TWI294616B

TWI294616B - Gamut conversion system and methods

Info

Publication number: TWI294616B
Application number: TW93131633A
Authority: TW
Inventors: Michael Francis Higgins
Original assignee: Clairvoyante Inc
Priority date: 2003-10-21
Filing date: 2004-10-19
Publication date: 2008-03-11
Also published as: TW200521955A; WO2005043507A1

Description

1294616 九、發明說明：【發明所屬之技術領域】本發明係有關一種液晶顯示器，更特別有關色域轉換系統及方法。【先前技術】在這些申請人共有的美國專利申請案中··（1)於2001年 7月25曰申請之美國專利申請第09/91 6,232號申請案，其標題爲「用於具有簡化位址之全彩影像元件之色彩像素配置」（ARRANGEMENT OF COLOR PIXELS FOR FULL COLOR IMAGING DEVICES WITH SIMPLIFIED ADDRESSING); (2)於 20 02年10月22曰申請之美國專利申請第1 0/278, 353號申請案，其標題爲「用於具有遞增調變轉換函數響應之次像素著色之色彩平面顯示器次像素配置與佈局之改進」1294616 IX. DESCRIPTION OF THE INVENTION: TECHNICAL FIELD OF THE INVENTION The present invention relates to a liquid crystal display, and more particularly to a color gamut conversion system and method. [Prior Art] In the U.S. Patent Application Serial No. 09/91, 232, filed on July 25, 2001, the entire disclosure of which is incorporated herein by reference. (ARRANGEMENT OF COLOR PIXELS FOR FULL COLOR IMAGING DEVICES WITH SIMPLIFIED ADDRESSING); (2) US Patent Application No. 1/0,278, 353, filed on October 22, 2002 The application, titled "Improvements in Subpixel Configuration and Layout for Color Planar Display with Subpixel Pixeling with Incremental Modulation Transition Function Response"

(IMPROVEMENTS TO COLOR FLAT PANEL DISPLAY SUB-PIXEL ARRANGEMENTS AND LAYOUTS FOR SUB-PIXEL RENDERING WITH INCREASED MODULATION TRANSFER FUNCTION RESPONSE)； (3)於20 02年10月22曰申請之美國專利申請第 1 0/278, 352號申請案，其標題爲「用於具有***藍次像素之次像素著色之色彩平面顯示器次像素配置與佈局之改進」（IMPROVEMENTS TO COLOR FLAT DISPLAY SUB-PIXEL ARRANGEMENTS AND LAYOUTS FOR SUB-PIXEL RENDERING WITH SPLIT BLUE SUB-PIXELS) ; (4)於 2002 年 9 月 13 日申請之美國專利申請第1 0/243, 094號申請案，其標題爲「用於次像素著色之改良型四色配置與發射器」（IMPROVED FOUR 6 1294616 COLOR ARRANGEMENTS AND EMITTERS FOR SUB-PIXEL RENDERING) ; (5)於2002年10月22日申請之美國專利申請第1 0/278, 328號申請案，其標題爲「減少藍色亮度而有良好能見度之色彩平面顯示器次像素配置與佈局之改進」 (IMPROVEMENTS TO COLOR FLAT PANEL DISPLAY SUB-PIXEL ARRANGEMENTS AND LAYOUTS WITH REDUCED BLUE LUMINANCE WELL VISIBILITY) ; (6)於2002年10月22日申請之美國專利申請第1 0/278,393號申請案，其標題爲「具有水平次像素配置與佈局之色彩顯示器」（COLOR DISPLAY HAVING HORIZONTAL SUB-PIXEL ARRANGEMENTS AND LAYOUTS);及 (7)於2003年1月16日申請之美國專利申請第01/347, 001 號申請案，其標題爲「改良型條紋顯示器次像素配置及其次像素著色用之系統及方法」（IMPROVED SUB-PIXEL ARRANGEMENTS FOR STRIPED DISPLAYS AND METHODS AND SYSTEMS FOR SUB-PIXEL RENDERING SAME)，其揭示了一些新穎、用來改善一些影像顯示設備成本/性能曲線的次像素配置，皆於此併入本文參考。對於某些沿一水平方向上具有偶數個子像素的子像素重複群，下面可影響適當的點反轉策略的系統與技術被揭示，且皆於此併入本文參考：（1)美國專利申請第 1 0/456, 839號申請案，其標題爲「新穎液蟲顯示器之影像劣化修正」（IMAGE DEGRADATION CORRECTION IN NOVEL LIQUID CRYSTAL DISPLAYS ) ; (2)美國專利申請第(IMPROVEMENTS TO COLOR FLAT PANEL DISPLAY SUB-PIXEL ARRANGEMENTS AND LAYOUTS FOR SUB-PIXEL RENDERING WITH INCREASED MODULATION TRANSFER FUNCTION RESPONSE); (3) US Patent Application No. 1 0/278, 352, filed on October 22, 2002 The application title is "Improvement of sub-pixel configuration and layout for color flat-panel display with sub-pixel rendering with split blue sub-pixels" (IMPROVEMENTS TO COLOR FLAT DISPLAY SUB-PIXEL ARRANGEMENTS AND LAYOUTS FOR SUB-PIXEL RENDERING WITH SPLIT BLUE SUB-PIXELS); (4) U.S. Patent Application Serial No. 10/243,094, filed on Sep. 13, 2002, entitled "Improved Four-Color Configuration and Transmitter for Sub-pixel Coloring" (IMPROVED FOUR 6 1294616 COLOR ARRANGEMENTS AND EMITTERS FOR SUB-PIXEL RENDERING); (5) US Patent Application No. 1/0,278, filed on October 22, 2002, entitled "Reducing Blue Brightness" IMPROVEMENTS TO COLOR FLAT PANEL DISPLAY SUB-PIXEL ARRANGEMENT S & LAYOUTS WITH REDUCED BLUE LUMINANCE WELL VISIBILITY); (6) U.S. Patent Application Serial No. 10/278,393, filed on Oct. 22, 2002, entitled <RTIgt; (COLOR DISPLAY HAVING HORIZONTAL SUB-PIXEL ARRANGEMENTS AND LAYOUTS); and (7) U.S. Patent Application Serial No. 01/347,001, filed on Jan. 16, 2003, entitled "Improved Stripe Display Sub-Pixel Configuration" And its sub-pixel coloring system and method" (IMPROVED SUB-PIXEL ARRANGEMENTS FOR STRIPED DISPLAYS AND METHODS AND SYSTEMS FOR SUB-PIXEL RENDERING SAME), which reveals some novelities to improve the cost/performance curve of some image display devices. Pixel configurations are incorporated herein by reference. For some sub-pixel repeating groups with an even number of sub-pixels in a horizontal direction, the following systems and techniques that can affect the appropriate dot inversion strategy are disclosed herein and are incorporated herein by reference: (1) U.S. Patent Application Serial No. 1 0/456, application No. 839, entitled "IMAGE DEGRADATION CORRECTION IN NOVEL LIQUID CRYSTAL DISPLAYS"; (2) US Patent Application No.

10/455, 925號申請案，其標題爲「具有促成點反轉交叉連接之顯示面板」（DISPLAY PANEL HAVING CROSSOVER 1294616 • CONNECTIONS EFFECTING DOT INVERSION) ; (3)美國專利申請第1 0/455, 931號申請案，其標題爲「於新穎顯示面板配置上執行具標準驅動及背板之點反轉系統及方法」（SYSTEM AND METHOD OF PERFORMING DOT INVERSION WITH STANDARD DRIVERS AND BACKPLANE ON NOVEL DISPLAY PANEL LAYOUTS); (4)美國專利申請第10/455,927號申請案，其標題爲「於具有減少量子化誤差固定形式雜訊面板之視覺效果補償系統及方法」（SYSTEM AND METHOD FOR COMPENSATING FOR VISUAL EFFECTS UPON PANELS HAVING FIXED PATTERN NOISE WITH REDUCED QUANTIZATION ERROR) ; (5)美國專利申請第10/456,806號申請案，其標題爲「具額外驅動器之新穎面板配置之點反轉」（DOT INVERSION ON NOVEL DISPLAY PANEL LAYOUTS WITH EXTRA DRIVERS);及（6)美國專利申請第1 0/456, 838號申請案，其標題爲「液晶顯示器背板設計及非標準次像素配置之位址」（LIQUID CRYSTAL DISPLAY BACKPLANE LAYOUTS AND ADDRESSING FOR NON-STANDARD SUBPIXEL ARRANGEMENTS)。上述申請案，皆於此併入.本案參考。當與上述專利申請案所揭示之技術相配合，下面一些申請人共有之美國專利申請案進一步揭示的一些次像素著色系統及方法，這些改進特別顯著··（1)於2002年1月16日申請之美國專利申請第1〇/〇51，612號申請案，其標題爲「紅綠藍像素格式數據轉換成波形瓦式矩陣次像素數據格式」 (CONVERSION OF RGB PIXEL FORMAT DATA TO PENTILE MATRIX SUB-PIXEL DATA FORMAT); (2)於 2002 年 5 月 17 1294616 曰申請之美國專利申請第1 0/1 50, 355號申請案，其標題爲「具有影像灰度調整之次像素著色用之系統及方法」 (METHODS AND SYSTEMS FOR SUB-PIXEL RENDERING WITH GAMMA ADJUSTMENT) ; (3)於2002年8月8日申請之美國專利申請第1 0/215, 843號申請案，其標題爲「具有自適應濾光之次像素著色用之系統及方法」（METHODS AND SYSTEMS FOR SUBPIXEL RENDERING WITH ADAPTIVE FILTERING) ； (4) 於2003年3月4日申請之美國專利申請第10/379,767號申請案，其標題爲「影像數據時態次像素著色用之系統及方法」（SYSTEMS AND METHODS FOR TEMPORAL SUB-PIXEL RENDERING OF IMAGE DATA) ; (5)於 2003 年 3 月 4 日申請之美國專利申請第1 0/379, 765號申請案，其標題爲「用於運動自適應濾光之系統及方法」（SYSTEMS AND METHODS FOR MOTION ADAPTIVE FILTERING) ; (6)於 2003 年 3 月 4 日申請之美國專利申請第10/379, 766號申請案，其標題爲「用於改良型顯示視角之次像素著色系統及方法」（SUB-PIXEL RENDERING SYSTEM AND METHOD FOR IMPROVED DISPLAY VIEWING ANGLES);及（7)於20 03牟4月7日申請之美國專利申請第1 0/409, 41 3號申請案，其標題爲「具有嵌入式預先次像素著色影像之影像數據集」（IMAGE DATA SET WITH EMBEDDED PRE-SUBPIXEL RENDERED IMAGE)。以上所述之申請案，皆於此併入本文參考。【發明内容】本發明係有關一種液晶顯示器，更特別有關色域轉換系 516 統及方法。本發明揭示了產生原始影像數據到標的影像數據集的、或轉換之系統及方法’其用以在可能不同色域的設備上_ 、行著色 #色域轉換系統，其包含：一輸入通道裝置、， -色，單元、一色度/亮度單元、一色調角計算器與一色域轉換單元」亥輪入通道裝置接收原始影像數據。該色域單 =將^始影像數據轉換成知覺上均句空間的數據。該色度/ ^度單元將知覺上均勻空間的數據轉換成-種包含一些色 · 度與免度分量的格式。該色調角計算器從色度/亮度單元至 ^接收㉟色度> 1 ’並根據所接收到白勺-些分量對原始衫像數據計算-些色調角。該色域轉換單元導出一些色域轉換值’利用色域轉換表將這些值應用到原始影像數據的一些分量上。【實施方式】於此將可詳細地參考一些具體實施方案與實施例，其範例將於附圖中描述之。於本文的額内，將盡可能在所冑 i 附圖中採用相同的參考標號，以標示相同或類似的元件。現今絕大多數監視器與電視設備，設計成顯示三數值色彩數據（three-valued color data )，例如 RGB ( RGB 二 RED、 GREEN、RED ’ 紅、綠、藍）及 /或 sRgb ( s = Spaee，空間紅綠藍）’或三數值色度/亮度（three_valued cluoma/luminance)信號，饷如YIq或YCbCr信號。術語 sRGB用來敍述非線性系統，絕大多數個人電腦影像均採用該非線性系統加以儲存，這有時也被稱爲R，G，B，或非線性 1294616Application No. 10/455, No. 925, entitled "DISPLAY PANEL HAVING CROSSOVER 1294616 • CONNECTIONS EFFECTING DOT INVERSION"; (3) US Patent Application No. 1 0/455, 931 No. Application No. "SYSTEM AND METHOD OF PERFORMING DOT INVERSION WITH STANDARD DRIVERS AND BACKPLANE ON NOVEL DISPLAY PANEL LAYOUTS"; 4) U.S. Patent Application Serial No. 10/455,927, entitled "SYSTEM AND METHOD FOR COMPENSATING FOR VISUAL EFFECTS UPON PANELS HAVING FIXED PATTERN", entitled "Visual Effects Compensation System and Method for Fixed Noise Panels with Reduced Quantization Errors" (NO) WITH REDUCED QUANTIZATION ERROR); (5) U.S. Patent Application Serial No. 10/456,806, entitled "DOT INVERSION ON NOVEL DISPLAY PANEL LAYOUTS WITH EXTRA DRIVERS"; And (6) U.S. Patent Application Serial No. 10/456,838, entitled "Liquid LCD backplane design and non-standard sub-pixel configuration address" (LIQUID CRYSTAL DISPLAY BACKPLANE LAYOUTS AND ADDRESSING FOR NON-STANDARD SUBPIXEL ARRANGEMENTS). The above application is hereby incorporated. In conjunction with the techniques disclosed in the above-identified patent application, some of the sub-pixel coloring systems and methods further disclosed in the U.S. Patent Application, which is hereby incorporated by reference in its entirety, is particularly significant. (1) on January 16, 2002 U.S. Patent Application Serial No. 1/No. 5,612, filed on-co., entitled,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, PIXEL DATA FORMAT); (2) The application of U.S. Patent Application Serial No. 1 0/1 50, 355, filed on May 17, 2009. (METHODS AND SYSTEMS FOR SUB-PIXEL RENDERING WITH GAMMA ADJUSTMENT); (3) U.S. Patent Application Serial No. 10/215,843, filed on Aug. 8, 2002, entitled "with Adaptive Filtering" "METHODS AND SYSTEMS FOR SUBPIXEL RENDERING WITH ADAPTIVE FILTERING"; (4) U.S. Patent Application Serial No. 10/379,767, filed on March 4, 2003 The application title is "SYSTEMS AND METHODS FOR TEMPORAL SUB-PIXEL RENDERING OF IMAGE DATA"; (5) US Patent Application filed on March 4, 2003 Application No. 1/0, 379, entitled "SYSTEMS AND METHODS FOR MOTION ADAPTIVE FILTERING"; (6) US application filed on March 4, 2003 Patent Application No. 10/379,766, entitled "SUB-PIXEL RENDERING SYSTEM AND METHOD FOR IMPROVED DISPLAY VIEWING ANGLES"; and (7) U.S. Patent Application Serial No. 10/409, file No. 3, filed on Apr. 7, the entire disclosure of which is entitled "IMAGE DATA SET WITH EMBEDDED PRE-SUBPIXEL RENDERED IMAGE). The above-mentioned applications are hereby incorporated by reference. SUMMARY OF THE INVENTION The present invention is directed to a liquid crystal display, and more particularly to a color gamut conversion system and method. The present invention discloses a system and method for generating raw image data to a target image data set, or for converting it to a device that may be in a different color gamut, a line coloring color gamut conversion system, comprising: an input channel device , -, color, unit, one chroma/brightness unit, one hue angle calculator and one color gamut conversion unit. The round-in channel device receives the raw image data. The color gamut list = data that converts the initial image data into a perceptually uniform sentence space. The chrominance/^ degree unit converts the perceptually uniform spatial data into a format containing some color and degree components. The hue angle calculator calculates 35 hue angles from the chroma/luminance unit to ^ 35 chroma > 1 ' and calculates the original shirt image data based on the received components. The gamut conversion unit derives some gamut conversion values' using the gamut conversion table to apply these values to some components of the original image data. [Embodiment] Reference will now be made in detail to the particular embodiments and embodiments, Wherever possible, the same reference numerals will be used in the Most monitors and television devices today are designed to display three-valued color data such as RGB (RGB RED, GREEN, RED 'Red, Green, Blue) and/or sRgb (s = Spaee) , space red green blue) 'or three-valued cluoma/luminance signal, such as YIq or YCbCr signal. The term sRGB is used to describe nonlinear systems. Most PC images are stored using this nonlinear system, which is sometimes referred to as R, G, B, or nonlinear 1294616.

GB爲了製造一些比較亮的顯干哭色糸褚沾骷-。的…員以及一些帶有較大色 7 S 、頌示器，製造商們正開妒器。這此顯亍莠踗女夕 "考慮某些夕原色顯示 k 一 4不益將有多於三個原色，多原色影像數據源，而且還疋，一、原色顯示器進行轉換。在將-些影像轉換成多系、u #擴展色域系統的過程中，將需要利用這些糸統的一些較大色域。本I月揭不一些能將現有的三數值色彩數據的色域轉換—擴展或收縮之方法及系統，以成爲可利用這種新等級 ^不-的色域。這樣一些方法及系統對於帶有任意數目原色的一些多原色顯示器均能產生效用。色域轉換的假定之一，是一些電視照相機（τν camera)、數位照相機（dig丨tal camera)以及其他一些輸入設備不會破壞現貫世界的擴展色域（expanded以瓜加），但將其大部分壓縮成能被展現的有限色域（limite(i gamut)，如第 4 A圖所建議的。爲了便於計算，對一些色域轉換演算法還假疋為項壓發生在某一 CIE色度圖（chromaticity diagram) 内的一個白色點。例如第4 A圖所示，在現實世界所感受到的一些色彩在色域外的一個區域402中可被找到，該色域是藉由影像擷取設備（image capture device)被展示出來（例如三角形404)。一種這樣的現實世界的色彩408被壓縮成與白色點406大體上共線的色彩點410。這暗示在這些影像源中某些喪失的色域，可藉由將該色域再伸展返回而被重新構造出來。 1294616 、- 爲了觀察不同的設備耷，GB in order to make some brighter and brighter 哭糸褚。 -. The ... and the larger 7 S, the display, the manufacturers are opening the device. This is a singer's eve " Considering some eve primary colors, k-4 will not have more than three primary colors, multiple primary color image data sources, and also 疋, one, primary color display for conversion. In the process of converting some images into a multi-system, u # extended color gamut system, some of the larger color gamuts of these systems will need to be utilized. This I month reveals some methods and systems that can convert the color gamut of existing three-value color data—expand or shrink, to become a color gamut that can take advantage of this new level. Such methods and systems can be useful for some multi-primary displays with any number of primary colors. One of the assumptions of gamut conversion is that some television cameras (τν cameras), dig丨tal cameras, and other input devices do not destroy the extended color gamut of the current world (expanded to melon), but Most of the compression is limited to the limited color gamut (i gamut), as suggested in Figure 4A. For ease of calculation, some gamut conversion algorithms also assume that the item pressure occurs in a certain CIE color. A white point in the chromaticity diagram. For example, as shown in Fig. 4A, some of the colors perceived in the real world can be found in an area 402 outside the gamut, which is obtained by the image capturing device. An image capture device is shown (e.g., triangle 404). One such real world color 408 is compressed into a color point 410 that is substantially collinear with the white point 406. This implies some loss of color in these image sources. The domain can be reconstructed by re-stretching the color gamut back. 1294616 , - To observe different devices,

備色域疋如何進行交互作用，第4B 圖展不出叠加在-個印表機色域422上的監視器的色域。在-些色彩或色調角(hueangle)的某些範圍，監視器色域的飽和度(saturation)必須被壓縮地更多，以便與印表機的色域配合。例如色彩點426被壓縮爲色彩點428，以便與印表機色域422配合（爯去，1、晉細a & & v丹I 其遇朝向白色點424壓知§ )。然而在其他一此多綱备老二邑凋角處’監視器色域的飽和度可被再擴展返回，以利用印表機色域，例如色彩點化擴展到色彩點427。 ' 第4C圖描述在—個假定的六原色多原色顯示色域— 内部的監視器色域440。在所有的色調角，該監視器色域被擴展來利用該多原色色域。例如，色彩點444可被擴展成碰上位於該多原色色域442的外邊界上的色彩點446。這種擴展可能與白色點450共線地發生。許多從一個色域轉換到另一色域的演算法已被發表在文獻中’使用得最多的是對一些CMY ( CMY = Cyan、 Magenta、Yell〇w，青綠、洋紅、黃）彩色印表機上用的輸出，進行如第4B圖中的sRgb顯示數據轉換。這些演算法的共同特點是能保持色調角。通常該被轉換的色彩，首先展現為一種單獨分開的色度/亮度格式（chr〇tna/iuma format)。一些色度分量可被乘以一個能增加或減少飽和度的數值。如果色度的兩個分量被乘以相同的數值，則色調和亮度被維持不變，而僅飽和度被改變。如果相乘的數值小於1 ’飽和度將減少；如果相乘的數值等於丨，飽和度將維持不變；如果相乘的數值大於1，飽和度將增加。通常 12 1294616 這些色域擴展的演算法是在一些CIE Lab或Luv色彩空間中貫施’因爲它們被設計成知覺上均勻的色彩空間。這意味著通常需要向CIE Lab進行轉換並轉換返回，而且這些步驟在計算上都报昂貴。爲了降低計算成本，現將敍述一個實施例。由於sRGB 已經是知覺上均勻的，因此不再需要轉換至計算上昂貴的 CIE Lab，反而可能需要將sRGB直接轉換成YCrCb，或者轉換至其他一些計算費用更為低廉的分開的色度/亮度。這與向CIE Lab轉換不同，可在硬體上非常容易地實現。如果忒二數值色彩數據碰巧已經是YCrCb，這也是一種知覺上均勻的系統，無需對它們進行轉換即可使用。對於大多數電視色彩空間來說，一旦已被數位化，例如數位化爲 YPrPb、YUV和YIQ，相同的情況將是真實的。如果一些原始色彩是線性RGB，首先可利用—個灰度轉換表（gamma ⑶nversion tabie )將其轉換成sRGB。一旦所形成的色彩是在個知覺上均勻、單獨分開的色度/亮度空間内，則可計算出色調角。-個這樣計算上低廉的計算色調角的方法，敘述於一些相關專利申請..案中的一個。色調角計算器於以上所列出的相關申請案中，有需要計算一給定影像數據集的色§周角，現將敘述一新穎的色調角計算器及方法。取常用於個人電腦上的色彩空間於本文中將稱為 sRGB，有時也叫做R，G，B，或非線性R(}B。許多色彩空間具有色度與亮度，兩者被運行在一些軸中一個軸上的一些 13 1294616 灰線所分開。這對SRGB來說不是真的，但其能夠被轉換到一個要轉換的地方。無論如何，對於其他的輸入格式，有可能接收已經為色度/亮度格式的數據，例如，Ycbcr，Yuv， CIE Lab與Luv。在這些色彩空間中，色度資訊可編碼為兩個預示的數字，該兩數字可描述一個二維的色彩向量。對於這兩個分量，.每-個系統均有其稱呼，❻為了說明下面實施例的目的，這兩個分量在此稱作「χ」與「y」。現將敘述色調角計算的一個可能實施例。對此向量計算色調角的一個步驟，是記錄兩分量的㈣（即正負號）並取其絕對值。這將減少計算至四個象限中的一個。於某一象限内該角度被計算出’接著符號位元決定真正的象限為何，亚且常數角度被加在末尾。例如，於第一象限内，兩分量的數字均為正，且簡單的角度計算是正確的。在第二象限内從絕對值計算角度之後，9〇度角並須被加進；在第三象限内180度角必須被加進，而於第四象限内27〇度角必須被加進。無論如何，另—色調角計算器的實施例對色調角、色域擴展與多原色轉換，提供一新穎的色彩空間。以此方式，有可旎選擇一個新的數字表示法，以使得硬體較容易實施。例如’我們可以—分開的符號位元，將數字儲存為正的7位元數字，以代#執行兩個補數⑽灿叫，這可消除取絕對值的情況。當相乘或相除兩個數字時，較低的7位το可立刻被相乘’而不須處理符號的作用，若同時施以異運算（X〇R)，則可獲得最終的正確符號。僅於執仃加法的情況下，符號位元須被檢驗，並且如果兩者的符 14 1294616 號不同’首先將對兩數字中的-個數字進行反運算㈣則让補數_pllment及增量increment)，或者如果造成一個借位（borrow)，將於之後進行反運算。於再另-個實施例中，環繞45度角有另一個對稱可被利用。於第-冑45冑角的内部（即於第一個八分圓（制㈣叫内），兩數字中的—個（χ分量）總是大於另一個〇分量）。因此’有可能檢驗這兩個分量，如果有需要的話，還可將兩者交換以| x分量總是較大。當執行此操作，y ^於X的事f被紀錄下來，符號位元被分開儲存。一旦執饤此插作’所有計算皆可被執行，就好像所有的角度均. 該第-個八分圓，’介於〇度與45度之間，並且對稱性將使我們理解其他的情況。如果交換該色域的χ肖y分量並計算該角度’稍後必須從90度減去該角度，以對該象限得到正確的結果。從三角學的考量，計算角度的公式為arctan(y/x)，可以任何的计算方式，完成y/x的除法計算。一個這樣的方式疋使用一個倒數表（inversion taMe)，先取x的倒數然後再采上y數值。為了精確地計算倒數值，可能需要一個大的倒數表，因此需要有12位元大的乘法器（muitiplier)。另一種计算方式為執行徐法。下面將提到的最佳化，可允許除法器模組產生小至5位元寬的結果。於任何情況中，結果總是位在第一個八分圓内的小於或等於i的一個數。所以此結果在最有效位元之前，總是為一個帶有二進小數點 (binary point)的二進固定點（binary fixed p〇int)的數字。應How the color gamut interacts, and Figure 4B shows the color gamut of the monitor superimposed on a printer gamut 422. In some areas of color or hueangle, the saturation of the monitor gamut must be compressed more to match the color gamut of the printer. For example, color point 426 is compressed into color point 428 for cooperation with printer color gamut 422 (go, 1, and a && v Dan I encounter the white point 424 to suppress §). However, the saturation of the monitor color gamut at the other end of the field can be re-expanded back to extend to the color point 427 using the printer gamut, such as color dot. 'Current 4C depicts the monitor color gamut 440 inside the hypothetical six primary color multi-primary color gamut. At all hue angles, the monitor color gamut is expanded to take advantage of the multi-primary color gamut. For example, color point 444 can be expanded to touch color point 446 located on the outer boundary of the multi-primary color gamut 442. This expansion may occur collinear with the white point 450. Many algorithms that have been converted from one color gamut to another have been published in the literature. 'The most used are on some CMY (CMY = Cyan, Magenta, Yell〇w, Green, Magenta, Yellow) color printers. The output used is subjected to sRgb display data conversion as shown in Fig. 4B. A common feature of these algorithms is the ability to maintain a hue angle. Usually the converted color is first presented as a separate separate chroma/luminance format (chr〇tna/iuma format). Some chrominance components can be multiplied by a value that increases or decreases the saturation. If the two components of the chrominance are multiplied by the same value, the hue and brightness are maintained, and only the saturation is changed. If the multiplied value is less than 1 'saturation will decrease; if the multiplied value is equal to 丨, the saturation will remain unchanged; if the multiplied value is greater than 1, the saturation will increase. Usually 12 1294616 These gamut-expanded algorithms are implemented in some CIE Lab or Luv color spaces because they are designed to be perceptually uniform color spaces. This means that conversions and conversions are usually required to the CIE Lab, and these steps are expensive to calculate. In order to reduce the computational cost, an embodiment will now be described. Since sRGB is already perceptually uniform, there is no longer a need to switch to the computationally expensive CIE Lab, but instead it may be necessary to convert sRGB directly to YCrCb, or to other discrete chrominance/brightness calculations that are less expensive. This is different from the conversion to CIE Lab and can be implemented very easily on hardware. If the binary color data happens to be YCrCb, it is also a perceptually uniform system that can be used without conversion. For most TV color spaces, the same situation will be true once it has been digitized, such as digitally converted to YPrPb, YUV, and YIQ. If some of the original colors are linear RGB, you can first convert them to sRGB using a grayscale conversion table (gamma (3) nversion tabie). The hue angle can be calculated once the resulting color is within a perceptually uniform, separately separated chromaticity/luminance space. - A method for calculating the low-cost calculation of the hue angle, which is described in one of the related patent applications. Hue Angle Calculator In the related applications listed above, there is a need to calculate the color ‧ angle of a given image data set. A novel hue angle calculator and method will now be described. The color space commonly used on personal computers will be referred to herein as sRGB, sometimes called R, G, B, or nonlinear R(}B. Many color spaces have chrominance and brightness, both of which are run at some Some 13 1294616 gray lines on one axis of the axis are separated. This is not true for SRGB, but it can be converted to a place to be converted. In any case, for other input formats, it is possible to receive the color already. Data in degrees/brightness format, for example, Ycbcr, Yuv, CIE Lab and Luv. In these color spaces, chrominance information can be encoded into two predictive numbers that describe a two-dimensional color vector. The two components, each of which has its name, are referred to herein as "χ" and "y" for purposes of illustrating the following embodiments. One possible embodiment of the hue angle calculation will now be described. One step in calculating the hue angle for this vector is to record the (four) (ie, sign) of the two components and take their absolute value. This will reduce the calculation to one of the four quadrants. The angle is calculated in a certain quadrant. 'connected The sign bit determines the true quadrant, and the sub-constant angle is added at the end. For example, in the first quadrant, the two components are positive, and the simple angle calculation is correct. In the second quadrant from the absolute After the value is calculated, the 9-degree angle must be added; the 180-degree angle must be added in the third quadrant, and the 27-degree angle must be added in the fourth quadrant. In any case, the other-tone angle Embodiments of the calculator provide a novel color space for hue angle, color gamut expansion, and multi-primary color conversion. In this way, a new digital representation can be selected to make the hardware easier to implement. For example, 'we It is possible to separate the sign bit and store the number as a positive 7-bit number to perform two complements (10) for the #, which eliminates the absolute value. When multiplying or dividing two numbers The lower 7 bits το can be multiplied immediately 'without the need to deal with the effect of the symbol, if an arithmetic operation (X〇R) is applied at the same time, the final correct symbol can be obtained. In the case of imposed addition, The sign bit shall be verified and If the two characters 14 1294616 are different, 'the first one will reverse the - number of the two numbers (four), then the complement _pllment and incremental increment), or if a borrow is caused, it will be carried out later. Inverse operation. In still another embodiment, another symmetry around the 45 degree angle can be utilized. In the interior of the 胄45胄 corner (that is, in the first octant (in the case of the system), the one (the χ component) of the two numbers is always greater than the other 〇 component). Therefore, it is possible to test these two components, and if necessary, exchange the two with | x components are always larger. When this operation is performed, the event f of y ^ X is recorded, and the sign bit is stored separately. Once this is inserted, 'all calculations can be performed, as if all angles are. The first octant, 'between 〇 and 45 degrees, and symmetry will make us understand other situations. . If the gamma y component of the gamut is exchanged and the angle is calculated, the angle must be subtracted from 90 degrees later to obtain the correct result for the quadrant. From the trigonometry considerations, the formula for calculating the angle is arctan(y/x), which can be used to calculate the division of y/x in any calculation. One such way is to use a reciprocal table (inversion taMe), first take the reciprocal of x and then take the y value. In order to accurately calculate the reciprocal value, a large reciprocal table may be required, so a 12-bit large multiplier is required. Another way to calculate is to implement Xufa. The optimizations mentioned below allow the divider module to produce results as small as 5 bits wide. In any case, the result is always a number less than or equal to i within the first octant. So this result is always a number with a binary fixed p〇int with a binary point before the most significant bit. should

X 注意的是，除以〇通常並不是一個問題，這是由於隨著 15 1294616 與y刀里的不斷交換’ 一直到χ大於或等於】為止，x為。的唯時械疋y也是〇。於此例子中，除法器電路會傳回· 個〇.結果，做為一個可能的合理預設值。一個有: 趣的情況發生在X等於y時，此情況中的結果將不符合為· 一個固定點二進制數（fixed p〇int binary __，除非於該二進小數點（binary point)之上有多於一個位元。為處理此情況’除法器電路為此可有一個額外的位元加入。另一種方式’除法器電路為此特殊情況可允許傳回一個稍微小的數字。僅需滿足上述情況，其他的技術也可能實行，如果< 系統遭遇到此問題，然後一些的裝置可提供以處理此問題。反正切（arc-tangent)函數可以一個小的表來實施。在實際上，此反正切函數簡直是接近一直線，而一些從事者 (pracutioner)(即於其他應用當中有使用到正切函數）在發現廷個表時會將之略過。但是，如果由於做此操作所引入的誤差大於可接受的限制時，希望在系統中保留這個反正切表。將於下面敘述，此表可能很小而因此不昂貴。當取色度的X與y分量的絕對值以及為了需要而交換兩者時，位元被儲存以允許對稍後的簡化進行修正。下面是這些位元的一個表，須對所有八分圓採取操作以進行修正。八分圓 Y<0 Χ<〇 Υ>χ 操作 1 0 0 0 不需要 2 0 0 1 90度角減一角度 3 0 1 1 — 90度角加一角度 16 1294616 4 0 1 0 180度角減一角度 5 1 1 0 1 8 0度角加一角度 6 1 1 1 270度角減一角度 7 1 0 1 270度角加一角度 8 1 0 0 360度角減一角唐 ------—--- 應注意的是，對於每個不同的位元組合（bit combination)是唯一的，但當其以類似上述的八分圓次序列出時，由串接（concatenate)此三個位元所產生的二進小數點不是八分圓數字。當然有可能建構具有此位元位址次序（bit address order)的表，或者建構成具有所要的任何位元位址次序，只要能夠唯一地解碼到任何給定的八分圓。因為有時上述的操作需要減去該角度，也應1 有可能包括-個位元，指示該角度必須於其被加至心：補償值（。腕)之前，進行反運算。下面的表是上述觀察= 一個實施例，並可建構成一個對照表、加怍對照表，action i〇〇k-uptable; LUT)或者一些其他的計算工具。X Note that dividing by 〇 is usually not a problem, because x is the same as 15 1294616 and y knives continue to exchange 'until χ is greater than or equal to 】. The only time weapon is 〇. In this example, the divider circuit will return a 〇. result, as a possible reasonable preset. One has: Interesting happens when X is equal to y, the result in this case will not match the fixed p〇int binary __, unless there is a binary point above the binary point More than one bit. To handle this situation, the 'divider circuit can have an extra bit added for this. Another way' the divider circuit allows for a slightly smaller number to be returned for this special case. In other cases, other techniques may be implemented. If the < system encounters this problem, then some devices can be provided to handle this problem. The arc-tangent function can be implemented in a small table. In fact, this The arctangent function is simply close to the straight line, and some practitioners (that is, the tangent function used in other applications) will skip it when it finds a table. However, if the error is introduced due to doing this operation, Above the acceptable limit, it is desirable to keep this inverse tangent table in the system. As will be described below, this table may be small and therefore inexpensive. When taking the X and y of the chromaticity The absolute value and the exchange of both for the need, the bits are stored to allow for later simplification correction. The following is a table of these bits, which must be operated on all octants for correction. Y<0 Χ<〇Υ>χ Operation 1 0 0 0 No need 2 0 0 1 90 degree angle minus one angle 3 0 1 1 — 90 degree angle plus one angle 16 1294616 4 0 1 0 180 degree angle minus one angle 5 1 1 0 1 8 0 degree angle plus one angle 6 1 1 1 270 degree angle minus one angle 7 1 0 1 270 degree angle plus one angle 8 1 0 0 360 degree angle minus one angle Tang --------- - It should be noted that it is unique for each different bit combination, but when it is listed in an octant order similar to the above, it is generated by concatenating these three bits. The binary decimal point is not an octant number. It is of course possible to construct a table with this bit address order, or to construct any bit address order that is desired, as long as it can be uniquely decoded to any Given a octant. Because sometimes the above operation needs to subtract the angle, it should also be 1 Include - a bit indicating that the angle must be added to the heart: before the compensation value (. wrist), perform the inverse operation. The following table is the above observation = an embodiment, and can be constructed to form a comparison table, crowning Comparison table, action i〇〇k-uptable; LUT) or some other calculation tool.

17 1294616 011 0 90 100 1 360 101 0 270 110 0 180 111 1 270 1 於此表的第一欄ΧΥ>，是y的符號的二進制串接（binary · concatenation)、X的符號以及檢驗y〉x的結果。這是該表的位址，現以二進制的計數次序出現。第二欄NEG為一個位元，指示於該反正切表中在某一角度之後被檢視（look up)，其必須被反運算。該反正切表是如此的小，以致於可於該表中儲存兩個正數以及兩者的負補數（compliment negative)，並使用此位元選擇正確的一個。於表中的第三欄是加到末尾的角度補償值（angle 〇ffset)，以做最後的修正，並傳回介於〇至360度的一角度。看從色域的x與y分量計算角度，對零色調角（hue angle zer〇)的色彩將導致一個任意的選擇。例如在YCbCr的情況下’將導致零（zero)是稍微比純藍色更為洋紅的一個色彩。在一個例子中，有可能調整色調角以使得零色調角位在一些原色色彩的一個色彩上。例如，於HSV中，紅色有一個零的色。周角。藉由加進一個常數，有可能選擇一些原色中的個使成為零，這會造成那個原色纏繞返回到零（模數 3 6 0 度）。一 yfm 個實施例不須要在末尾包括一個額外的運算， 18 1294616 替換該操作表ADD攔位中的所有項（entry)能夠做事先的調整，以必須的零點產生數字。 - 在上面的討論中，一些角度的測定是假定圍繞一個圓爲 . 3 60度。但可能需要選擇角度的計量單位，以使得所形成的數值在一些硬體的具體實施方案中更易於處理。例如，如果認爲圍繞一個圓爲256度，則一些色調角可與一個8 位疋數字完美地配合。此外，在一個八分圓内只有32個這樣的「度數」，使得除法器電路僅需具有5個位元，而且反正切表僅具有32項，每項皆爲5位元。以「模數256」計算角度可以較簡單，例如在一個8位元的加法器中實現。至於一些大於2;r弧度的角，該系統僅需容許加法器可以溢位（overflow)，即可將正確的答案保留下來。第9圖展示根據本發明的一些原理所製成的一種色調角計算器900的一個可能實施的實施例。色度的X與y分量具有從方塊902所取得的絕對值，而1其符號也可被儲存起來。S y>x作-次檢測，如果方塊9〇4的判斷結果爲籲真，則將此兩個數交換。將y的值除以χ的值，在方塊9〇6 中給出結果的前面的5個位元。該相除結果用作一個反正切對照表（arc tangent LUT) 9〇8 .中的索引。一些符號位元與y>x的位元，對於「操作」冑照表912也被用來作爲一個索引，指示該角是否要執行反運算，以及應當加上多少。忒角根據表912可選擇地在方塊9 i 〇中執行反運算。根據才木作表，一個角度補償值被加到該八分圓角上，以得到最終的色調角。 19 1294616 * 、 * - - 一旦該色調角被計算，就常被用來確定色彩是屬於某一17 1294616 011 0 90 100 1 360 101 0 270 110 0 180 111 1 270 1 The first column of this table ΧΥ > is the binary concatenation of the symbol of y (binary concatenation), the sign of X and the test y>x the result of. This is the address of the table and is now in binary count order. The second column, NEG, is a bit indicating that it is looked up after an angle in the arctangent table, which must be reversed. The inverse tangent table is so small that two positive numbers and a negative negative of both can be stored in the table, and the correct one is selected using this bit. The third column in the table is the angle compensation value (angle 〇 ffset) added to the end for the final correction and returns an angle between 〇 and 360 degrees. Looking at the angle of the x and y components of the color gamut, the color of the hue angle zer〇 will result in an arbitrary choice. For example, in the case of YCbCr, 'zero' is a color that is slightly more magenta than pure blue. In one example, it is possible to adjust the hue angle such that the zero hue angle is on one of the colors of the primary colors. For example, in HSV, red has a zero color. Week angle. By adding a constant, it is possible to select one of the primary colors to become zero, which causes the primary color to wrap back to zero (modulo 3 6 0 degrees). A yfm embodiment does not need to include an extra operation at the end, 18 1294616 replaces all entries in the ADD block of the operation table to make prior adjustments, producing numbers with the necessary zeros. - In the above discussion, some angles were determined to be around a circle of .60 degrees. However, it may be desirable to select the unit of measure for the angle so that the values formed are easier to handle in some hardware specific embodiments. For example, if you think that it is 256 degrees around a circle, some hue angles can perfectly match an 8-digit 疋 number. In addition, there are only 32 such "degrees" in an octant, so that the divider circuit only needs to have 5 bits, and the inverse tangent table has only 32 items, each of which is 5 bits. Calculating the angle with "modulo 256" can be simpler, for example in an 8-bit adder. For some angles greater than 2;r radians, the system only needs to allow the adder to overflow, leaving the correct answer. Figure 9 shows an embodiment of a possible implementation of a hue angle calculator 900 made in accordance with some principles of the present invention. The X and y components of chrominance have absolute values taken from block 902, and their symbols can also be stored. S y > x is used for the second detection. If the judgment result of the block 9 〇 4 is true, the two numbers are exchanged. Dividing the value of y by the value of χ gives the first 5 bits of the result in block 9〇6. The division result is used as an index in an arc tangent LUT 9〇8. Some of the symbol bits and the bits of y>x are also used as an index for the "operation" reference table 912, indicating whether the corner is to perform an inverse operation and how much should be added. The corners optionally perform an inverse operation in block 9 i 根据 according to table 912. According to the wood table, an angle compensation value is added to the eight-point fillet to obtain the final hue angle. 19 1294616 * , * - - Once the hue angle is calculated, it is often used to determine if the color belongs to a certain

色度圖（chromaticity diagram)的哪個三角形。第1〇A與 10B圖展示以色調角計算色度三角形數（ triangle number)的兩個實施例1000與1〇2〇。參閱第1〇A 圖，例如在多原色的顯示器實例中，複數個原色預先被轉換爲一些色調角，並被儲存在一個原色對照表（pHmary look-up table ’ LUT : Look-Up Table) 1002 中。這些原色在複數個比較器（comparato〇1004中針對所計算出的色調角1〇1〇進行檢測。一個多工器模組（de_multiplexar mQduie) 1006將這些比較的結果轉換成一個色度三角形數ι〇〇8。如在一些相關的專利申請案中所述，帶有三個原色，紅、綠、藍的色度圖可以被分成爲三個三角形或區域·· RGW (紅綠白）、gbw(綠藍白）與BRW(藍紅白），如第12A圖所示。根據色調角，色度三角形數可識別某一色彩屬於這些三角形中的哪一個，而不需進行昂貴的計算，以將X、乂轉換成色度數。參閱第10B圖，其展示出一個關於色度三角形數的色調角1010的另一實施例1020。每一個可能的色調角用的三角死/數白預先被计鼻出來並儲存在一個對照表1 022中。該色調角對於這個對照表用來作爲一個索引，在一個步驟中讀取色度三角形數1024。當把圍繞圓的色調角的「度數」限制爲2的某一次方（power)，這將可特別有效地實現。這可比第10A圖的實施例迅速，但可能需要較多的閘（gate ) 來實覌。弟11圖展示出用在元整「色域管線」（ganlut pipeHne ) 20 1294616 1100中的一個色調角轉換器1丨06，該色域管線對一些數位電視YCrCb信號1102進行轉換，以做為多原色顯示器11〇4 的輸出。一些YcrCb的Cr與Cb信號已經是一個色度向量，並直接被供應給色調角計算器11 〇 6。該所形成的色調角，如在一些相關的專利申請案中所述，.被用來做為一個對於色域擴展（gamut expansion)對照表1108的索弓|，以生成一個擴展的比例因數（scale factor)。這個比例因數被乘以Cr 與Cb的一些分量，可藉由改變色彩飽和度來擴展色域，而無須改變色調角或亮度（luma)。應當注意的是，色域擴展可發生在一些不正確的YcrCb色彩上，這些不正確的YcrCb 色彩隱含着對它們已應用了非線性變換（non_linear transformation)。這對於其他一些電視信號（YPbPr、YIq、Which triangle of the chromaticity diagram. Figures 1A and 10B show two embodiments 1000 and 1〇2〇 for calculating the chromaticity triangle number from the hue angle. Referring to Figure 1A, for example, in a multi-primary display example, a plurality of primary colors are previously converted to some hue angles and stored in a pHmary look-up table 'LUT: Look-Up Table 1002 in. These primary colors are detected in a plurality of comparators (comparato 〇 1004 for the calculated hue angle 1 〇 1 。. A multiplexer module (de_multiplexar mQduie) 1006 converts the results of these comparisons into a chromaticity triangle number ι 〇〇 8. As described in some related patent applications, a chromaticity diagram with three primary colors, red, green, and blue, can be divided into three triangles or regions·· RGW (red, green, white), gbw ( Green blue and white) and BRW (blue red white), as shown in Figure 12A. Depending on the hue angle, the number of chrominance triangles identifies which of these triangles a color belongs to, without expensive calculations to X Converting to a chromaticity number. Referring to Fig. 10B, another embodiment 1020 is shown for a hue angle 1010 for the number of chromatic triangles. Each possible hue angle is pre-counted with a triangle dead/number white It is stored in a look-up table 1 022. The hue angle is used as an index for this look-up table, and the number of chromaticity triangles is read 1024 in one step. When the "degree" around the hue angle of the circle is limited to two This will be particularly efficient at a certain power. This can be faster than the embodiment of Figure 10A, but may require more gates to implement. The 11th figure shows the color used in the meta A color tone converter 1丨06 in a ganlut pipeHne 20 1294616 1100 that converts some digital television YCrCb signals 1102 as outputs of a multi-primary display 11〇4. Some YcrCb The Cr and Cb signals are already a chrominance vector and are supplied directly to the hue angle calculator 11 〇 6. The resulting hue angle, as described in some related patent applications, is used as a For the gamut expansion, compare the score of the table 1108 to generate an extended scale factor. This scaling factor is multiplied by some components of Cr and Cb, which can be extended by changing the color saturation. Color gamut without changing the hue angle or luma. It should be noted that gamut expansion can occur on some incorrect YcrCb colors. These incorrect YcrCb colors imply that they have been applied offline. Non-linear transformation. This is for some other TV signals (YPbPr, YIq,

Yuv等）與SRGB也是真實的。隱含在這些信號中的非線性變換，意味着它們在知覺上是均勻的，這使它們構成理想的色彩空間以進行色域轉換。其他一些從業人員是首先轉換成CIE Lab或Luv色彩空間以達到知覺上的均勻性，吾人已發現，在利用像YcrCb這樣的非線性空間時，該計算上昂貴的步驟並非必要的。一旦已進行了色域擴展，一個輸入灰度對照表（input gamma LUT)1110被用來將一些 YcrCb值轉換成一個正確領域的線性色彩空間，來進行色彩空間轉換與子像素着色。如上所述，在第11圖中，色調角亦可用來計算模組i i 12 中的色度三角形。該色度三角形數被用来作爲一個對於多原色矩陣對照表（multi-primary matrix LUT )1114的索引。結果是3xn個係數（此處η爲顯示器中原色的數目），這些 21 ^94616 =數t模組1116中被乘以一些線性YcrCb的值，以將之轉、成夕原色顯不。多原色的一些分量可能被發送到一個子、象素着色（SPR; Sub-Pixel Rendering)模組 1118，再被發送到個輸出灰度表，最後則發送到多原色顯示器 1104 〇第12B圖是色域管線的另一個實施例12〇〇。在這個系、、先中些紅、綠、藍的數值1202被輸入並在1204轉換成些色度/亮度數值。從那裏，色調角計算器1206把色調角供給一個角度三角形單元（angle_triangleunit) 12〇8，以確疋影像點落在哪個色度三角形内。這被用來選擇方塊 1210處的多原色轉換矩陣。這個矩陣被供應給兩個3別乘法器（multiplier) 1212a與1212b。乘法器1212a將輸入的、、工”，彔、監色彩轉換成顯示單元（display unit ) 1224的多多原色的色彩空間。一些紅、綠、藍輸入數值也被供應給尋找這三個色彩分量最大值的最大化單元（max unit ) 1214,並將該最大值供應給一個倒數對照表（inverseLUT) 1.216 ’該倒數對照表産生一個比例因數，使得這些rgb (紅綠藍）數值規格化爲帶有相同色調角的最大值。該比例因數被乘以一些紅、綠、藍輸入數值，以産生具有與一些原始紅、綠、藍數值相同的色調角的最大值容許色彩。乘法裔1212b取出那個最大化的色調色彩（hue color)，並將之轉換成標的多原色的色彩空間。那些數據被供應給最大化單元1218，該最大化單元尋找一些多原色分量的最大數值，並且倒數對照表1220將之轉換成一個比例因數，與從乘法器1212a輪出的一些特殊的多原色數值相乘。這個系 22 1294616 、統的結果疋已被擴展或收縮爲多原色顯示器1224色域的標的多原色空間内的一個色彩點。但如在以上併入參考的任一專利申請案中所述，該項數據在該顯示器上着色之前，可視情況利用子像素着色（spR)單元1222來進行子像素着色。在以上一些實施例中所涉及的一些功能方塊，可利用硬體及/或軟體的任何組合，包括一些零件或模組，例如以一個或多個圯憶體設備或電路系統來實施。例如，一種可程式的閘陣列（programmable gate array )或類似的電路系統可被裝配成來貫施這樣一些功能方塊。在其他例子中，一個在記憶體内執行程式的微處理機，亦可實施這樣的一些功能方塊。色域擴展一旦色調角被計算出（利用任何適當的方式），即被用來作爲進入色域擴展數值表的一個索引。在一個實施例中，色域擴展表的輸出可用一個介於〇與2的數來表達該數可能非常接近於1。這些數可以用在二進小數點（binary point)之上帶有一位元的固定點二進數字（fixed p〇int binary number)來表達。以適當的向後移動，允許將整數乘法器作爲一個廉價的系統用來實現色域擴展。該擴展數可被用來與色度的兩個分量相乘，在不影響色調或亮度的情況下，因此改變了飽和度。爲了設計一個系統（例如一個監視器），這樣一個系統可具有好幾條輸X路徑，例如SRBG、YCbCr、S-VHS等。 23 1294616 _ 對於這些路徑的每一條，不同的硬體及/或軟體可爲好幾條路徑的每一條提供服務，特別如果對所假定的影像數據輸入不需要作任選的處理時，例如，YCbCr就不需要一個色度/冗度分離( chroma/luma separator )，因爲該空間已經為那種形式。也能有可能設計一種系統以讀取影像數中的標題（header)，以辨別哪種影像數據格式正被輸入，使得某些任選子系統與程式（routine)可旁路繞開（bypass)。其他一些適當的實施例與具體實施方案，用於設計一個採用不同輪入的完全系統，以及用於處理這樣一些不同輸入的需要也是可能的。第1圖描述本發明的提供色域擴展及/或轉換的一個實 &例糸統1〇〇輸入原始影像數據（source image data) 102 (例如RGB )。如果需要，該原始影像數據可在灰度單元 (gamma unit) 103處被轉換成知覺上均勻的色彩空間。該數據被傳遞到色度/亮度轉換器1〇4。來自轉換器1〇4的一些色度分量被輸入色調角計算器106内，而色調角於此後被輸入至色域擴展方塊1 〇 8 (可像在一個表内查找那樣來實施或者本身做為一個计异方塊）。該方塊的輸出是一個色域擴展數（gamut expansion number)，該色域擴展數可藉由一些乘法器用來對色度的兩個分量進行縮放。這些色度分I接著被輸出，以便由其他一些零件丨丨〇使用。將可瞭解到，本發明的其他一些實施例也是可能的。如前所述，色域擴展可以用許多合適的方式來實現，這些方式包括但並不僅f艮於查表方式、計算子系統方式或其他類似方式。此外，其他一些路徑也是可能的。第丨圖展示出色域擴展 24 1294616 ⑽的輸出可被應崎104輪出的—些色度數值上；但另一替代方案，是色域擴展單元⑽直接將色域擴展數值供應給其他一些零件1 1 Q。如上所述，如果該擴展數值小於1，它減少色彩的飽和度’如果該數值等⑨卜它大體上保持相同的飽和度，如果該數值大於卜它增加色彩的色域的飽和度。當献，亦可能執行-些替代的實施例’於這些實施例中，&他一此數值比例（numerie seale)也表示對影像數據作同樣的處建立色域轉換表、現將敍述色域轉換表的-個實施例，以及如何建構該色域轉換表。建立色域轉換表的—個可能的途徑，是橫跨一些色彩空間的輸入與輪出色域的某些邊緣。在其色域的一些邊緣爲直線的CIE色度空間内，經常是做這樣的處理。第2A、2B與2C圖展示出三個不同的色彩空間用的三個不同色度圖㈣i:_ieity diagram)。這些不同的色度圖將用來做為生成色域轉換表的實例。第2A圖描述SRGB用的色度圖。第2B圖展示-個CMY 輸出設備，例如一個印表機用的色度圖。第2C圖展示一個假定的六原色多原色顯示器用的色域的色度圖，於此情況’該色域爲第2A與2B圖的兩個色域的組合。應當注意的是’第2C圖的多原色色域所覆蓋的面積，事實上可^ 於第2A肖2B圖的兩個三角形色域簡單相交的面積。此外，此間所揭示的一些方法與系統同樣可應用到任何__般 25 1294616 - 的原色的顯示器。建立色域轉換表的一個實施例，是預先離線（〇ff-line) 藉由橫跨色域的一些邊緣開始。所形成的數據接著被用來生成一些被燒錄在一些硬體工具（hardware implementations)上的色域轉換表。橫跨每個色域的一些邊緣的目的，是生成一些最大飽和度數值的一個中繼表格 (intermediate table)。當然，可以有一些離開最大值的飽和度中繼值，其對於系統是有價值的。這個表可被量化，以對每一度色調包含一個值。爲了解說這個實例方便起見，將假定這些色調圍繞該色彩色域佔據36〇個整數度數。當然，在一個相關且併入的專利申請案中所揭示的另一個實施例，可對角度的計量單位加以改變，造成圍繞色域的「角度」的數目爲2的某次方，例如爲256「度」。這種對度數的改變，將使得色域擴展的實現在硬體中更爲容易。橫跨這些色域可用幾種方式來做到。一種方式將是對於每一個色調角生成一種色彩，除了帶有任意選擇的飽和度與色度，而且位於一個知覺上均勻的亮度/色度空間内。這些色彩可被轉換到CIE色度且被校正到落在該色域多邊形的邊緣上，接著被轉換返回色度/亮度以得到一些最大的飽和度值。另一種方式，將是沿着CIE色度空間内色域多邊形的一些邊緣，線性地分步驟進行對一些色域的橫跨以及對一些最大飽和度值的生成，將一個色彩點集轉換到亮度/ 色度並計算每一個色彩點的色調角。如果這些線性步驟足夠的小，則對於每一個色調角，可生成幾個亮度/色度數對 (luma/chroma pair)。所有這些數對對於每一個色調角的 26 1294616 ' * - 平均飽和度，可被用來做爲那個角的最大飽和度。當然，其他形成的一些飽和度值，除了平均飽和度之外，皆可根據一些色度/亮度數對的多重性來進行計算，並且使用在色域轉換表中。此外，建立一些最大飽和度表中的一個表，還有其他一些方式，以上兩個演算法只不過是許多可能的實施例中的兩個。第3圖展示出爲第2圖中三個色域多邊形所生成的三個最大飽和度表。第3圖的水平軸爲色調角，在這個實例中，色調角的角度連續地從〇到359度。垂直軸爲在CIE Lab 空間内離開白色點的距離。在一個實施例中，該系統可使用離開白色點距離的真實數值，可是在另一個實施例中，不舄要直接地使用這些真實數值。如下面所討論的，可使用一些不同的曲線之比值。第3圖中的點線（d〇tted Hne)3〇2 爲第2A圖的SRGB色域用的最大飽和度曲線，虛線（dashed lme)304爲第2B圖的CMY色域用的最大飽和度曲線，而實線（solid line)3〇6爲第2C圖的多原色色域用的最大飽和度曲線。其各自的曲線形狀，是由於這些實例所用的的CIE Lab色彩空間在知覺上的非線性均勻度所造成的結果。在這些最大飽和度曲線中一些數值之比，可被用來生成一些最後的色域轉換表。例如，CMY與sRGB的一些數值之比’可生成一個色域轉換表，該表將將sRGB的一些數值轉換到一個對CMY印表機能更適合的色域。應當注意地是’在第3圖的某些色調角處，sRGB曲線落在CMY曲線的下方’這將導致一些色域轉換值大於1，增加帶有那些 •色調角的一些色彩的飽和度。亦應當注意地是，在第3圖 27 1294616 的某些色調角處，sRGB曲線落在CMY曲線的上方，這將導致一些色域轉換值小於1，減少帶有那些色調角的一些色彩的飽和度。此爲第4B圖所展示的情況。爲了選擇一個不同的實例，第3圖内多原色最大飽和度曲線對sRGB最大飽和度曲線之比，將生成一個色域轉換表’該表將把sRGB的一些數值轉換到一個能更適合六原色顯示器的色域。應當注意地是，在第3圖的所有色調角處，sRGB曲線落在多原色曲線之上或者落在多原色曲線的下方’這將導致一些色域轉換值皆大於或等於1，在所有色調角處常增加色彩的飽和度。此爲第4C圖所展示的情況0 藉由縮放亮度及/或飽和度進行色域轉換如在相關且標題爲「將原始色彩空間轉換至紅綠藍白標的色彩空間之方法及裝置；METHOD AND APPARATUS FOR CONVERTING FROM SOURCE COLOR SPACE TO RGBW TARGET COLOR SPACE」的專利中請案中所討論的，敍述一種用於檢測一些色域外色彩（〇ut_〇f_gamut color)，並移動它們到輸出色域空間的表面上成爲一個「好斑點」（good spot)，此處可將好斑點定義爲一個大體上具有與色域外色彩相同色調的色彩。第5圖展示出一個經過簡化的一個實例的二維圖，其中一種RGBW色域在RGB 色域内邛配合。此處特別要敍述兩種技術：色域钳位 (clamping)與色域縮放（scaiing)。色域鉗位將把介於a點與 B點之間的任何色彩移動到RGBW色域表面上的b點。由 28 1294616 於介於BLACK(黑色）與B點之間的所有色彩都不在μ 外’它們將留在原來所在的地方。例如，色域鉗位對c點處的色彩將無任何影響。〇、但是，它對於縮放介於BLACK (黑色）與八點之間的所有色彩是有利的，這將造成在BLACK_A直線上所有的色彩被縮放，使得A點處的色彩移動到第丨圖内的b點。介於A點與B點之間的一些色彩將被移動到色域的内部。介於BLACK與B點之間的一些色彩也被移動，例如，c 點處的色彩被移動到D點。Yuv, etc.) and SRGB are also true. Non-linear transformations implied in these signals mean that they are perceptually uniform, which allows them to form the ideal color space for color gamut conversion. Other practitioners first converted to CIE Lab or Luv color space to achieve perceptual uniformity, and we have found that this computationally expensive step is not necessary when using non-linear spaces like YcrCb. Once the gamut expansion has been performed, an input gamma LUT 1110 is used to convert some of the YcrCb values into a linear color space of the correct domain for color space conversion and sub-pixel rendering. As described above, in Fig. 11, the hue angle can also be used to calculate the chromaticity triangle in the module i i 12 . The number of chrominance triangles is used as an index to the multi-primary matrix LUT 1114. The result is 3xn coefficients (where η is the number of primary colors in the display), and these 21^94616 = number t modules 1116 are multiplied by some linear YcrCb values to turn them into a primordial color. Some components of the multiple primary colors may be sent to a sub-Pixel Rendering module (SPR) 1118, sent to an output grayscale table, and finally sent to the multi-primary display 1104. Figure 12B is Another embodiment of the color gamut pipeline is 12A. In this system, the first red, green, and blue values 1202 are input and converted to a number of chrominance/luminance values at 1204. From there, the hue angle calculator 1206 supplies the hue angle to an angle triangle unit (angle_triangleunit) 12〇8 to determine which chromaticity triangle the image point falls within. This is used to select the multi-primary transformation matrix at block 1210. This matrix is supplied to two 3 multipliers 1212a and 1212b. The multiplier 1212a converts the input, work, and monitor colors into a color space of a plurality of primary colors of the display unit 1224. Some red, green, and blue input values are also supplied to find the three color components. The max unit of the value 1214 is supplied to a reciprocal comparison table (inverseLUT) 1.216 'The reciprocal comparison table produces a scaling factor that normalizes these rgb (red, green, and blue) values to The maximum value of the same hue angle. This scale factor is multiplied by some red, green, and blue input values to produce a maximum allowable color with the same hue angle as some of the original red, green, and blue values. Multiply 1212b takes the maximum Hue color, and convert it into the color space of the target multi-primary color. Those data are supplied to the maximization unit 1218, which seeks the maximum value of some multi-primary components, and the reciprocal comparison table 1220 will It is converted into a scaling factor multiplied by some special multi-primary values that are rotated from multiplier 1212a. This is 22 1294616 , The result 疋 has been expanded or shrunk to a color point within the target multi-primary color space of the multi-primary display 1224 gamut. However, as described in any of the above-referenced patent applications, the data is on the display Prior to coloring, sub-pixel rendering (spR) unit 1222 may be utilized for sub-pixel rendering. Some of the functional blocks involved in some of the above embodiments may utilize any combination of hardware and/or software, including some parts or The modules are implemented, for example, in one or more memory devices or circuitry. For example, a programmable gate array or similar circuitry can be assembled to implement such functional blocks. In other examples, a microprocessor that executes a program in memory can also implement such functional blocks. Gamut expansion Once the hue angle is calculated (in any suitable manner), it is used as the entering color gamut. Extending an index of the value table. In one embodiment, the output of the gamut extension table can be expressed by a number between 〇 and 2 The numbers may be very close to 1. These numbers can be expressed by a fixed p〇int binary number with a single digit above the binary point. With appropriate backward movement, allow The integer multiplier is used as an inexpensive system to achieve color gamut expansion. This extension can be used to multiply the two components of chrominance without changing the hue or brightness, thus changing the saturation. Design a system (such as a monitor) such that a system can have several X-paths, such as SRBG, YCbCr, S-VHS, etc. 23 1294616 _ For each of these paths, different hardware and/or software It can serve each of several paths, especially if the input of the assumed image data does not require optional processing, for example, YCbCr does not require a chroma/luma separator. Because the space is already in that form. It is also possible to design a system to read the headers in the number of images to identify which image data format is being input, so that some optional subsystems and routines can be bypassed by bypass. . Other suitable embodiments and specific embodiments are also possible for designing a complete system with different wheelings, as well as for handling such different inputs. BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 depicts a real image of a color gamut expansion and/or conversion of the present invention input source image data 102 (e.g., RGB). The raw image data can be converted to a perceptually uniform color space at the gamma unit 103, if desired. This data is passed to the chroma/brightness converter 1〇4. Some chrominance components from converter 〇4 are input into hue angle calculator 106, and the hue angle is thereafter input to gamut expansion block 1 〇8 (can be implemented as if it were looked up in a table or itself as A different square). The output of the block is a gamut expansion number that can be used by some multipliers to scale the two components of chrominance. These chrominance fractions are then output for use by other parts. It will be appreciated that other embodiments of the invention are also possible. As mentioned earlier, gamut expansion can be implemented in a number of suitable ways, including but not limited to lookup methods, computational subsystems, or the like. In addition, some other paths are also possible. The second diagram shows that the output of the excellent domain extension 24 1294616 (10) can be rotated by the chromaticity values of the Akizawa 104; but another alternative is that the gamut expansion unit (10) directly supplies the gamut extension value to some other parts. 1 1 Q. As described above, if the spread value is less than 1, it reduces the saturation of the color 'if the value is equal to 9 and it substantially maintains the same saturation, if the value is larger than the saturation of the color gamut which increases the color. It is also possible to perform some alternative embodiments. In these embodiments, & numerie seale also means that the same color map conversion table is established for the image data, and the color gamut will now be described. An embodiment of the conversion table and how to construct the color gamut conversion table. One possible way to create a gamut conversion table is to span some of the input of the color space and some edges of the wheel's excellent domain. This is often done in CIE chromaticity spaces where some of the edges of the gamut are straight. Figures 2A, 2B, and 2C show three different chromaticity diagrams for three different color spaces (4) i: _ieity diagram). These different chromaticity diagrams will be used as examples of generating a gamut conversion table. Figure 2A depicts the chromaticity diagram for SRGB. Figure 2B shows a CMY output device, such as a chromaticity diagram for a printer. Figure 2C shows a chromaticity diagram of the color gamut for a hypothetical six primary color multi-primary display, where the color gamut is a combination of the two color gamuts of Figures 2A and 2B. It should be noted that the area covered by the multi-primary color gamut of the 2Cth image can in fact be the area where the two triangular color gamuts of the 2A XI 2B diagram simply intersect. In addition, some of the methods and systems disclosed herein can be applied to any of the primary color displays of the 25 1294616 -. One embodiment of establishing a color gamut conversion table is to pre-offline (〇ff-line) by starting at some edge across the color gamut. The resulting data is then used to generate some color gamut conversion tables that are burned on some hardware implementations. The purpose of crossing some edges of each color gamut is to generate an intermediate table of some maximum saturation values. Of course, there can be some saturation relay values that leave the maximum value, which is valuable to the system. This table can be quantized to contain a value for each tone. To facilitate the understanding of this example, it will be assumed that these tones occupy 36 integer degrees around the color gamut. Of course, in another embodiment disclosed in a related and incorporated patent application, the unit of measure of the angle can be varied to result in a number of "angles" around the color gamut of 2, for example 256. "degree". This change in degree will make the implementation of gamut expansion easier in hardware. There are several ways to do this across these gamuts. One way would be to create a color for each hue angle, with the exception of arbitrarily chosen saturation and chrominance, and within a perceptually uniform luminance/chrominance space. These colors can be converted to CIE chromaticity and corrected to fall on the edge of the gamut polygon, which is then converted back to chrominance/luminance to get some maximum saturation value. Another way would be to follow some edges of the gamut polygon in the CIE chromaticity space, linearly step through the gamut of some gamuts and generate some maximum saturation values, and convert one color point set to brightness. / Chromaticity and calculate the hue angle of each color point. If these linear steps are small enough, for each hue angle, several luma/chroma pairs can be generated. All of these pairs can be used as the maximum saturation for that angle for each of the hue angles of 26 1294616 ' * - average saturation. Of course, some of the other saturation values formed, in addition to the average saturation, can be calculated based on the multiplicity of some pairs of chrominance/luminance pairs and used in the gamut conversion table. In addition, some of the tables in the maximum saturation table are created, and there are other ways in which the above two algorithms are just two of many possible embodiments. Figure 3 shows the three maximum saturation tables generated for the three gamut polygons in Figure 2. The horizontal axis of Fig. 3 is the hue angle, and in this example, the angle of the hue angle is continuously from 〇 to 359 degrees. The vertical axis is the distance from the white point in the CIE Lab space. In one embodiment, the system can use real values that are away from the white point distance, but in another embodiment, these real values are not intended to be used directly. As discussed below, a number of different curve ratios can be used. The dotted line (d〇tted Hne)3〇2 in Fig. 3 is the maximum saturation curve for the SRGB color gamut of Fig. 2A, and the dashed lme 304 is the maximum saturation for the CMY color gamut of Fig. 2B. The curve, and solid line 3〇6 is the maximum saturation curve for the multi-primary color gamut of Figure 2C. Their respective curved shapes are the result of perceptual nonlinear uniformity of the CIE Lab color space used in these examples. The ratio of some of these values in these maximum saturation curves can be used to generate some final gamut conversion tables. For example, the ratio of some values of CMY to sRGB produces a color gamut conversion table that converts some of the sRGB values into a color gamut that is more suitable for CMY printers. It should be noted that at some of the hue angles of Figure 3, the sRGB curve falls below the CMY curve' which will cause some gamut conversion values to be greater than one, increasing the saturation of some of the colors with those • hue angles. It should also be noted that at some of the hue angles of Figure 3, Figure 1294616, the sRGB curve falls above the CMY curve, which causes some gamut conversion values to be less than 1, reducing the saturation of some colors with those hue angles. degree. This is the case shown in Figure 4B. In order to choose a different example, the ratio of the multi-primary maximum saturation curve to the sRGB maximum saturation curve in Figure 3 will generate a color gamut conversion table. This table will convert some values of sRGB to one that is more suitable for the six primary colors. The color gamut of the display. It should be noted that at all hue angles of Figure 3, the sRGB curve falls above the multi-primary curve or falls below the multi-primary curve. This will cause some gamut conversion values to be greater than or equal to 1, in all tones. The saturation of the color is often increased at the corners. This is the case shown in Figure 4C. 0 is a color gamut conversion by scaling the brightness and/or saturation, as in the related method and apparatus entitled "Converting the original color space to the red, green, blue and white color space; METHOD AND AND APPARATUS FOR CONVERTING FROM SOURCE COLOR SPACE TO RGBW TARGET COLOR SPACE", discussed in the request, describes a method for detecting some out-of-gamut colors (〇ut_〇f_gamut color) and moving them to the output gamut space. The surface becomes a "good spot" where a good spot can be defined as a color that generally has the same hue as the color outside the gamut. Figure 5 shows a two-dimensional diagram of a simplified example in which an RGBW color gamut is matched in the RGB gamut. Two techniques are specifically described here: color gamut clamping and gamut scaiing. The gamut clamp will move any color between point a and point B to point b on the surface of the RGBW gamut. From 28 1294616, all colors between BLACK (black) and point B are not outside the μ's and they will remain where they were. For example, the color gamut clamp will have no effect on the color at point c. 〇, however, it is advantageous for scaling all colors between BLACK (black) and eight points, which will cause all colors on the BLACK_A line to be scaled so that the color at point A moves to the second map. Point b. Some colors between point A and point B will be moved to the inside of the gamut. Some colors between BLACK and B are also moved, for example, the color at point c is moved to point D.

在某些系統中，輸出色域是一種簡單的幾何事例，而且可能檢測一些域（domain )，其中的某一色域已在其他的域内部，而且計算出比例因數。可是，纟大多數多原色色域中，一些色域形狀太複雜以致於無法檢測那些域，而且那些域不具有簡單的幾何形狀。第6圖展示出一個稍微較複雜的色域對（pair of gamuts)。這是貫穿兩個色域的一個片段圖像，該兩色域也許被畫在某個第三色彩空間，例如clE XYZ内。在CIE XYZ的三維空間内，那些色域的一個可具有扭曲的立方體形狀，然而其他色域爲帶有較多的邊與^ 伸到新原色色彩的脊線（ridges)的更爲複雜的形狀。第6 圖展示出爲何有時輸入色域會延伸到輸出色域之外，以及為何有時輸出色域能延伸到輸入色域之外。這種複雜的相互貫通的情況，在簡單的CIE色度圖上並未展現出，cIE 色度圖主要示出色域的粗略外形，而未展現出色域在一些不同亮度（luminosity)的細節。 29 1294616 如果單獨使用色域鉗位，則介於A1與B1點之間的任何色彩將被移動到B1點，而且介於B1點與black之間的一些色彩，包括C1色彩將被留下而不予觸動。應當注^ 地是，介於A2點與BLACK之間的所有色彩將被留下不= 觸動，此因其已在色域内。但是，如果代替使用色域縮放，A1點處的色彩將被移動到B 1點，而且在介於A1與B丨點之間直線上的所有色彩將被移到色域的内部，介於B1與BLACK之間的所有色彩也被縮放，例如C1點處的色彩將移動到D1點。此外，在，I於BLACK與A2點之間直線上的一些色彩將被擴展，將A2點處的色彩移出其原來的色域之外而到B2點。用來做爲在該直線上所有其他色彩的一個實例的C2點，將也被稍微移至色域之外並至D2點。這兩直線展示出，在某些域内的:些色衫被按比例縮小，以與輸出色域配合，而在其他區域内的一些色彩被按比例放大，以擴展到輸出色域内。這些比例因數對於每個亮度與飽和度的獨特組合是不同的，所以在一個實施例中，這些比例因數可事先計算出來亚儲存在一個對照表内。但是，這個對照表可能很大，對於大多數系統並不實用。因此，另一個實施例將對每個色衫動態地（on_the-fiy )計算比例因數，再對某一顯示系統加以轉換。再次參閱第5圖，一個用於將c點處色彩轉換到d點處色杉的可能比例因數，可以是直線BLACK_B的長度與直線BLACK-A的長度之比。6知的一個RGB色彩c，帶有 1294616 :-些介於0與1之間的分量，可能計算出輪出色彩空間内的兩個點A與B。A點將在色域外，而b %腌 •點將正好在輸出色域的表面上。距離公式被用來找出這兩根直線的長度，然後計算出它們的比值。爲了找出A點，可能對輸入色彩空間内的c點進行「規格化」（normalize )，這可藉由將色彩c的所有分量除以其最大分里來完成。應注意的是，在一個典型的系統中，一些色彩分量係介於0至255之間的整數。在那些情況中，除運算的結果藉由簡單乘以256或左移8個位元。爲了方便解釋，一些色彩分量從〇到i的取值範圍被用於本討論中口人亦應當注思’除運鼻可精由求最大值的倒數，並將色彩C的所有分量乘以該倒數值來實施。一個數的倒數可被儲存在一個對照表（Look-Up Table; LUT)中，並可迅速地從表中取得。當C點規格化之後，其被轉換到輸出色彩空間以得到A點處的色彩。若已知從輸入色彩空間到輸出色彩空間轉換的一個函數爲CONV，返回一個數的倒數的個函數爲INV，與返回一個色彩最大分量的一個函數爲 MAX，則A點的計算公式爲： A = CONV(C*INV(MAX(C))) C〇NV做爲一個函數，可以藉由計算色調角、利用該色調角確定色度三角形數以及利用選擇一個多原色轉換矩陣以執行該矩陣乘運算求得標的色彩空間内某一色彩結果來實現。爲了找出B點，可能將C點轉換到輸出色彩空間，接著將之規袼化。在下列公式中，c點首先被轉換爲一個 31 1294616 - 臨時的C2點，然後再被用來計算β點。 C2 - CONV(C) . B - C2*INV(MAX(C2)) , 因此對於色彩C的一個色域縮放比（scaling ratio)可以是從BLACK到A點的直線長度與從BLACK到B點的直線長度之比。由於BLACK是一個零點，所以A與B點兩個點可視為向量來處理，而且其大小即相當於兩直線的長度。這使得比值爲：籲 R = |B| / |A| 隶後，經過轉換的色彩C 2 (根據上面的公式）被乘以 $亥細放比來計算出第5圖内的色域縮放色彩d。In some systems, the output gamut is a simple geometric example, and it is possible to detect some domains, one of which is already inside the other, and the scaling factor is calculated. However, in most multi-primary gamuts, some gamut shapes are too complex to detect those domains, and those domains do not have simple geometries. Figure 6 shows a slightly more complex pair of gamuts. This is a segment image that runs through two color gamuts that may be drawn in a third color space, such as clE XYZ. In the three-dimensional space of CIE XYZ, one of those gamuts can have a twisted cube shape, while the other gamuts are more complex shapes with more edges and ridges that extend to the new primary color. . Figure 6 shows why sometimes the input gamut extends beyond the output gamut and why sometimes the output gamut extends beyond the input gamut. This complex inter-penetration case does not appear on a simple CIE chromaticity diagram. The cIE chromaticity diagram mainly shows the coarse outline of the color gamut, but does not show the details of the excellent field in some luminosity. 29 1294616 If gamut clamping is used alone, any color between the A1 and B1 points will be moved to point B1, and some color between B1 and black, including C1 color, will be left Do not touch. It should be noted that all colors between A2 and BLACK will be left untouched because they are already in the gamut. However, if instead of using gamut scaling, the color at point A1 will be moved to point B1, and all colors on the line between A1 and B丨 will be moved inside the gamut, between B1 All colors between the BLACK and the BLACK are also scaled, for example the color at the C1 point will move to the D1 point. In addition, some of the colors on the line between I and BLACK and A2 will be expanded, shifting the color at point A2 out of its original color gamut to point B2. The C2 point used as an instance of all other colors on the line will also be moved slightly beyond the gamut to point D2. These two lines show that in some fields: the shades are scaled down to match the output gamut, while some of the colors in other areas are scaled up to extend into the output gamut. These scaling factors are different for each unique combination of brightness and saturation, so in one embodiment, these scaling factors can be calculated in advance and stored in a look-up table. However, this checklist can be large and not practical for most systems. Therefore, another embodiment will calculate the scaling factor dynamically for each color shirt (on_the-fiy) and then convert a certain display system. Referring again to Fig. 5, a possible scaling factor for converting the color at point c to the color point at point d may be the ratio of the length of the straight line BLACK_B to the length of the straight line BLACK-A. A known RGB color c, with 1294616: some components between 0 and 1, may calculate two points A and B in the rounded color space. Point A will be outside the gamut, while b% will be spotted on the surface of the output gamut. The distance formula is used to find the length of the two lines and then calculate their ratio. To find point A, it is possible to "normalize" point c in the input color space by dividing all components of color c by its maximum score. It should be noted that in a typical system, some of the color components are integers between 0 and 255. In those cases, the result of the divide operation is simply multiplied by 256 or shifted left by 8 bits. For the convenience of explanation, the range of values of some color components from 〇 to i is used in this discussion. The mouth should also pay attention to the 'reciprocal of the maximum value, and multiply all the components of the color C by the number. Reverse the value to implement. The reciprocal of a number can be stored in a Look-Up Table (LUT) and can be quickly retrieved from the table. When C point is normalized, it is converted to the output color space to get the color at point A. If it is known that a function from the input color space to the output color space conversion is CONV, the function that returns the reciprocal of a number is INV, and the function that returns the maximum component of a color is MAX, then the calculation formula of point A is: A = CONV(C*INV(MAX(C))) C〇NV is a function that can be performed by calculating the hue angle, determining the number of chroma triangles using the hue angle, and performing a multiplication by selecting a multi-primary color transformation matrix. The operation achieves a certain color result in the target color space. In order to find point B, it is possible to convert point C to the output color space and then normalize it. In the following formula, point c is first converted to a 31 1294616 - temporary C2 point, which is then used to calculate the beta point. C2 - CONV(C) . B - C2*INV(MAX(C2)) , so a color gamut scaling ratio for color C can be the length of the line from BLACK to point A and from BLACK to point B. The ratio of the length of the line. Since BLACK is a zero point, the two points A and B can be treated as vectors, and their size is equivalent to the length of the two lines. This makes the ratio: R = |B| / |A| After the converted color C 2 (according to the above formula) is multiplied by the $ fine ratio to calculate the color gamut zoom color in Figure 5. d.

D = C2*R 應注意的是，計算色彩A與B大小的過程可能在計算上疋卬貝的。第7圖展示出計算這些值的另一個實施例。一根正父的直線（orthogonal line )從A點平行於該色彩的 _ 最大分量向下延伸到p點，利用相似三角形，從black 到B點的距離與從BLACK到A點的距離之比，等同於從 P點到B ’點的距離與從p點到A點的距離之比。沿對角線從BLACK到A點與B點的距離，必須利用距離公式來計算，該距離公式將涉及到乘運算與求平方根運算。但沿直線從P點到A點與B，點的距離，簡單地爲一些色彩的最大分量，這容許我們對縮放比將其方程式簡化爲： R ^ MA.X(B)/MAX(A) 32 1294616 、無論如何，B點被選成落在輸出色域的表面上，因此應當至少有一個分量，該分量具有最大值爲丨。因此數值i 可被用來替代MAX(B)，而一點也不需要真正地計算B點，對所形成的硬體造成可觀的節省。現在R的值變成ΜΑχ(Α) 的倒數，並可從另一對照表中查到·· R = INV(MAX(A)) 第8圖展示出本系統的一種硬體實施方案（hardware implementation ) 800的一個實施例。輸入的一些RGB數值 8〇2可被劃分成三條平行的路徑。底部路徑首先在模組8〇4 中將RGB轉換爲單獨分開的色度/亮度。該色度被用來計算色調角806。該亮度值在此處並未使用，但被儲存起來以用於其他一些實施例中的其他一些步驟。該色調角被使用在一個對照表808中，而且角度對照表8〇8用來確定該色彩落在哪個色度三角形内。這被用來選擇一個用於3χη 矩陣乘法器812a與812b的轉換矩陣（在表81〇内），以將色彩轉換到其他兩條平行的路徑。第8圖上的上部平行路徑尋找原始rgb色彩的最大分量（在模組814中），並從一個倒數對照表816中查出該最大分量的倒數。這個倒數值被乘以原始RGB色彩的所有分量以對其進行規格化。然後該色彩被轉換到輸出色彩空間 (在矩陣乘法器812a内），找到該最大分量（在模組8 i 8 内）’並對這個最大分量進行轉換（在倒數對照表82〇内），以建立色域的縮放比。第8圖上當中的平行路徑，將原始rgb色彩轉換到輸 33 1294616 出色彩空間（在矩陣乘法器812b ^ 八旦毛”_ 内）’然後母個所形成的上部路徑來的色域的縮放比（在—些乘法器似弟8圖展中示出四個輪出原色’但應當瞭解到可以有任忍多個輸出原色。這此一巴冬點被發运到顯示器 (8 2 4 )上以被顯示。在色域轉換的第一個實施例中，如第！圖所示，假定擴展對於所有的亮度大體上爲常數可能是合理的。否則，將產生某些色域外的色彩，而且可能需要進行色域钳位元 (cUmpUg)或縮放’或者進行其他類似的操作。在色域 1換的第二個實施例中，如帛8圖所示，可以將那兩個步驟合併爲一個步驟。在此實施例中’可能要動態地執行色，比的計算，這可被用來做爲—種色域轉換演算法，該演异：大體上也將用輪入色域的所有色彩來填入輸出色域。在來自RGB的RGBW的情況，RGBW色域完全落在應色域的内部。如果該兩色域的邊緣的比，對於每一個輸入的像素數值被計算出，則比值（RGBW/RGB)將總是小於 1 ’亚將把一些色彩按比例縮小到RGB W色域内，如第5 圖所示。在來自RGB的RGBC (C爲青綠色（Cyan)或其他的色彩）的情況，對於某些色彩，RGBC色域是在RGB色域的内部，而且對於其他一些色彩，RGBC色域是在RGB 色域的外部，如第6圖所示。當RGBC在RGB的外部時， β亥比值將大於1 ’且將擴展該色域來配合。當RGBC在RGB 的内部時’該比值將小於1，且將收縮該色域來配合。在以上一些實施例中所涉及的一些功能方塊，可利用硬體及/或軟體的任何組合，包括一些零件或模組，例如以一 34 * 1294616 Λ · 、 - 個或多個記憶體設備或電路系統來實施。例如，一種可程式的閘陣列（Pr〇grammable gate array )或類似的電路系統可被裝配成來實施這樣一些功能方塊。在其他例子中，一個在記憶體内執行程式的微處理機，亦可實施這樣的一些功能方塊。儘管本發明引用一個示範性·實施例來加以描述，但熟習本發明的技藝者應瞭解，在不脫離本發明範疇的情況下，可對本發明做出各種修改，或者對此間的某些元件以均等物加以替代。此外，在不脫離本發明基本範疇的情況下，根據本發明的教示，可對本發明做出許多修改，以適應某一特殊的情況或材料。因此，本發明並非僅限於以考慮作爲實現本發明最佳模式而加以揭示的特殊實施例，而是包括所有落在所附的申請專利範圍内的所有實施例。【圖式簡單說明】第1圖：展示色域轉換/擴展系統及/或方法的一個可能的實施例。第2A、2B與2C圖：描述用於三種不同色彩空間的三個不同色度圖（chromaticity diagram)。第3圖：展示爲第2A、2B與2C圖中三個色域多邊形所產生的三個最大飽和度表。第4A、4B與4C圖：描述不同影像設備的一些色域與不同影像設備的色域壓縮及/或擴展的一些可能的效應。第5圖··爲大體上重疊的兩種色彩的一個色域圖與色域 35 1294616 鉗位及/或縮放的一些效應。第6圖：爲具有相當大非重疊面積的兩種色彩色域的另一個圖與色域鉗位及/或縮放的一些效應。第7圖：爲一種藉由計算一個比例因數將一個色域轉換到另一個色域用的技術的一個可能的實施例。第8圖··爲一種實現將色域從一個色彩空間轉換到另一個色彩空間的系統的一個可能的實施例。第9圖：展示根據本發明的一些原理所製成的一種色調角計算器的一個實施例。第10A與10B圖：展示利用色調角計算色度三角形數目的兩個實施例。第11圖·展不用於色域擴展與多原色轉換的色調角計算器的使用。 " 第12Α圖1展示由_ (紅綠藍白）原色所造成三個三角形區域的色度圖。第12Β圖：爲根據本發明D = C2*R It should be noted that the process of calculating the color A and B size may be calculated on the mussel. Figure 7 shows another embodiment of calculating these values. An orthogonal line extends from point A parallel to the _maximum component of the color to point p, using a similar triangle, the ratio of the distance from black to point B to the distance from BLACK to point A, Equivalent to the ratio of the distance from point P to point B to the distance from point p to point A. Along the diagonal, the distance from BLACK to point A and point B must be calculated using the distance formula, which will involve the multiplication and square root operations. But along the line from point P to point A and point B, the distance is simply the maximum component of some colors, which allows us to simplify the equation for the scaling ratio: R ^ MA.X(B)/MAX(A) 32 1294616 , In any case, point B is chosen to fall on the surface of the output gamut, so there should be at least one component with a maximum of 丨. Therefore, the value i can be used instead of MAX(B), and there is no need to actually calculate point B, which causes considerable savings in the formed hardware. Now the value of R becomes the reciprocal of ΜΑχ(Α) and can be found from another comparison table. R = INV(MAX(A)) Figure 8 shows a hardware implementation of the system. An embodiment of 800. Some of the RGB values entered can be divided into three parallel paths. The bottom path first converts RGB into separate chrominance/brightness in module 8〇4. This chromaticity is used to calculate the hue angle 806. This brightness value is not used here, but is stored for use in some other steps in some other embodiments. The hue angle is used in a look-up table 808, and the angle comparison table 8〇8 is used to determine which chromaticity triangle the color falls within. This is used to select a conversion matrix (in Table 81) for the 3χη matrix multipliers 812a and 812b to convert the color to the other two parallel paths. The upper parallel path on Fig. 8 finds the largest component of the original rgb color (in module 814) and finds the reciprocal of the largest component from a reciprocal lookup table 816. This reciprocal value is multiplied by all the components of the original RGB color to normalize it. The color is then converted to the output color space (within matrix multiplier 812a), the largest component is found (in module 8 i 8)' and this maximum component is converted (in the inverse count table 82〇) to Establish the zoom ratio of the color gamut. The parallel path in Fig. 8 converts the original rgb color to the gamut of the color path (in the matrix multiplier 812b ^ 八毛毛 _)) and then the gamut of the upper path formed by the parent (In the case of some multipliers, the figure shows four rounds of primary colors, but it should be understood that there can be more than one output primary color. This one is sent to the display (8 2 4) In the first embodiment of the color gamut conversion, as shown in the Figure!, it may be reasonable to assume that the extension is substantially constant for all luminances. Otherwise, some color gamut colors will be produced, and It may be necessary to perform a color gamut clamp (cUmpUg) or scaling 'or perform other similar operations. In the second embodiment of gamut 1 change, as shown in Figure 8, the two steps can be combined into one Steps. In this embodiment, it is possible to dynamically perform color, ratio calculations, which can be used as a color gamut conversion algorithm, which will also generally use all colors of the gamut. To fill in the output gamut. In RGBW from RGB In the case, the RGBW color gamut falls entirely within the gamut. If the ratio of the edges of the two gamuts is calculated for each input pixel value, the ratio (RGBW/RGB) will always be less than 1 ' Scale some colors down to the RGB W color gamut, as shown in Figure 5. In the case of RGBC from RGB (C is Cyan or other colors), for some colors, the RGBC color gamut is Inside the RGB gamut, and for some other colors, the RGBC gamut is outside the RGB gamut, as shown in Figure 6. When RGBC is outside RGB, the 亥H ratio will be greater than 1 'and will expand the color The domain is matched. When RGBC is inside RGB, the ratio will be less than 1, and the color gamut will be shrunk to match. Some of the functional blocks involved in some of the above embodiments can utilize any hardware and/or software. Combinations, including some parts or modules, for example implemented in a 34 * 1294616 Λ ·, - or more memory devices or circuitry. For example, a programmable gate array (Pr〇grammable gate array) or the like The circuit system can be assembled into a real Such functional blocks. In other examples, a microprocessor that executes a program in memory may also implement such functional blocks. Although the present invention is described with reference to an exemplary embodiment, it is familiar to the present invention. It will be appreciated by those skilled in the art that various modifications may be made in the present invention, or some of the elements may be replaced by equivalents. Further, without departing from the basic scope of the invention, The invention may be modified to adapt a particular situation or material to the teachings of the present invention. The invention is therefore not limited to the specific embodiments disclosed as the preferred mode of the invention. All of the embodiments are within the scope of the appended claims. BRIEF DESCRIPTION OF THE DRAWINGS Fig. 1: A possible embodiment showing a color gamut conversion/expansion system and/or method. 2A, 2B, and 2C diagrams: Three different chromaticity diagrams for three different color spaces are described. Figure 3: shows the three maximum saturation tables produced for the three gamut polygons in Figures 2A, 2B, and 2C. Figures 4A, 4B, and 4C: Describe some of the possible effects of color gamut compression and/or expansion of some color gamuts and different imaging devices of different imaging devices. Figure 5 is a gamut and gamut of two colors that are roughly overlapping. 35 1294616 Some effects of clamping and/or scaling. Figure 6: Some effects of clamping and/or scaling of another map and color gamut for two color gamuts with fairly large non-overlapping areas. Figure 7: A possible embodiment of a technique for converting one color gamut to another by calculating a scaling factor. Figure 8 is a possible embodiment of a system for implementing a color gamut from one color space to another. Figure 9: shows an embodiment of a hue angle calculator made in accordance with some principles of the present invention. Figures 10A and 10B: Two embodiments showing the calculation of the number of chromatic triangles using the hue angle. Figure 11 shows the use of a hue angle calculator that is not used for color gamut expansion and multi-primary color conversion. " Figure 12 shows the chromaticity diagram of the three triangle regions caused by the _ (red, green, blue and white) primary colors. Figure 12: in accordance with the present invention

一些原理進行色域擴展與多 ί吏用的另一個實施例。【主要元件符號說明】 1G〇系統 103 灰度單元 1 〇6 色調角計算器 110 其他零件 102 原始影像數據 104 色度/亮度轉換器 1〇8 色域擴展 3 02 點線 36 1294616 3 04 虛線 306 實線 402 區域 404 三角形 406 白色點 408 現實世界色彩 410 色彩點 420 監視器色域 422 印表機色域 426 色彩點 427 色彩點 428 色彩點 440 監視器色域 442 多原色色域 444 色彩點 446 色彩點 450 色彩點 802 RGB數值 804 模組 806 色調角 808 對照表 810 表 812a 3xn矩陣乘法器. 812b 3xn矩陣乘法器 814 模組 816 倒數對照表 818 模組 820 倒數對照表 822 乘法器 824 顯示器 900 色調角計算器 902 方塊 904 方塊 906 方塊 908 反正切對照表 910 方塊 912 操作對照表 1000 實施例 1002 原色對照表 1004 比較器 1006 多工器模組 1008 色度三角形數Some embodiments perform another embodiment of color gamut expansion and more. [Main component symbol description] 1G〇 system 103 Grayscale unit 1 〇6 Hue angle calculator 110 Other parts 102 Raw image data 104 Chroma/brightness converter 1〇8 Color gamut expansion 3 02 Dotted line 36 1294616 3 04 Dotted line 306 Solid line 402 Area 404 Triangle 406 White point 408 Real world color 410 Color point 420 Monitor color 422 Printer color 426 Color point 427 Color point 428 Color point 440 Monitor color 442 Multi-primary color gamut 444 Color point 446 Color Point 450 Color Point 802 RGB Value 804 Module 806 Hue Angle 808 Comparison Table 810 Table 812a 3xn Matrix Multiplier. 812b 3xn Matrix Multiplier 814 Module 816 Countdown Table 818 Module 820 Countdown Table 822 Multiplier 824 Display 900 Hue angle calculator 902 Block 904 Block 906 Block 908 Anti-tangential comparison table 910 Block 912 Operational comparison table 1000 Example 1002 Primary color comparison table 1004 Comparator 1006 Multiplexer module 1008 Chromatic triangle number

37 1294616 1010 色調角 1022 對照表 1100 色域管線 1102 數位電視YCrCb信號 1106 色調角轉換器 1110 輸入灰度對照表 1114 多原色矩陣對照表 1118 子像素着色模組 1200 色域管線實施例 1204 色度/亮度值 1208角度三角形單元 1212A3XI1乘法器 1214 最大化單元 1218 最大化單元 1222 子像素着色單元 1020 實施例 1024 色度三角形數 . 1104 多原色顯示器 1108 色域擴展對照表 1112 模組 1116 模組 φ 1120 輸出灰度表 1202 紅綠藍數值 1206色調角計算器 1210 多原色轉換矩陣 1212B3xn乘法器 1216 倒數對照表 1220倒數對照表 _ 1224 顯示單元 3837 1294616 1010 Hue angle 1022 Comparison table 1100 Gamut line 1102 Digital TV YCrCb signal 1106 Hue angle converter 1110 Input gray scale comparison table 1114 Multi-primary matrix comparison table 1118 Sub-pixel coloring module 1200 Color-domain pipeline embodiment 1204 Chroma / Luminance value 1208 angle triangle unit 1212A3XI1 multiplier 1214 maximization unit 1218 maximization unit 1222 sub-pixel shading unit 1020 embodiment 1024 chroma triangle number. 1104 multi-primary color display 1108 color gamut expansion table 1112 module 1116 module φ 1120 output Gray scale table 1202 red green blue value 1206 hue angle calculator 1210 multi-primary color conversion matrix 1212B3xn multiplier 1216 countdown table 1220 countdown table _ 1224 display unit 38

Claims

1294616 十、申請專利範圍： 1、一種色域轉換系統，其包含：一輸入通道裝置，接收原始影像數據；一灰度單元，將該原始影像數據轉換成知覺上均勻的空間數據；一色度/壳度單元，將該知覺上均勻的空間數據轉換成一包含一些色度與亮度分量的袼式；一色調角計算器，從該色度/亮度單元接收至少該等色度分量，並根據所接收的一些分量對該原始影像數據計算一些色調角；一色域轉換單元，導出一些色域轉換值，以利用一色域轉換表應用到該原始影像數據的該等分量。 2'依申請專利範圍第1項之色域轉換系統，其中如果該原始影像數據已成爲一包含一些色度與亮度分量的格式，該色度/亮度單元可旁路（bypass)該原始影像數據。 3、依申請專利範圍第1項之色域轉換系統，其中該色調角計算器以從〇到2的某次方的角度計算出一些色調角。 4、依申請專利範圍第1項之色域轉換系統，其中該色域轉換表包含一具有一些於離線狀態所計算出的色域轉換值的對照表。 5、依申請專利範圍第1項之色域轉換系統，其中該等色域轉換值係藉由橫跨該被施加轉換的系統的複數個色域的一些邊緣來確定。 39 !294616 « * 、 ' 6、依申請專利範圍第5項之色域轉換系統，其中該被施加轉換的系統的該等色域包含複數個群，每—群包含：一些輸入色域、一些監視器色域與一些多原色色域。 7、依申請專利範圍第5項之色域轉換系統，其中該等色域轉換值係藉由橫跨一色域的圍繞邊緣而被計算出來，以產生一些飽和度值。 8、依申請專利範圍第7項之色域轉換系統，其中該等色域轉換值對於每一色調角，在一知覺上均勻的色度/亮度空間内產生一色彩。 9、依申請專利範圍第8項之色域轉換系統，其中該色域轉換單元可對於每一色調角來任意地選擇一些飽和度與色度值、將該色彩轉換爲CIE色度、校正該所形成的色彩使其大體上落在該色域多邊形的邊緣上以及將該色彩轉換爲色度/亮度。 10、依申請專利範圍第7項之色域轉換系統，其中該等色域轉換值係沿着該色域的邊緣被產生，而且對沿着該邊緣的一點集（set of points)，產生色度/亮度與色調角的數據。 11、依申請專利範圍第10項之色域轉換系統，其中對每一色調角，產生複數個色度/亮度數據。 12、依申請專利範圍第1〇項之色域轉換系統，其中一些飽和度值之比被計算出，以將一色域空間轉換爲另一色域空間。 13、一種色域轉換單元，其包含： 1294616 14 15 16 17 18 19 一裝置，用以產生一些色域轉換數值，該等數值係藉由%複數個被或單元施加轉換的色域的一些邊緣而被計算出來。 ‘依申請專利範圍第13項之色域轉換單元，其中該等色域包含複數個群，每一群包含：一些標準色域、一些監視器色域與一些多原色色域。、依申請專利範圍第13項之色域轉換單元，其中該等數值係藉由橫跨一色域的圍繞邊緣被計算出來，以產生一些飽和度值。、依申請專利範圍第13項之色域轉換單元，其中該等數值對於每一色調角，在一知覺上均勻的色度/亮度空間内產生一色彩。、依申請專利範圍第16項之色域轉換單元，其中該裝置對於每一色調角，選擇一些飽和度與色度值、將該色彩轉換爲CIE色度、校正該所形成的色彩使其大體上落在該色域多邊形的邊緣上以及將該色彩轉換返回爲色度/亮度。、依申請專利範圍第1 6項之色域轉換單元，其中該等數值係沿着該色域的邊緣被產生，而且沿着該邊緣的一點集（set of points)，產生色度/亮度與色調角的數據。、依申請專利範圍第18項之色域轉換單元，其中對每一色調角，產生複數個色度/亮度數據。、依申請專利範圍第18項之色域轉換單元，其中一些飽和度值之比被計算出，以將一色域空間轉換爲另一色 20 1294616 . 一種用於將該第二色彩點乘上該比例因數之裝置。 29、依申請專利範圍g 28項之影像處理***，另包含：一種用於選擇一第一色彩點的一最大分量之装置；種用於計异該最大分量的倒數之裝置；及一種用於將該倒數乘上該第一色彩點的一些分裝置。一 30、依申請專利範圍第28項之影像處理系統，另包含：一種用於從一第一色彩點的一些色度分量計算一色調角之裝置；種用於從該色調角計算一色度三角形之裝置· 一種用於使用該色度三角形選擇—色彩空間轉換矩陣之裝置；及種用於將該第一邊緣點乘上該轉換矩陣之穿置。 31、依申請專利範圍第28項之影像處理系統，其中該比例因數將該第一色彩點收縮到一第二色彩空間内。 32、依申請專利範圍第28項之影像處理系統，其中該比例因數將該第一色彩點擴展到一第二色彩空間内。Λ 33、一種色調角計算器，其包含：至少一個輪入.通道，接收影像數據的一些色度分量· 一個八分圓確定單元，確定某一給定的影像數栌哪一個八分圓；一角度確定單元，確定該給定的影像數據的一色調角 45 1294616 在某一個八分圓内；及一角度補償單元’根據已確定的該影像數據所佔據的該八分圓，對該影像數據進行該色調角調節。 34、依申請專利範圍第33項之色調角計算器，其中該八分圓確定單元包含：一絕對值單元，確定該等色度分量的絕對值；一交換單元，根據一條件對該等色度分量進行交換；及一個八分圓記憶體，根據該絕對值單元與該交換單元的一些結果，儲存該影像數據的八分圓。 35、依申請專利範圍第33項之色調角計算器，其中該角度確定單元包含：一除法器，對該等色度分量進行除運算；及一反正切確定單元，磙定該影像數據的角度在一個八分圓内。 36、依申請專利範圍第33項之色調角計算器，其中該角度補償單元包含：一個八分圓補償單元，根據已確定的該影像數據所佔據的八分圓，藉由一補償值對該角度確定單元所輸出的該色調角進行調節。 37 ; —種在一影像處理單元中計算色調角之方法，其包含步驟：將一些色調角儲存爲一帶有符號位元的一正的二進 1294616 位數字； :將-些色調角相乘時，將該等做爲兩數子的色調角相乘，的一進位 (XOR)操作；及 '對其各自的符號位元進行異當將該等色調角相加時，檢測料色元，並在相加時對任何負的色調角進行反運算 38、之方象處理單元中計算-影像數據的色調角之方法，其包含步驟：匕门月汁异-輸入影像數據的色調角位在哪個八分圓内. 計算在該影像數據所占據的八分圓内的該角；及根據該影像數;^ , 加到在哪m將—補償值在㈣像所占據的八分圓内的該角上。元::::Γ觀圍第38項之方法，其中該影像處理單據加以儲存。的色調角被計算出來，並做爲#色影像數 40 種影像處理單元，其包含：形數；及一選擇器，根據輸入影像點的色調角，選擇形數。色度三角形存儲區，爲複數個色調角儲存一個三角個三角 41 依申明專利範圍第40項之鸟# a +上貝之衫像處理早70，其中該等色调角根據圍繞一個圓& U馬2的某次方的度數而被計算出來。 47 1294616 . 42、 43 44 45 依申請專利範圍第41項之影像處理單元，其中圍繞一個圓的該度數爲256度。一種影像處理單元，其包含·· 複數個原色’影像數據可利用其着色； -色調角計算器’其内儲存有該等原色的色調角，供該影像數據着色用；複數個比較器，將該等原色的色調角與一輸入影像點的色调角進行比較；及，擇裔一埯擇一色度三角形，其中該輸入影像點即位在該色度三角形内。 ‘一種影像處理系統，其包含： -輸入裝置’接收影像數據的一些色度分量；及 =路系統’與該輸人裝置相連接，該電路系統確-某-給定的影像數據占據哪個的影像數據的一色喟备左^^ 定的η 色调角在-個八分圓内以及根據已確像數據所佔據的人分圓對該影像數據调角進行調節。匕路利範圍第44項之影像處理系統，其中該電等色度二=以分量的料值、根據某-條件對該 …0換歧根據該絕對值單元與該交換早兀的—些結果’將該影像的人分圓加以财卜、、，申請專利範圍第44項之影像統路系統對該等色度分量、中該電仃除運异，亚確定該影像數據 48 46 1294616 的色调角在一個八分圓内。 47、依申請專利範圍第44項之与你各路系統根據已確定的^％像處理系統，其中該電 -補償值對該角度確定單元所佔據的八分圓，藉由 48 斤輪出的色調角進行調節。 -種計算色調角之影像處理單元，其包含：些色調角儲存爲—帶有符號位元的-正的一進位數字之裝置；一種將做爲兩個正的—進乘的-進位數子的該等色調角進行 n、各自的符號位元進行異（職)操作之裝置，及 1種檢測該等色調角的符號位元，並在相加該等色調角時對任何負的色調角進行反運算之裝置。依申凊專利範圍第48項之影像處理單元，另包含：種3十鼻一輸入影像數據的色調角位在哪個八分圓内之裝置；種计异在該影像數據所占據的八分圓内的該角之裝置；及 a 一種根據該影像數據所占據的是哪個八分圓，將一補 4貝值加到在該影像所占據的八分圓内的該角上之裝置。〇依申凊專利範圍第49項之影像處理單元，其中該影像處理單元的—些原色的色調角被計算出，並加以儲存以供着色影像數據用。 491294616 X. Patent application scope: 1. A color gamut conversion system, comprising: an input channel device for receiving original image data; a gray scale unit for converting the original image data into perceptually uniform spatial data; a shell unit that converts the perceptually uniform spatial data into a modality containing some chrominance and luminance components; a hue angle calculator that receives at least the chrominance components from the chrominance/luminance unit and receives Some components calculate some hue angles for the original image data; a color gamut conversion unit derives some color gamut conversion values to apply to the components of the original image data using a color gamut conversion table. 2' The color gamut conversion system of claim 1, wherein the chrominance/luminance unit bypasses the original image data if the raw image data has become a format containing some chrominance and luminance components . 3. A color gamut conversion system according to item 1 of the patent application scope, wherein the hue angle calculator calculates some hue angles from a certain angle from 〇 to 2. 4. The color gamut conversion system according to item 1 of the patent application scope, wherein the color gamut conversion table comprises a comparison table having some color gamut conversion values calculated in an offline state. 5. A color gamut conversion system according to claim 1 of the patent application, wherein the color gamut conversion values are determined by traversing edges of a plurality of color gamuts of the system to which the conversion is applied. 39 !294616 « * , ' 6. The color gamut conversion system according to item 5 of the patent application scope, wherein the color gamut of the system to which the conversion is applied comprises a plurality of groups, each group includes: some input color gamuts, some Monitor gamut with some multi-primary gamuts. 7. A color gamut conversion system according to item 5 of the scope of the patent application, wherein the color gamut conversion values are calculated by surrounding the edges of a color gamut to produce some saturation values. 8. A color gamut conversion system according to clause 7 of the patent application, wherein the color gamut conversion values produce a color in a perceptually uniform chromaticity/luminance space for each hue angle. 9. The color gamut conversion system according to item 8 of the patent application scope, wherein the color gamut conversion unit can arbitrarily select some saturation and chrominance values for each hue angle, convert the color to CIE chromaticity, and correct the The resulting color is such that it falls substantially on the edge of the gamut polygon and converts the color to chrominance/luminance. 10. The color gamut conversion system according to item 7 of the patent application scope, wherein the color gamut conversion values are generated along an edge of the color gamut, and a color is generated for a set of points along the edge. Degree/luminance and hue angle data. 11. A color gamut conversion system according to claim 10, wherein a plurality of chrominance/luminance data is generated for each hue angle. 12. A color gamut conversion system according to the first aspect of the patent application, wherein a ratio of saturation values is calculated to convert a gamut space into another gamut space. 13. A color gamut conversion unit comprising: 1294616 14 15 16 17 18 19 a device for generating a number of color gamut conversion values by applying a number of edges of a converted color gamut by a plurality of or units It was calculated. The color gamut conversion unit according to item 13 of the patent application scope, wherein the color gamut comprises a plurality of groups, each group comprising: some standard color gamuts, some monitor color gamuts and some multi-primary color gamuts. The color gamut conversion unit according to item 13 of the patent application scope, wherein the numerical values are calculated by surrounding edges around a color gamut to generate some saturation values. The color gamut conversion unit according to claim 13 of the patent application, wherein the values produce a color in a perceptually uniform chromaticity/luminance space for each hue angle. According to the color gamut conversion unit of claim 16, wherein the device selects some saturation and chrominance values for each hue angle, converts the color into CIE chromaticity, and corrects the formed color to make it substantially Drops on the edge of the gamut polygon and returns the color conversion to chrominance/luminance. a color gamut conversion unit according to item 16 of the patent application scope, wherein the numerical values are generated along the edge of the color gamut, and a set of points along the edge produces a chromaticity/brightness Hue angle data. According to the color gamut conversion unit of claim 18, wherein a plurality of chromaticity/luminance data is generated for each hue angle. According to the color gamut conversion unit of claim 18, some ratios of saturation values are calculated to convert one color gamut space into another color 20 1294616. One is used to multiply the second color point by the ratio Factor device. 29. The image processing system according to claim 28, further comprising: a device for selecting a maximum component of a first color point; a device for counting the reciprocal of the maximum component; and a device for The reciprocal is multiplied by some of the sub-devices of the first color point. 30. The image processing system according to claim 28, further comprising: a device for calculating a hue angle from some chrominance components of a first color point; and a method for calculating a chromaticity triangle from the hue angle Apparatus - A means for using the chromaticity triangle selection - a color space conversion matrix; and a means for multiplying the first edge point by the transformation matrix. 31. The image processing system of claim 28, wherein the ratio factor shrinks the first color point into a second color space. 32. The image processing system of claim 28, wherein the ratio factor extends the first color point into a second color space. Λ 33, a hue angle calculator, comprising: at least one wheeled channel, receiving some chrominance components of the image data, an octant determining unit, determining which octet of a given image number; An angle determining unit determines a tonal angle 45 1294616 of the given image data within an octant; and an angle compensation unit 'according to the determined octant occupied by the image data, the image The data is subjected to the hue angle adjustment. 34. The hue angle calculator according to claim 33, wherein the octant determining unit comprises: an absolute value unit that determines an absolute value of the chrominance components; and an exchange unit that colors the color according to a condition The degree component is exchanged; and an octant memory stores an octant of the image data according to the result of the absolute value unit and the exchange unit. 35. The hue angle calculator according to claim 33, wherein the angle determining unit comprises: a divider for performing division operations on the chrominance components; and an arctangent determining unit for determining an angle of the image data In an eight-point circle. 36. The tonal angle calculator according to claim 33, wherein the angle compensation unit comprises: an octant compensation unit, according to the determined octant occupied by the image data, by a compensation value The hue angle output by the angle determining unit is adjusted. 37: A method for calculating a hue angle in an image processing unit, comprising the steps of: storing some hue angles as a positive binary 1294616 digit with a sign bit; : multiplying the hue angles , as a carry-over (XOR) operation of multiplying the hue angles of the two numbers; and 'differently adding the hue angles to the respective sign bits, detecting the chromatic elements, and A method of inversely calculating any of the negative hue angles at the time of addition, 38, a method of calculating a hue angle of the image data in the processing unit, comprising the steps of: the hue angle of the image data of the input image is Within the octant circle, calculate the angle within the octant occupied by the image data; and according to the number of images; ^, to which m is to be - the compensation value is within the octant occupied by the (4) image On the corner. Yuan:::: Method of item 38 of Guanuanwei, in which the image processing document is stored. The hue angle is calculated and used as the #color image number 40 image processing units, which include: a number; and a selector that selects the shape according to the hue angle of the input image point. Chromatic triangle storage area, storing a triangular triangle for a plurality of hue angles 41 According to the claim patent range 40th bird # a + upper shell image treatment early 70, wherein the hue angle is based on a circle & U The degree of the power of the horse 2 is calculated. 47 1294616 . 42, 43 44 45 The image processing unit according to item 41 of the patent application, wherein the degree around a circle is 256 degrees. An image processing unit comprising: a plurality of primary colors 'image data can be colored by using the same; - a hue angle calculator' storing therein a hue angle of the primary colors for coloring the image data; a plurality of comparators The hue angles of the primary colors are compared to the hue angles of an input image point; and the selected color point triangle is selected, wherein the input image point is within the chromaticity triangle. An image processing system comprising: - an input device 'receives some chrominance components of image data; and = a system </ RTI> connected to the input device, the circuit system - which - a given image data occupies which The color data of the image data is adjusted to the left of the η hue angle within an octant and the image data is adjusted according to the circle of people occupied by the confirmed image data. The image processing system of the 44th item of the 匕路利 range, wherein the unequal chromaticity 2 is the value of the component, and the result of the gamma according to a certain condition is based on the result of the absolute value unit and the exchange 'The image of the person is divided into a wealth, and the image system of the 44th patent application scope is used to determine the color of the chromatic component, the color of the image data 48 46 1294616 The corner is in an octant. 47. According to the 44th item of the patent application scope and your system, according to the determined image processing system, the electric-compensation value is occupied by the angle determining unit, which is rounded by 48 kg. The hue angle is adjusted. An image processing unit for calculating a hue angle, comprising: a plurality of hue angles stored as - a positive carry-number device with a sign bit; one will be treated as two positive - multiply - carry-in bits The hue angles are n, the respective sign bit is operated by the device, and a sign bit detecting the hue angle, and any negative hue angle is added when the hue angles are added A device that performs an inverse operation. The image processing unit of claim 48 of the patent application scope further includes: a device for arranging the tonal angle of the input image data in which octant circle; the octagonal circle occupied by the image data The device of the corner; and a means for adding a B-norm value to the corner within the octant occupied by the image based on which octant the image data occupies. The image processing unit of claim 49, wherein the hue angles of the primary colors of the image processing unit are calculated and stored for coloring image data. 49