| Summary: | 博士 === 國立成功大學 === 光電科學與工程學系 === 102 === With the development in science and technology, the optimal performance for a display is essential. The image quality is one of the significant parameters to evaluate the performance of displays. Various methods are proposed to improve rendering image qualities such as multi-primary color mixing, and pixel pattern designed. The multi-primary can show more rich colors in rendering images due to four or more primary colors are employed to extend the color gamut. Various pixel patterns have different qualitative influences on the rendering images because of their different intensity distributions in the spatial frequency space. In this thesis, first of all, we focus on the arrangements of traditional RGB sub-pixel which have influence on image quality. Including analysis for modulation transfer function (MTF), pseudo colors and the performance for rendering images in high frequency condition. From the results, we can find the Green sub-pixel at the center in a pixel shows better performance than others. Second, we study the problem which promotes the luminance for displays. Adding the white sub-pixel to a pixel is one of the effective methods to enhance the luminance. In this part, we focus on the conversion algorithm from three primary colors to four primary colors. Compare with other algorithms, we can find the proposed conversion algorithm can enhance luminance without significant color shift by means of simulation and instrumental measurement on rendering images. The pseudo color phenomenon is also analyzed and discussed on the basic of human vision model. The study indicated that image performance is better when using quad RGBW with W and G sub-pixels in diagonal positions rather than them in non-diagonal positions. Except for the luminance, how to extend wide color gamut is important. In this part, we proposed a method of color mapping conversion between RGB and RGBCW to improve image luminance and maintain minimal color shift. We measured and simulated the rendered images to prove the capabilities of our proposed method. In study, the RGB-to-RGBCW mapping algorithm is designed so that the hue and saturation remains unchanged from the original color and images with higher luminance than RGB system. We also comparatively evaluate the image performance for two potential types of multi-primary color LCDs with stripe RGBCW and RGBCY pixel units, respectively. According to our simulation study and experimental results, we proved that the colors in RGBCW are similar to the original colors in RGB pixel unit including hue and saturation, which also show higher luminance than those in the RGBCY unit.
|
|---|
