| Summary: | 碩士 === 國立臺灣科技大學 === 光電工程研究所 === 101 === In this thesis, we proposed the gamut similarity descriptions and gamut mapping algorithms among different displays. In the gamut similarity part, we used the shape and area information to compare the color gamut similarity between multi-displays which are under the CIE L*a*b* and CIE L*C*h color space. The shape compare method use the polar quantization to get the shape similarity, and the area compare method uses the intersection of gamut to calculate gamut area similarity.
We used the non-linear tone compression and clipping methods for designing various Gamut Mapping Algorithms (GMAs). In the non-linear tone compression, we divided the color gamut into six main hue pages: [R],[Y],[G],[C],[B] and [M], and use the first-order regression model and non-linear compression ratio to store the interpolation matrices for the gamut mapping. The clipping algorithm could divided the lightness of input pixels into adjusted lightness or not, and use the information of the six boundaries of small gamut to get the destination color which could be used for determine whether the input pixel is out or in the boundary of gamut. The psychophysical experiment shows the improved clipping models are better for the natural images.
|
|---|
