Design And Implementation Of Gamut Extension Techniques

Gamut mapping techniques have become a research hotspot in color man-agement as people begin to understand gradually the reason why color distortionhappens during image duplicating between di?erent devices. Gamut mappingtechniques are divided into two parts, one is mapping from wide gamut to smallgamut named gamut clip&compression and the other is mapping from smallgamut to wide gamut named gamut extension. Gamut clip&compression tech-niques are widely researched for a long time so that gamut mapping become thename of gamut clip&compression. However, as wide gamut display is put intouse, how to make full use of the color representation ability of wide gamut dis-play become a critical issue. Therefore, people begin to realize the importanceof gamut extension. Using gamut extension techniques, the color representationability of wide gamut devices can be utilized to display small gamut images morevividly, which provides high-quality pictures.The main goal of this dissertation is to do research on gamut extensionalgorithms and implementation methods. Based on analysis of current gamutextension algorithms, in this dissertation, we presented analysis on transitiongamut boundary and mapping types, proposed and verified piecewise linear tran-sition gamut boundary is a better transition gamut boundary, made comparisonsamong three di?erent algorithms(linear, piecewise linear and nonlinear), and pro-posed the solution for the issue how to do gamut extension well for certain rangeof hue. Other than the common 3-D look up table and neural network im-plementation methods, look up table and arithmetic logic units was used toimplement gamut extension system which only needs certain range of hue pro-cessed, and two implementation schemes for linear and nonlinear algorithms wereproposed,furthermore,synthesis results and discussion were presented. Goodsimulation results of the system are obtained, and the hardware implementa-tion is high-precision, configurable and scalable, moreover, the system works inpipeline, and the goal of high-definition video realtime processing is achieved.