簡易檢索 / 詳目顯示
| 研究生： |
Christin Erniati Panjaitan Christin - Erniati Panjaitan |
|---|---|
| 論文名稱： |
一個適用於Kinect深度影片壓縮之低複雜度前處理 A Low Complexity Pre-Processing for Kinect Depth Video Compression |
| 指導教授： |
阮聖彰
Shanq-Jang Ruan 沈中安 Chung-An Shen |
| 口試委員: |
林昌鴻
Chang-Hong Lin 陳筱青 Hsiao-Chin Chen |
| 學位類別： |
碩士 Master |
| 系所名稱： |
電資學院 - 電子工程系 Department of Electronic and Computer Engineering |
| 論文出版年： | 2016 |
| 畢業學年度： | 104 |
| 語文別： | 英文 |
| 論文頁數： | 41 |
| 中文關鍵詞： | 深入的視頻 、壓縮 、3D 、Kinect 、改革 、向上/向下樣本 |
| 外文關鍵詞： | Depth Video, Compression, 3D, Kinect, Reformation, Up/Donwsample |
| 相關次數： | 點閱：295 下載：11 |
| 分享至: |
| 查詢本校圖書館目錄 查詢臺灣博碩士論文知識加值系統 勘誤回報 |
目前可以撥放立體影片(3DV)的3DTV共有被動立體及主動立體兩種方式使用被動立體式的3DTV，使用者需要穿戴特別的眼鏡，而主動立體式則不需配戴額外的配件。3DV以紋理及深度影像組合而成。一般來說，深度影像藉由具有深度感測器的相機產生。為了啟用3DV，一個好的合成及壓縮是必須的。然而，深度資訊遭受到不可忽略的量測時的雜訊，這將會在重建3D資料時，破壞壓縮率及影響合成品質。本論文專注於前處理技術得到高效率及減少運算複雜度。將一個Kinect所拍攝的影像以每個不重複的區塊的中間值縮小。一個縮小後的深度影片以DNB改良及填充。在濾波的過程中，每個非重疊的區塊使用一個不同大小的偵測邊緣資訊的子視窗。在導入H.264之前，利用被放置在前處理末端的向上取樣用於恢復被縮小的深度影片的大小到原始大小。實驗結果顯示本論文提出的方法增加效能而所需的運算時間比傳統的H.264還低
There is two kinds of three dimensional video television (3DTV) which are stereoscopic and autostreoscopic display that can deliver three dimensional video (3DV). In stereoscopic display, users need to wear a special glasses and autostreoscopic display does not require any special accessories. 3DV consist of texture and depth video. Generally, depth video can be generated from some cameras which equipped with depth sensors. In order to enable 3DV, a good synthesis and a good compression are required. However, depth data is suffering from non-negligible estimation noise where it destroys compression ratio and affects synthesis quality to reconstruct 3D data. This thesis focuses on pre-processing technique to obtain high performance and reduce the computational complexity. A raw Kinect depth video is downsampled with selecting median value for each non-overlapped block. A downsampled depth video is reformed with divisive normalized bilateral filter (DNBL) and padding. In the filtering process, each non-overlapped block uses a different window size based on detected edge information. Upsample is placed at the end of pre-processing part to return the size of a downsampled depth video into the original size before feed into H.264/AVC. The experimental results demonstrate that the proposed method increases the performance and reduces the computational time over classical H.264/AVC.
[1]D. Minoli, 3DTV Content Capture, Encoding and Transmission : Building the Transport Infrastructure for Commercial Services. New Jersey : John Willey & Sons Inc. 2010.
[2]A. Smolic, K. Mueller, N. Stefanoski, J. Ostermann, A. Gotchev, G. B. Akar, G. Triantafyllidis, and A. Koz, “Coding algorithms for 3dtv – a survey,” IEEE Trans. Circuits and Systems for Video Technology., vol. 17, pp. 1606 – 1621, 2007.
[3]Microsoft Kinect. [Online]. Available : http://www.xbox.com/kinect.
[4]S. B. Gokturk, H. Yalcin, and C. Bamji, “A time-of-flight depth sensor – system description, issues and solutions,” in Proc. CVPRW, 2004, pp. 35.
[5]Z. Jing, W. L. – Hao, L. D. – Xiao, and Z. Ming, “High quality depth maps from stereo matching and tof camera,” in Proc. SoCPaR, 2011, pp. 68 – 72.
[6]Z. Yang, C. Meng, and Y. Lei, “Application of 3d laser scanner in topographic change monitor and analysis,” in Proc. ICEMI, 2009, pp. 4-382 – 4-385.
[7]Depth Estimation Reference Software (DERS). [Online]. Avalaible : http://wg11.sc29.org/svn/repos/MPEG-4/test/tags/3D/.
[8]M. Tanimoto, T. Fujii, and K. Suzuki, “Video depth estimation reference software (DERS) with image segmentation and block matching” ISO/IEC JTC1/SC29/WG11 MPEG/M16092 ed. Laussane, Switzerland, 2009.
[9]O. Stankiewicz and K. Wegner, “Depth Map Estimation Software Version 3,” ISO/IEC JTC1/SC29/WG11 MPEG/M15540 ed. Hanover, Germany, 2008.
[10]Y. Q. Shi, and H. Sun, Image and Video Compression for Multimedia Engineering. USA: CRC Press. 2008.
[11]I. E. G. Richardson, H.264 and MPEG-4 Video Compression. UK : John Willey & Sons Ltd, 2003.
[12]I. E. G. Richardson, The H.264 Advanced Video Compression Standard. UK : John Willey & Sons Ltd, 2010.
[13]K. Müller, P. Merkle, and T. Wiegand, “3D video representation using depth maps,” Proceedings of IEEE, vol. 99, pp. 643 – 656, 2011.
[14]G. Cheung, A. Ortega, W. –S Kim, V. Velisavljevic and A. Kubota, “Depth map compression for depth-image-based-rendering,” New York : Springer, 2013.
[15]View Synthesis Reference Software (VSRS) Version 3.5. [Online]. Available : http://wg11.sc29.org/svn/repos/MPEG-4/test/tags/3D/view-synthesis/VSRS-3-5.
[16]T. Mallick, P. P. Das, and A. K. Majumdar,”Characterization of noise in kinect depth images : a review,” IEEE Sensors Journal, vol. 14, pp. 1731 - 1740, 2014.
[17]P. Zongju, J. Gangyi, Y. Mei, P. Shihua, and C. Fen, “Temporal pixel classification and smoothing for higher depth video compression performance,” IEEE Trans. Consumer Electronics, vol. 57, pp. 1815 – 1822, 2011.
[18]C. Panjaitan, C.-A Shen, and S.-J Ruan, “A low complexity depth map compression approach for Microsoft Kinect devices,” IEEE Global Conference Consumer Electronic, pp. 339 – 340, 2015.
[19]F. Jingjing, M. Dan, Y. Weiren, W. Shiqi, L. Yan, and L. Shipeng, “Kinect-like depth data compression,” IEEE Trans. Multimedia, vol. 15, pp. 1340 – 1352, 2013.
[20]R. C. Gonzalez and R. E. Woods, Digital Image Processing, 3rd ed., Taiwan : Pearson, 2008.
[21]K. Khoshelham, “Accuracy analysis of Kinect depth data,”in. Proc. ISPRS, vol. XXXVIII – 5/W12, 2011.
[22]O. Kwan-Jung, Y. Sehoon, A. Vetro, and H. Yo-Sung,”Depth reconstruction filter and down/up sampling for depth coding in 3-d video,” IEEE Signal Processing Letters, vol. 16, pp. 747 – 750, 2009.
[23]H. 264 software resipository (JMkta 14.2). [Online]. Available : http://iphome.hhi.de/suehring/tml/.
