摘 要:針對由樣本集直接建立的三維查找表數(shù)據(jù)不均勻,不能滿足ICC Profile 規(guī)范要求的問題,提出了反向均勻化的解決方法。重點討論了反向均勻化過程中快速定位四面體、色域邊界點外推技術(shù)以及色域邊界外點的色域匹配等幾個關(guān)鍵技術(shù)。根據(jù)LAB空間采樣數(shù)據(jù)特點給出了一種快速定位四面體的算法,利用圖像邊界處理的外推思想得出了色域邊界點插值模型,提高了插值的校正速度;討論了色域匹配的裁剪匹配算法并提出了利用二維查找表來提高算法效率的思想,縮短了判斷時間,降低了算法復(fù)雜度。最后通過實驗驗證了反向均勻化算法,給出了實驗效果圖。
關(guān)鍵詞:反向均勻化; 三維插值; 外推技術(shù); 色域匹配
中圖分類號:TP391 文獻標識碼:A
文章編號:1004-373X(2010)12-0125-04
Key Technologies on Inverse Uniformity for Three Dimensional Interpolation
LI Xiang-wei
(Xi’an Institute of Post Telecommunications, Xi’an 710121, China)
Abstract:The inverse uniformity method is proposed to solve the problem that 3D look-up table created directly by color patch sample is usually not uniform, which can not meet the ICC Profile specification. Three key technologies are discussed in detail: fast searching tetrahedron, extrapolation of gamut boundary points and gamut mapping. A fast searching tetrahedron algorithm is given based on the characteristics of LAB spatial sampling data. The interpolation model of gamut boundary points is provided by extrapolation ideas of image boundary treatment ideas, the corrected speed of interpolation is improved. The matching cut algorithm of gamut mapping is studied, and the thinking of using two-dimensional look-up table to improve the efficiency of algorithm are stated, which shortened the time to determine and reduce the complexity of the algorithm. The inverse uniformity algorithm is verified by experiments, and itsexperimental results figures are given finally.
Keywords:inverse uniformity; three dimensional interpolation; extrapolation; gamut mapping
0 引 言
為解決各種成像設(shè)備如顯示器、打印機等等之間成像色彩的一致性, 色彩管理技術(shù)通過引入與設(shè)備無關(guān)的顏色空間,建立設(shè)備的彩色特性描述,并按用戶要求進行顏色空間轉(zhuǎn)換、色域匹配(彩色增強)、以及非線性特性校正等軟件技術(shù)方法來彌補現(xiàn)有硬件設(shè)備的不足,使其可給出用戶滿意的高品質(zhì)、與設(shè)備無關(guān)(所見即所得)的彩色再現(xiàn)[1-2]。彩色校正的主要任務(wù)是補償外設(shè)非線性產(chǎn)生的偏差,其實現(xiàn)方法可以劃分為多重回歸、三維插值、神經(jīng)網(wǎng)絡(luò)和模糊邏輯等[1,3-4]。
作為三維插值的一種算法,四面體插值算法由于其在實現(xiàn)代價和計算速度上的優(yōu)勢已廣泛應(yīng)用于實際的設(shè)備校正中[1]?!?br>