臺灣博碩士論文加值系統

在本篇論文中，我們提出JPEG影像解碼硬體化的設計，JPEG的解碼設計流程首先是JPEG標頭的處理，接著處理可變長度解碼以及反量化，再經由反Zig-Zag掃描將資料重新排列，並使用二維離散餘弦逆轉換將資料由頻率域的值轉回空間域的值，最後則是將YCbCr的色彩座標值轉回RGB色彩座標值，上述部分均由硬體描述語言(Verilog)完成。JPEG讀檔掃描，以及解碼後的RGB資料加入BMP標頭檔使之成為BMP影像這兩部分則是高階程式語言(C language)完成。
由實驗的數據中得知JPEG硬體解碼的速度比軟體解碼的速度平均快了2.9至4.3倍左右，當輸入的圖片像素越大時，表示需要做YCbCr轉RGB的資料量越大，相對效能提升的效果也越好。

In this paper, we propose hardware design for JPEG image decoder. The design process of decoder is handled JPEG header file. Then it handled variable length decoding and inverse quantization. We use inverse zig-zag scan rearrange the data, and use the two-dimensional inverse discrete cosine transform the data from the frequency domain values back to the apace domain values. Finally, it transformed YCbCr coordinate values into RGB coordinate values. The foregoing are completed by hardware description languages (Verilog). JPEG read file, and the data of RGB after decoding is added BMP header file let it became BMP image. The two parts are completed by high level language(C language).
According to the experimental data, the average speedup of the hardware JPEG decoder is about 2.9 to 4.3 times more than that of the software JPEG decoder. The larger the pixels of the input JPEG image are, the better the performance will be, particularly during the transformation from YcbCr to RGB.

摘要 III
Abstract IV
誌謝 V
目錄 VI
表目錄 IX
圖目錄 X
第一章緒論 1
1.1 研究動機 1
1.2 研究方法 1
1.3 論文架構 1
第二章背景與相關研究 2
2.1 JPEG介紹 2
2.1.1 JPEG編碼流程 3
2.1.1.1 二維離散餘弦轉換 4
2.1.1.2 量化 5
2.1.1.3 Zig-Zag掃描 6
2.1.1.4 差分編碼 7
2.1.1.5 變動長度編碼 7
2.1.1.6 霍夫曼編碼與霍夫曼表 8
2.2 JPEG解碼介紹 11
第三章 JPEG解碼架構分析 13
3.1 檔案格式介紹 13
3.1.1 JPEG檔案格式介紹 13
3.1.2 BMP檔案格式介紹 14
3.2 軟體程式設計 16
3.2.1 JPEG讀檔掃描 16
3.2.2 BMP影像建立 16
3.3 霍夫曼解碼說明 17
3.3.1 JPEG影像資料流結構 17
3.3.2 差分編碼 18
3.3.3 反量化 19
3.3.4 反Zig-Zag掃描 19
3.4 二維離散餘弦逆轉換(IDCT) 19
3.5 YCbCr轉RGB 21
第四章 JPEG解碼硬體設計與實現 23
4.1 JPEG解碼設計 23
4.1.1 JPEG標頭處理 24
4.1.2 霍夫曼解碼 27
4.1.3 二維離散餘弦逆轉換(IDCT) 34
4.1.4 YCbCr轉RGB 37
第五章 實驗結果與效能分析 40
5.1 開發工具 40
5.1.1 ModelSim模擬器 40
5.1.2 Quartus II 41
5.2 驗證與量測 42
5.3 實驗結果分析 43
第六章 結論與未來展望 48
6.1 結論 48
6.2 未來研究 49
參考文獻 50

[1] 李明峻(2011)，「嵌入式JPEG影像壓縮之軟硬體協同傳輸最佳化設計」， 國立成功大學電機工程學系碩士論文。
[2] CCITT (1992), “Information technology - Digital compression and coding of continuous-tone still images – Requirements and guidelines, ITU-T Recommendation T.81, ISO/IEC 10918-1.
[3] Chang, Y.W., Truong, T.K., Chang, Y. (2006), “Direct mapping architecture for jpeg huffman decoder, IEE Proceedings Communications 153, pp.333~340.
[4] En-hui Yang and Longji Wang (2009), “Joint Optimization of Run-Length Coding, Huffman Coding, and Quantization Table With Complete Baseline JPEG Decoder Compatibility, IEEE Transaction on Image Processing, VOL.18, NO.1, pp.63~74.
[5] Gregory K. Wallace (1991), “The JPEG Still Picture Compression Standard, Submitted in December for publication in IEEE Transactions on Consumer Electronics.
[6] Hanli Wang and Sam Kwong (2005), “Novel Variance Based Approach to Improving JPEG Decoding, IEEE International Conference on Industrial Technology, pp475~479.
[7] Jahanzeb Ahmad and Mansoor Ebrahim (2009), “FPGA based implementation of Baseline JPEG decoder. International Journal of Electrical ＆ Computer Sciences IJECS Vol: 9 No: 9
[8] Shanq-Jang Ruan and Wei-Te Lin (2007), “Bipartition Architecture for Low Power JPEG Huffman Decoder, The 12th Asia-Pacific Computer SystemsArchitecture Conference, August, pp. 235~243.
[9] Sung-Hsien Sun and Shie-Jue Lee (2003), “A JPEG Chip for Image Compression and Decompression, Journal of VLSI Signal Processing 35, pp.43-60.
[10] T. M. Czyszczon, R. S. Czernikowski, M. Shaaban, K. W. Hsu (1998), “Real-Time Implementation of JPEG Encoder/Decoder, Part of the SPIE Conference on Input/Output and Imaging Technologies. Taipei, Taiwan, July ,pp281~292.
[11]William B. Pennebaker and Joan L. Mitchell (1993), “JPEG: Still Image Data Compression Standard, New York: Van Nostrand Reinhold.