網路的蓬勃發展，使得人們透過網路傳輸資料的頻率愈來愈頻繁，許多網路上的彩色圖片廣為流傳，但一張彩色圖片若不經過處理，需要不小的儲存空間，所以透過色彩量化的技術，將圖像做一定比例的壓縮，在可接受的品質損失下，顯示出與原圖差異不大的圖像，也可節省儲存空間，加速在網路上傳輸的速率。
本論文提出一個新的彩色影像壓縮技術稱為LBG-CD，LBG-CD可改進LBG設計調色盤的方式，加入了過濾重複像素來降低訓練像素的數量，以及在初始調色盤的產生改用細胞分裂演算法來取代原本LBG演算法以隨機的方式所產生的初始調色盤，藉此來提升壓縮品質及降低壓縮圖像時所耗費的時間。

Because the rapid development of Internet, people transport data via Internet are more and more frequent. Many full-color photographs on the Internet are widespread, but a full-color photograph needs a large storage space if it does not perform by image processing. Therefore, image conducts a certain percentage of compression by the color quantization scheme at an acceptable loss of quality to show little different image compared with original image. It can not only save storage space, but also accelerate the network transmission rate.
This thesis proposes a color image compression technique called LBG-CD which improves the design of palette in the LBG by adding a filter to reduce duplication of training pixels of the number of pixels, as well as in the initial selection of color palettes to use cell division to replace the original LBG algorithm to generate a random way of the initial color palette in order to enhance compression quality and time cost of image compression.

摘 要 I
Abstract II
謝 誌 III
目 錄 IV
圖表索引 V
第1章 緒論 1
1.1 研究背景與動機 1
1.2 研究目的 2
1.3 研究範圍與限制 2
1.4 研究流程 2
1.5 論文架構 4
第2章 文獻探討 5
2.1 彩色影像簡介 5
2.2 影像壓縮技術簡介 6
2.2.1 無失真壓縮技術 ( Lossless compression ) 8
2.2.2 失真壓縮技術( Lossy compression ) 9
2.3 色彩量化簡介 11
2.4 色彩量化方法介紹 13
2.4.1 LBG演算法 13
2.4.2 Hsieh和Fan所提出的適應性分群演算法 16
2.4.3 Color Finite-State LBG(CFSLBG)演算法 18
2.5 影像品質評估標準 20
2.5.1 均方誤差( Mean Square Error, MSE ) 21
2.5.2 尖峰訊號與雜訊比(Peak Signal to Noise Ratio, PSNR) 21
第3章 研究方法 22
3.1 思考流程 22
3.2 LBG-CD演算法 24
3.2.1 過濾重複像素 24
3.2.2 細胞分裂法 25
3.2.3 LBG-CD演算法流程 26
第4章 實驗結果與討論 28
4.1 相關演算法實驗參數之設定 29
4.2 實驗結果 29
4.2.1 Lena圖測試結果 29
4.2.2 Airplane測試結果 33
4.2.3 Peppers測試結果 36
4.2.4 Tiffany測試結果 39
4.2.5 House圖測試結果 42
第5章 結論與未來展望 45
5.1 結論 45
5.2 未來展望 46
參考文獻 47
作者簡介 49

[1] Braquelaire, J. P. and Brun, L., “Comparison and optimization of methods of color image quantization,” IEEE Transactions on Image Processing, Vol. 16, No. 7, pp. 1048-1052, Jul. 1997.
[2] Cao, H. Q. and Li, W., “A Fast Search Algorithm for Vector Quantization Using a Directed Graph,” IEEE Transactions on Circuit and Systems for Video Technology, Vol. 10, No. 4, Jun. 2000.
[3] Chang, C. C., Chou, Y. C. and Shen, J. J., “A Discrete Wavelet Transform Based State-codebook Search Algorithm for Vector Quantization,” Proceedings of the First International Conference on Innovative Computing, Information and Control(ICICIC`06) Vol. 1, pp. 197-200, Aug. 2006.
[4] Chang, C.-H., Xu, P., Xiao, R. and T., S., “New Adaptive Color Quantization Method Based on Self-Organizing Maps,” IEEE Transactions on Neural Networks, Vol. 16, No. 1, pp. 237 - 249, Jan. 2005.
[5] Hsieh, I.S. and Fan, K.C., “An Adaptive Clustering Algorithm for Color Quantization,” Pattern Recognition Letters, Vol. 21, No.4, pp. 337-346, 2000.
[6] Huang, Y.L. ang Chang, R.f., “A Fast Finite-State Algorithm for Generating RGB Palettes of Color Quantized Images,” Journal of Information Science and Engineering, Vol.20, No.4, pp. 771-782, 2004.
[7] Linde, Y., Buzo, A. and Gray, R. M., “An algorithm for vector quantization design,” IEEE Transactions on Communications, Vol. COM-28, No. 1, pp. 84-95, Jan.1980.
[8] Liu, S. H. and Lin, J. S., “Vector quantization in DCT domain using fuzzy possibilistic c-means based on penalized ang compensated constraints,” Pattern Recognition, Vol. 35, pp. 2201-2211, 2002.
[9] Tagdizen, T., Akarun, L. and Ersoy, C., “Color quantization with genetic algorithms,” Signal Processing: Image Communication, Vol. 12, pp. 49-57, 1998.
[10] Wu, Y., Yang, C. and Wang, T., “A new approach of color quantization of image based on neural network,” International Joint Conference on Neural Networks. Vol. 2, pp. 973-977, Jul. 2001.