯示另― 實驗結果’其中色度座標X與y值顯示一突起的曲線區域 20與亮度變化產生的偏移，即表示顯示器中彩色模糊的區 域。然而因為人眼視覺並非均勻色度空間，故此利用xy 色度空間表達的量測結果無法插述色差距離之概念。 如韓國三星電子（Samsung Electronics )發表針對液晶 顯示态動作假影的評估與補償的技術（Evaluation of Motion Artifacts and Evolving Compensation Techniques for LCD Monitors )，動作假影（artifact)為晝面移動時產生殘 留的假影，對於液晶顯示的品質有關鍵性的影響，此篇論 文針對各種假影進行分類，並加以分析，在個別判斷後藉 軟體手段進行補償，然其中對於顯示瑕疵的判斷僅計算主 1321299 色色差，未能描述邊緣色彩模糊的表現。 上述習知量測方式.以光二極體（photodiode )等光感 測器量測液晶顯示器反應時間，但因並無引入人眼的檢測 參數，故其檢測結果無法真正反映出人眼反應。 【發明内容】 有鑑於習知技術對於液晶顯示器的檢測方法無法考量 到人眼觀測時之反應，本發明即以模擬人眼追蹤方式拍攝 影像，並載入視覺模型，並分析動態影像品質，使能呈現 與人眼接近之判斷。 本發明所提供之應用於檢測顯示器動態影像色彩偏移 的檢測系統與方法，内容包含動態影像量測模擬人眼機制 拍攝、完整色彩特徵化轉換以及色彩偏移量化的功能，藉 以提供顯示器廠商能合理正確的分析顯示器動態影像品 質。本發明所提供之方法紀錄動態影像模糊區域所有色彩 數據，並繪成視覺化圖形，使用者可以直覺的看到色彩軌 跡的分佈狀況，而非只計算兩色色差。另外，更針對每套 顯示器以動態取像方式取得色彩影像，並且與標準色度量 測設備之量測值進行色彩特徵化轉換，如此可解決不同顯 示器不同色域與參考白問題。 本發明有以下優點： (1) 利用彩色攝影機取像，，免除人工i叙測的缺點。 (2) 利用追蹤式系統模擬人眼觀看動態影像機制拍 攝動態影像色彩偏移程度，並量化成參數，以作 為液晶顯示器效能之標準參考。 11 1321299 (3) 因此進 彩參數。 處理’可得到校正後的色 (4) 區域所有資訊，並在指定 觀察色;二;二透過此色彩軌跡直覺 例包較佳實施 =:中 段;動態取像進行色】校;追像定位手 行量測之色彩特徵化手段；最後進行】=色度空間中進 行被拍攝之影像亮度、色度偏移之分析之影=:段進 上述色彩特徵化手段更藉—進行色度校正的色卢二 m 控制#色攝影裳置調整拍攝角度，以進行拍攝。 拍攝機構對動態圖荦進行拍摄， 心、圖木，再以 ^影像，並進行色度校正與進行=:= 量析模組將影像轉換於-色度， ，發明所示之一種顯示器動態影像色彩偏移檢 以影像處理模組產生測試圖案，接著由拍攝機構進 丁 拍攝過程以模仿人眼視覺方式控制攝影裝置追蹤 12 拍攝顯示1¾ Φ μ闰安 β 圖案的畫素訊息，的色度值， =nr化結果’其實 ,谀查表，再進行影像分析步 王轉換矩 像说息轉換為色度空間的座標，最後產生一旦攝之移動影 於拍士述顯圖案之拍攝步驟更包括先校正=1 之色偏值。 ; 之初偵測移動之測試圖宰，攝機構，並 機進，影與，並於拍攝時進：::¾ 上述平均色偏值之計算步驟包括先 、運异子。 二該轉換運算子’再轉換拍攝之圖宰空間， 曲線間之差距計算該平均色偏值。 、_之色度 【實施方式】 本舍明為一種應用於檢測& ，系統與方法’包含動態影像量二模^彡目^彩偏移的 完整色彩特徵化娜以及色彩偏、、機制拍攝、 人眼，方式拍攝影像，並載入—視覺:，、’係以模擬 動態影像品質，使能呈現與人 、f進行分析 示器廠商能合理正確的分析顯示器=^:以提供顯 其中⑴猎模擬機制考量人貝 13 1321299 卻未能考量到义^ ’並裁來$明以模擬人眼追縱方 月匕呈現與人目艮接近之判斷；、:年斤動態影像品質，此較 ί化纪錄動態影像模糊區域:Γ呈現色彩偏移狀 而非只計算兩色色差 色彩執跡的分佈狀況， :針對每套顯示器以動態)取二 標準色度量柳設備之量測值二伸色彩影像，並且與 解決不同顯示器不同色域與袁;^徵化轉換，如此可 本發明並有以下優點·· 心 ⑴利用彩色攝影機取像，免除人^ ⑺利用追縱式系賴擬人_看=1 的缺點。 攝動態影像色彩偏移程度，機制拍 為液晶顯示器效能之標準參考。乡，以作 ⑶顯示器色域表現不同無法啸，因此進 彩參數。 了侍到权正後的色 W影像產生色偏區域所有資訊，在指定 間上逐點晝出，可透過此色彩執 硯祭色彩偏移的現象和數據。 本發明主要構成請參閱第三圖所示之顯哭 色彩偏移之檢測系統示意圖，架構分成四個主^ ^衫勺 括影像擷取料3卜影像^位單元33、色彩特徵^二二 characterization)單元35與影像分析單元37。其中影 像擷取單元31係為模擬人眼追蹤方式的拍攝系統，^括= 1321299 Γ如咖〇s感測元件)的攝_ 置，與控制其轉二:進行色度校正的色 係於以高解析度進行旦^"控制裝置等’影像定位單元33 定位尺標進行螢幕邊拍攝後’調整放大倍率’以邊界 線定義顯示器螢幕邊^位（可藉由特定圖案、線條或框 驟；色彩倾化單元祕影像追料確度定位等步 色後’再以-色度裝置二 態取像方式取得彩 或是其他色度校正裝置;度值，亚利用電腦系統 則進行動態色彩特徵又正’處理後之彩色影像 校正後的色彩參數，即色^ ^此特徵化處理可得到 色度偏移分析等動以；=影像之亮度分析、 度空間中晝出量測曲線： 第四A圖顯示本發明顯示 妾近之判斷。 針對液晶顯示器（LCD)、雷將顧—二，°〇 41，本發明係 =進行檢測，先由影像處理二= PDP)等平面顯示 :實施方式可以一電腦系統進行產生試圖案4 9，其 二:構進行該顯示器41中的公=49 ’再由拍 包括有彩色攝影裝置42、色戶/拍攝，拍攝機構 :旋轉控制裝置44，此拍攝二針3f=接攝影裝置42 進仃檢測。其中攝影裝置42 、同型怨的顯示器 =達）進行移動或旋轉動態取;寻：色裝置44 (如設置 ，所攝取的影像並非準確 j ’而利用拍攝機 置測能力的色度以43量測==財標準色度 〇 Y衫像，經之後的 15 1321299 特徵化步驟再轉換至標準色度空間。 之後，由一色彩特徵化模組45接收拍攝之影像與色度 量測值進行色彩特徵化的轉換，例如，當取得顯示中動態 影像的色彩值後’與^準的色度裂置（color meter )所量 測的色度值比對，進行影像校正，轉換後取得校正後之色 彩特徵值。 而影像分析模組46則接收上述色彩特徵化結果與拍 攝機構擷取的影像與色度訊息後，先選定一色度空間，將 所拍攝的影像座標轉換為色度空間座標，經過影像處理 組47產生量化的色偏值。 ' 上述色彩特徵化模組45、影像分析模組％ 理模組47之實施例可利用—電腦系統處理 ^ = 統接收拍攝機構所拍攝之影像訊息，刪徵化、2糸 經轉換為色度空間座標後產生量化 刀， 48。 色偏值’即檢測結果 動態拍攝測試圖案之實施例 影裝置42實施例。#顯示於顯 ^圖顯示之者 攝機構包括—影像_器，當_1 =t49，），該击 :旦偵測到移動晝面，即啟動旋二卜圖中) 42旋轉控制’追蹤該移 ；^:制裝置進行攝影裝】 驟亦包括-校正機制 二^伙。㈣攝影的多 订一對準兰 拍攝機構^ 直到| 驟，當對準失敗，番田攝影時，同時進子 1321299 弟五圖為弟四A圖顧-夕A ~ 檢測方=流程圖，流程包括_ =移彳 徵化與影像分析等步驟，對應第三 ^又位、色 開始時，影像處理模組產生列^檢夠系練〆 係•接至顯*器，於顯示器二影像處理 驟 置 :實施例更可由，接至顯示器=::，，； 木（步驟S5G1)’·接著由拍攝機構進行「產生剩、 拍攝過程以彩色攝影裳置追縱移動中的(步⑽5〇3)， =仿人眼視覺方式控制攝影裝置追縱^案，其中& 案’亦由標準的色度裝置取得標準的色度值^器中上圖 驟擷取測試圖案的畫素訊息（步驟S5〇5 ，，由此拍攝步 進行色度校正，產生標準的色度 亚鉍由色度妒 示器產生一轉換運算子，此轉換 ^，同顯 訊息與一標準色度空間之間轉換關係產生，=相 之晝素訊息轉換為標準色度空間，其實施=試圖案 陣（transf〇rmati〇n )或—轉、轉 (transformation i〇〇kup table)(步驟 S5〇9、、〜表 影像分析轉，此步_#上述職運算切^進行 動影像訊息轉換為色度空間的座標（步驟S5 = 多 ^色度"之訊息產生―平均色偏值’此量化的色偏值= 為本發明之重要目的之一（步驟S513)。 第/、圖所述為本發明之影像拍攝步驟，如上所述之拍 ，機，可先進行校正步驟（步驟S6G1)，其中方式包括_ d象疋位手·^又，如放大倍率調整（詳細揭露於本案第九 影像定位後，取得色彩特徵化“值再: 子係由該測試 之晝素訊息轉換為標準色度空間例試圖案 陣（transformation matHY、^ ^為轉換起 17 細揭露於本案第十⑴動嘴追縱準確度校正等步驟(詳 工作完成時上述測試圖案卿 中係於拍攝狀態下、，，=„像步驟（步驟觸），其 案，當有偵測到移動象伯測器债測該移動之測試圖 影裝置進行圖案的拍J時；即驅Ϊ旋轉控魏置控制攝 影裳置進行追職A / H即11旋轉控制裝置控制攝 否拍攝成功士 i衫（步驟S605);拍攝完畢後需判斷是 作；==到影像訊息、彩色攝影 補償或是修正拍攝:=S6:7);若拍攝失敗，先進行 新、隹一站儀^ . ’構（，驟S609)，並回到步驟S6〇3重 功擷二…：移ί影像偵測、追蹤攝影等步驟;若成 驟別11) 訊息’表示拍攝成功，_攝結束（步 而第七關顯*本發明影像娜與色料徵化步驟之 細郎流程’其最終藉—轉換運算子進行色度空間之轉換。 步驟開始’先產生待測色塊，較佳實施例係由上述影 像，理模組或圖案產生器產生此待測色塊（步驟s7〇i)， 接著由上述拍攝機構拍攝移動之待測色塊，由彩色攝影裝 置中之影像感測元件擷取待測色塊之色彩拍攝訊息（步驟 S703) ’並由色度裝置擷取之其中色度值（步驟S7〇5)後， 由擷取之晝面色度平均值，在此步财若被拍攝測試 圖案為一二維圖案，由此產生的量測色塊亦為一二維圖 案，藉此汁昇平均值之步驟產生一維的色度平均值（步驟 S707)。值得一提的是，顯示器上水平以矩陣型式顯示的像 素間隙（black matrix)應會影響其中亮度之量測，但是， 18 1321299 當拍攝機構利用追縱方式拍攝水平移動之測試圖案時，顯 示器上水平顯示的像素間隙會被模糊掉而不易顯出，而垂 直方向的像素間隙則可藉此平均值步驟而忽略不計。 之後，藉色彩特徵化模組由上述擷取之色度值進行色 度才父正，付出標準色度值（步驟S709)，再藉以產生一轉 換查表（lookup table)(步驟S711)，此轉換查表係紀錄 各被拍攝之晝素訊息（如各晝素之紅色r、綠色G、藍色B 三色值）與其標準色度值。接著，得出一轉換運算子（步 驟S713)’此轉換運异子可將經調整後之該測試圖案訊息 (如、’、工色R綠色G、監色B二色值）經座標轉換為某選定 之色度空間的座標（如座標X、y值）。 第八圖所不為上述第五圖中所示之影像分析步驟之流 粒圖。传出上述轉換運算子之後，即需依據待測顯示器選 擇一適當色度空間（步驟S8()1)，接著引入上述經特徵化 付出的轉換運算子（步驟S8G3)，並藉轉換運算子將已緩 案訊ΐ進行色度空間轉換為一色度空間座標 电二 )’接著描繪經轉換後的色度空間座標，如本 Β圖所示之曲線（步驟s謝），由轉換後的色度 ，不標準（沒有色幻的色度座標值作比較，計瞀 平均色偏值（步驟湖），此平均色偏值即為—量化的: 偏值。 敫·^述有關拍攝機構的影像定位手段中之一放大倍率調 ^手段放大倍率之調整係調整影像感測元件與被拍攝之 ϋιχει)比例’如調整為u晝素）：ι〇(感光單元）， 貝1、不以10個影像感測元件中10個感光單元拍攝一個晝 19 rI321299 素，以此增加檢測的解析度。 如第九圖所示之較佳實施例，顯示器91晝面中顯示一 可調整比例之十字框97，並於四周有多個方形框95。此實 施例係利用顯示器91中晝面上十字框97來決定放大倍 率，先調整十字框97之涵蓋範圍，再改變顯示器91與攝 影裝置（未顯示於本圖中）之距離，或者改變影像感測元 件與顯示器91之距離，使該攝影裝置所拍攝的範圍（如圖 中攝影裝置涵蓋範圍93)填滿該十字框97,即達到所需之 放大倍率。 第十圖則顯示拍攝機構在影像定位手段中之動態影像 追縱準確度定位手段實施例不意圖’當拍攝機構追縱榮幕 中以固定速度移動之測試圖案，動態影像追蹤準確度定位 係用來確認拍攝到之動態影像均在同一個位置，其中旋轉 控制裝置會自動修正起始角度，以進行下一次拍攝，直到 影像追蹤誤差在某適當移動距離之内。其實施例如圖式， 由旋轉控制裝置控制的攝影裝置可自動調整追蹤角度，直 到每次正中間的拍攝能拍到顯示器91中模糊區域101，則 完成此動態影像追蹤準確度定位。 第十一 A圖所示為顯示器91上顯示兩種色塊移動時邊 緣色偏程度之轉變，如前景與背景之轉換，而本發明之檢 測目的係為檢測出於轉變時的色偏程度。如圖中所示之實 施例，左方的第一顏色區域111為前景顏色，右方的第二 顏色區域112表示為背景顏色，當拍攝移動之測試圖案 時，前景與背景顏色之轉變將可能會產生色偏。 第十一 B圖所示則為上述步驟S807中所述描繪經轉換 20 1321299 後的色度空間座標實施例，如第十一A圖所示由第一顏色 區域111轉變為第二顏色區域112時，圖中於色度座標中 的理想參考色度曲線113表示一沒有產生色偏的理想曲 線，與實際量測之色度曲線114顯示一差距△，兩者曲線 有複數個差距△，每一個差距△表示各量測值的色偏值， 其平均值（如步驟S809)顯示為整體的色偏值，為本發明 檢測方法所產生量化的色偏值。 所附圖式僅提供參考與說明用，並非用來對本發明加 以限制者。 綜上所述，本發明為一顯示器動態影像色彩偏移檢測 系統與檢測方法，係以動態影像量測模擬人眼機制拍攝一 移動之測試圖案，並進行色彩特徵化轉換以及色彩偏移量 化的功能，可提供顯示器廠商能合理正確的分析顯示器動 態影像品質並當作分級與銷售依據。 惟以上所述僅為本發明之較佳可行實施例，非因此即 拘限本發明之專利範圍，故舉凡運用本發明說明書及圖示 内容所為之等效結構變化，均同理包含於本發明之範圍 内，合予陳明。 【圖式簡單說明】 第一圖係為習用技術動態影像顯示評價裝置示意圖； 第二A圖與第二B圖係為習用技術xy色度空間A線 圖； 第三圖係為本發明顯示器動態影像色彩偏移之檢測系 統架構圖； 21 1321299 第四A圖係為本發明顯示器動態影像色彩偏移之檢測 系統裝置示意圖； 第四B圖係為攝影裝置之運作實施例示意圖； 第五圖係為本發明色彩偏移檢測方法流程圖； 第六圖係為本發明檢測方法中影像拍攝步驟之流程 圖； 第七圖係為本發明檢測方法中色彩特徵化步驟之流程 圖； 第八圖係為本發明檢測方法中影像分析步驟之流程 圖； 第九圖係為本發明拍攝機構的影像定位手段中之放大 倍率調整手段實施例示意圖； 第十圖係為本發明拍攝機構的影像定位手段中之動態 影像追蹤準確度定位手段實施例示意圖； 第十一 A圖所示係為顯示器中色彩變化示意圖； 第十一 B圖係為色度空間座標示意圖。 【主要元件符號說明】 攝影機3 鏡子2 螢幕5 驅動控制器7 電腦糸統6 影像訊號產生器9 22 1321299 曲線區域20 顯示器41 攝影裝置42 色度裝置43 旋轉控制裝置44 色彩特徵化模組45 影像分析模組46 影像處理模組47 檢測結果4 8 測試圖案49, 49’ 顯示器91 攝影裝置涵蓋範圍93 十字框97 方形框95 模糊區域101 第一顏色區域111 第二顏色區域112 理想參考色度曲線113 量測之色度曲線114 23SYSTEM FOR EVALUATING MOVING IMAGE QUALITY OF DISPLAYS ) ' This case uses the optical debt detector (ph〇t〇detect〇r) to detect motion, and the 4 image position is still limited to the boundary blur determination of black and white images, and it is not It can evaluate the quality of color images, but it is not accurate enough in color and brightness conversion, and there is no information on chroma (chr_). 9 1321299 There is a method for dynamic quality measurement of color images using a tracking camera system by Yoshi Enami, Keiko Kitagishi and Koichi Oka (Otsuka Electronics Co. LTD), which uses a color camera to white-view the moving image in the display. The fuzzy state due to the movement is classified and classified into an xy Chromaticity space. The measurement results of the embodiment as shown in FIG. 2A are the Luminance and time curves of the kneading surface taken from the color camera in the Xy chromaticity space when the color changes from white to black, wherein The captured red R, green G, and blue B values are converted to Tristimulus Values X, γ, and Z, which are converted to the chromaticity coordinates (x, y) by the energy of the reflectance source of the object. As shown in the second A picture, the red R, green G, and blue B have the same rise timing of the luminance change when switching in black and white, and the chromaticity coordinates X and y (the horizontal line) are unchanged; The second B-picture shows another "experimental result" in which the chromaticity coordinates X and y values show a convex curve region 20 and an offset caused by the change in luminance, that is, an area indicating color blur in the display. However, since the human eye vision is not a uniform chromaticity space, the measurement result expressed by the xy chromaticity space cannot interpolate the concept of the chromatic aberration distance. For example, South Korea's Samsung Electronics issued the "Evaluation of Motion Artifacts and Evolving Compensation Techniques for LCD Monitors". The motion artifacts are residual when moving under the surface. False shadows have a crucial impact on the quality of liquid crystal display. This paper classifies and analyzes various artifacts, and compensates them by software means after individual judgment. However, only the main 1321299 color is calculated for the judgment of display 瑕疵. Chromatic aberration, failed to describe the performance of edge color blur. The above-mentioned conventional measurement method measures the reaction time of the liquid crystal display with a photodiode or the like, but since the detection parameters of the human eye are not introduced, the detection result cannot truly reflect the human eye reaction. SUMMARY OF THE INVENTION In view of the fact that the detection method of the liquid crystal display cannot be considered to reflect the reaction of the human eye, the present invention captures an image by a simulated human eye tracking method, loads a visual model, and analyzes the dynamic image quality. Can present judgments close to the human eye. The invention provides a detection system and method for detecting color shift of a dynamic image of a display, which comprises the functions of dynamic image measurement, human eye mechanism shooting, complete color characterization conversion and color shift quantization, thereby providing display manufacturers capable of Reasonable and correct analysis of monitor dynamic image quality. The method provided by the invention records all the color data of the blurred area of the moving image and draws the visualized figure, and the user can intuitively see the distribution of the color track instead of calculating only the two color difference. In addition, the color image is obtained by dynamic image capture for each set of displays, and color characterization is performed with the measured values of the standard color measurement device, so that different color gamuts and reference white problems of different displays can be solved. The invention has the following advantages: (1) Using a color camera to take images, thereby eliminating the disadvantages of manual i-description. (2) Using the tracking system to simulate the human eye to view the motion picture mechanism to capture the degree of color shift of the motion picture and quantize it into parameters as a standard reference for the performance of the liquid crystal display. 11 1321299 (3) Therefore, the color parameters are entered. Processing 'can get all the information of the corrected color (4) area, and specify the observation color; 2; 2 through the color trajectory intuition package better implementation =: middle section; dynamic image acquisition color] school; chase positioning hand Color characterization method of line measurement; finally] 】 分析 空间 空间 进行 分析 分析 分析 = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = Lu Erm Control #色摄影裳 Adjust the shooting angle for shooting. The shooting mechanism shoots the dynamic image, the heart, the figure, and then the image, and performs the chromaticity correction and the =:= measurement module converts the image to the chromaticity, and the invention displays a display motion image. The color shift detection uses the image processing module to generate a test pattern, and then the photographing mechanism controls the photographic device to control the photographic device to track the chrominance value of the pixel information of the 13⁄4 Φ μ闰β pattern. , =nrization result 'In fact, check the table, and then perform image analysis. The step-by-step transformation of the moment image is converted into the coordinates of the chromaticity space. Finally, the shooting step of the motion of the film is reflected in the filming step. Corrected the color shift value of = 1. At the beginning of the detection of the mobile test chart, camera, and machine, shadow and, and when shooting::: 3⁄4 The above calculation of the average color shift value includes the first, the transport of the son. Second, the conversion operator 're-converts the captured image space, and the difference between the curves calculates the average color shift value. The chromaticity of _ [Implementation] This is a kind of application for detection & system and method 'including dynamic image quantity two-mode ^ 彡 eye color offset full color characterization Na and color bias, mechanism shooting , human eyes, way to shoot images, and load - visual:,, 'to simulate the dynamic image quality, enable rendering and human, f to analyze the display manufacturers can reasonably and correctly analyze the display = ^: to provide the display (1) The hunting simulation mechanism considers the people 13 1321299 but fails to consider the meaning of ^ 'and cuts the $ Ming to simulate the human eye to trace the 匕 匕 匕 匕 匕 匕 匕 匕 匕 匕 匕 ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; Recording the blurred area of the moving image: Γ Presenting the color offset instead of only calculating the distribution of the color chromatic aberration of the two colors, : for each set of displays, dynamically taking the measured value of the two standard color measuring device, And to solve the different color gamut and Yuan of different monitors; ^ ̄ ̄ conversion, so the invention has the following advantages · · Heart (1) using color camera to take images, exempt people ^ (7) using the squatting type of avatar Disadvantage . The degree of color shift of the motion picture is taken, and the mechanism is taken as a standard reference for the performance of the liquid crystal display. Township, for the work (3) display color gamut performance can not be whistle, so enter the color parameters. The color W image after the right of the waiter produces all the information in the color-shifted area, and it is displayed point by point in the specified interval. This color can be used to perform the phenomenon and data of the color shift. The main structure of the present invention, please refer to the schematic diagram of the detection system of the crying color shift shown in the third figure, the structure is divided into four main ^ ^ 勺 勺 撷 撷 撷 撷 3 3 3 3 3 ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ 、 The unit 35 and the image analyzing unit 37. The image capturing unit 31 is a shooting system that simulates a human eye tracking mode, and includes a photo-control system that controls the rotation of the lens. High-resolution processing is performed on the image positioning unit 33. The positioning scale is used to perform the screen-side shooting and the 'adjustment magnification' is defined by the boundary line (by a specific pattern, line or frame); Color tilting unit secret image tracking accuracy positioning and other color steps, then use the - chroma device two-state image capture method to obtain color or other color correction device; degree value, the use of computer systems for dynamic color features and positive 'The color parameters after the color image correction after processing, that is, the color ^ ^ This characterization process can obtain the chromaticity shift analysis and the like; = the brightness analysis of the image, the measurement curve in the degree space: the fourth A picture The display of the present invention shows the proximity judgment. For the liquid crystal display (LCD), Lei Jiang Gu - two, ° 〇 41, the present invention = detection, first by image processing two = PDP) and other plane display: the implementation can be a computer system The test pattern 4 is generated in a row, and the second is configured to perform the public=49' in the display 41. The photographer includes a color photographing device 42, a color household/photographing, and a photographing mechanism: a rotation control device 44, which photographs two stitches 3f= The photographing device 42 is connected to the detection. The photographic device 42 and the display of the same type of blame = up) are moved or rotated dynamically; the finder color device 44 (if set, the captured image is not accurate j' and the chromaticity of the camera's ability to be measured is measured by 43 == financial standard color 〇Y shirt image, after the subsequent 15 1321299 characterization step and then converted to the standard chromaticity space. Then, a color characterization module 45 receives the captured image and color metric value for color characteristics The conversion is performed, for example, after obtaining the color value of the moving image in the display, and comparing the chromaticity values measured by the color meter with the color, the image is corrected, and the corrected color is obtained after the conversion. The image analysis module 46 receives the color characterization result and the image and chrominance information captured by the shooting mechanism, and then selects a chromaticity space to convert the captured image coordinates into chromaticity space coordinates, and passes the image. The processing group 47 generates quantized color shift values. The embodiments of the above-described color characterization module 45 and image analysis module % management module 47 can be utilized by a computer system to process the system. The captured image message is deleted, and the quantized knife is generated after being converted into a chromaticity space coordinate, 48. The color shift value is the embodiment of the image capturing device 42 of the detection result dynamic shooting test pattern. #显示在显^ The figure shows that the camera mechanism includes - image_device, when _1 = t49,), the hit: once the mobile surface is detected, that is, the rotation is controlled. 42 rotation control 'tracking the movement; ^: system The device is also equipped with a photography device. (4) Multi-order photography for the camera. ^ Until | _ = moves and image analysis steps, corresponding to the third ^ position, when the color starts, the image processing module generates a column to check the system, and then connects to the display device. : The embodiment can be connected to the display =::,,; wood (step S5G1)'. Then, the shooting mechanism performs "there is a surplus, and the shooting process is followed by a color photography (step (10) 5〇3). = The human eye visual mode controls the photographic device to track the 縱^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ 5, the chromaticity correction is performed by the shooting step, and a standard chromaticity 铋 is generated by the chromaticity indicator to generate a conversion operator, and the conversion ^, the conversion relationship between the same message and a standard chromaticity space is generated, = phase metamorphic message is converted to standard chromaticity space, its implementation = attempt The file array (transf〇rmati〇n) or - transformation, transfer (transformation i〇〇kup table) (step S5〇9, ~ table image analysis turn, this step _# the above job calculation cut ^ motion image information converted to The coordinates of the chromaticity space (step S5 = multi-color chromaticity " message generation - average color shift value 'this quantized color shift value = one of the important purposes of the present invention (step S513). For the image capturing step of the present invention, as described above, the machine may first perform a correcting step (step S6G1), wherein the method includes _d symbolizing the position of the hand, and further, such as magnification adjustment (detailed in the ninth of the case) After the image is positioned, the color characterization is obtained. The value is further converted into a standard chromaticity space test pattern matrix (transformation matHY, ^^ is converted from the finest disclosure in the tenth (1) of the case. Vertical accuracy correction and other steps (the above test pattern is in the shooting state when the work is completed, =, like the step (step touch), the case, when the movement is detected, the movement is measured by the detector The test picture device performs the pattern shot J That is, the driving control, the rotating control, the control, the shooting, the tracking, the A/H, the 11 rotation control device, the shooting, the shooting, the successful shooting (step S605); after the shooting is completed, the judgment is made; == to the image message, Color photography compensation or correction shooting: =S6:7); If the shooting fails, first perform a new, one-stop station ^. 'Construction (, step S609), and return to step S6〇3 heavy work two...: shift ίImage detection, tracking photography, etc.; if it is a step 11) The message ' indicates that the shooting is successful, _ photo ends (step and the seventh key display * the invention of the image and the color of the smear step of the process of the lang process' its final The conversion-operator converts the chroma space. The step starts to generate the color block to be tested first. In the preferred embodiment, the image block to be tested is generated by the image, the rational module or the pattern generator (step s7〇i), and then the moving color is detected by the shooting mechanism. Block, the color capture information of the color block to be detected is captured by the image sensing component in the color photographing device (step S703)' and the chroma value is extracted by the chroma device (step S7〇5), The mean value of the chromaticity of the face, if the test pattern is a two-dimensional pattern, the resulting color patch is also a two-dimensional pattern, and the step of averaging the juice produces a one-dimensional chromaticity. The average value (step S707). It is worth mentioning that the black matrix displayed horizontally on the display should affect the measurement of the brightness. However, 18 1321299 When the shooting mechanism uses the tracking method to shoot the horizontally moving test pattern, the display is on the display. The pixel gap displayed horizontally is blurred and is not easily displayed, and the pixel gap in the vertical direction can be neglected by the average step. Then, the color characterization module performs the chrominance on the chromaticity value of the captured color, and pays the standard chromaticity value (step S709), and then generates a lookup table (step S711). The conversion look-up table records the individual information (such as the red, green, and blue B tristimulus values of each element) and its standard chromaticity values. Then, a conversion operator is obtained (step S713). The converted test element can convert the adjusted test pattern information (eg, ', work color R green G, monitor color B two-color value) into coordinates to be converted into coordinates. The coordinates of a selected chromaticity space (such as coordinates X, y). The eighth figure is not a flow chart of the image analysis step shown in the fifth figure above. After the above-mentioned conversion operator is transmitted, an appropriate chromaticity space is selected according to the display to be tested (step S8()1), and then the above-described characterization-paid conversion operator is introduced (step S8G3), and the conversion operator is used. The mitigation space is converted to a chromaticity space coordinate. 2) Then the converted chromaticity space coordinates are plotted, as shown in the figure (step s thanks), from the converted chromaticity , not standard (no color illusion chromaticity coordinate value for comparison, calculate the average color shift value (step lake), the average color shift value is - quantized: partial value. 敫·^ describe the image location of the shooting mechanism One of the means of magnification adjustment means the adjustment of the magnification is to adjust the ratio of the image sensing element to the captured ϋιχει" as adjusted to u 昼 )): ι〇 (photosensitive unit), shell 1, not 10 images Ten sensing units in the sensing element take a 昼19 rI321299 element to increase the resolution of the detection. As shown in the preferred embodiment of the ninth embodiment, an adjustable ratio cross frame 97 is displayed in the face of the display 91, and a plurality of square frames 95 are provided around the periphery. In this embodiment, the cross frame 97 on the upper surface of the display 91 is used to determine the magnification. First, the coverage of the cross frame 97 is adjusted, and the distance between the display 91 and the photographing device (not shown in the figure) is changed, or the image sense is changed. The distance between the measuring component and the display 91 is such that the range photographed by the photographing device (the photographing device covers the range 93 in the figure) fills the cross frame 97 to achieve the desired magnification. The tenth figure shows that the moving image tracking accuracy positioning means of the shooting mechanism in the image positioning means is not intended to be used when the shooting mechanism tracks the test pattern moving at a fixed speed in the honor screen, and the dynamic image tracking accuracy positioning system is used. To confirm that the captured motion pictures are in the same position, the rotation control will automatically correct the starting angle for the next shot until the image tracking error is within an appropriate moving distance. For example, the photographing device controlled by the rotation control device can automatically adjust the tracking angle until the blurring region 101 in the display 91 can be photographed every time the middle photograph is taken, and the dynamic image tracking accuracy positioning is completed. Fig. 11A shows the transition of the degree of edge color shift when the two color patches move, such as the transition between foreground and background, on the display 91, and the detection purpose of the present invention is to detect the degree of color shift due to the transition. In the embodiment shown in the figure, the first color area 111 on the left is the foreground color, and the second color area 112 on the right is the background color. When the test pattern is moved, the transition between the foreground and the background color may be Will produce color cast. FIG. 11B is a chromaticity space coordinate embodiment after the conversion 20 1321299 is described in the above step S807, and is transformed from the first color region 111 to the second color region 112 as shown in FIG. 11A. When the ideal reference chromaticity curve 113 in the chromaticity coordinates in the figure represents an ideal curve without color shift, the chromaticity curve 114 of the actual measurement shows a difference Δ, and the two curves have a plurality of Δ, each A gap Δ indicates the color shift value of each measured value, and the average value (such as step S809) is displayed as the overall color shift value, which is the quantized color shift value generated by the detecting method of the present invention. The drawings are provided for reference and description only and are not intended to limit the invention. In summary, the present invention is a display dynamic image color shift detection system and detection method, which adopts a dynamic image measurement simulation human eye mechanism to take a moving test pattern, and performs color characterization conversion and color shift quantization. The function can provide the display manufacturer with reasonable and correct analysis of the dynamic image quality of the display and as a basis for grading and sales. However, the above description is only a preferred embodiment of the present invention, and thus the scope of the present invention is not limited thereto, and equivalent structural changes made by using the present specification and the illustrated contents are equally included in the present invention. Within the scope of the agreement, Chen Ming. [Simple diagram of the diagram] The first diagram is a schematic diagram of a conventional dynamic image display evaluation device; the second diagram A and the second diagram B are conventional techniques xy chromaticity space A-line diagram; the third diagram is the display dynamics of the present invention Image color shift detection system architecture diagram; 21 1321299 The fourth diagram is a schematic diagram of the detection system of the dynamic image color shift of the display of the invention; the fourth diagram B is a schematic diagram of the operation example of the photographic apparatus; The flowchart of the color shift detection method of the present invention is a flowchart of the image capturing step in the detecting method of the present invention; and the seventh drawing is a flowchart of the color characterization step in the detecting method of the present invention; The flowchart of the image analysis step in the detection method of the present invention; the ninth figure is a schematic diagram of the embodiment of the magnification adjustment means in the image localization means of the photographing mechanism of the present invention; the tenth figure is the image localization means of the photographing mechanism of the present invention A schematic diagram of an embodiment of a dynamic image tracking accuracy positioning method; FIG. 11A is a schematic diagram of a color change in a display; B is a color space coordinate system in FIG. FIG. [Description of main component symbols] Camera 3 Mirror 2 Screen 5 Drive controller 7 Computer system 6 Video signal generator 9 22 1321299 Curve area 20 Display 41 Photographing device 42 Color device 43 Rotation control device 44 Color characterization module 45 Image Analysis module 46 Image processing module 47 Detection result 4 8 Test pattern 49, 49' Display 91 Photographing device coverage 93 Cross frame 97 Square frame 95 Blur area 101 First color area 111 Second color area 112 Ideal reference color curve 113 Measurement of chromaticity curve 114 23