Digital Color Image Quantization Methods Based on Various Color Spaces

[[abstract]]在本論文中，基於各種色彩空間的色彩量化技術被探討。首先，種種不同的 色彩空間表示被描述，共包括RGB，YCbCr，HSI 和CMY 等空間；然後，三種 量化演算法被陳述，共包括中間值切割法，K-means 演算法和模糊C-means 演算 法；其次峰值信雜比(SNR)被用來評估三種量化演算法在各種色彩空間的量化性 能；最後，使用各種彩色影像來實驗示範三種彩色量化方法的有效性。[[abstract]]In this thesis, color image quantization based on various color spaces are investigated.First, several color space representations are described including RGB, YCbCr , HSI and CMY spaces etc. Then, three color quantization algorithms are presented including Median-Cut algorithm, K-Means algorithm and Fuzzy C-means algorithm. Next, the peak signal-to-noise ratio (SNR) is used to evaluate the performance of three color quantization algorithms in various color spaces. Finally, experimental results are conducted to demonstrate the effectiveness of three color quantization algorithms by using various images