跳到主要內容

臺灣博碩士論文加值系統

(100.28.2.72) 您好!臺灣時間:2024/06/16 09:04
字體大小: 字級放大   字級縮小   預設字形  
回查詢結果 :::

詳目顯示

: 
twitterline
研究生:溫家瑜
研究生(外文):WEN,CHIA-YU
論文名稱:使用運動估計之立體視訊生成方法
論文名稱(外文):Stereoscopic Video Generation Using Motion Estimation
指導教授:杜維昌杜維昌引用關係
指導教授(外文):Wei-Chang Du
口試委員:杜維昌周遵儒黃健興
口試委員(外文):Wei-Chang DuChou, Tzren-ruChien-Hsiang Huang
口試日期:2013-07-26
學位類別:碩士
校院名稱:義守大學
系所名稱:資訊工程學系
學門:工程學門
學類:電資工程學類
論文種類:學術論文
論文出版年:2014
畢業學年度:101
語文別:中文
論文頁數:60
中文關鍵詞:立體影像運動估計快速搜尋法
外文關鍵詞:Motion EstimationStereoscopic VideoFast Search
相關次數:
  • 被引用被引用:0
  • 點閱點閱:1355
  • 評分評分:
  • 下載下載:1
  • 收藏至我的研究室書目清單書目收藏:0
自從2009年電影阿凡達成功開啟立體視覺新紀元,近年來有關立體的視覺數位媒體有越來越風行的趨勢。雖然市面上已有相當多廉價的立體顯示器,但其影像來源主要是利用立體攝影機進行拍攝,對於過去既有的視訊內容將無法呈現立體性視覺感受。本研究主要目的是將傳統視訊自動轉換成立體視訊資料,其中如何正確的分割出前景物件為主要議題之一。過去雖有「背景差分」與「畫面差分」等方法,但是都有其使用上的限制,特別是無法有效解決為動態背景的情況。本文採用運動估計的策略計算出運動向量,並以向量的大小、方向及分布作為判斷的依據,以期可解決動態背景的問題,並在靜態背景也可以達到不錯的物件分割效果。最後,透過視差處理、遮罩濾波、空洞填補等步驟產生出立體之影像。
Since 2009, the stereoscopic movie “Avatar” opened a new era successfully. In recent few years, stereoscopic digital media has become more and more popular. Although there are already a lot of inexpensive 3D display devices, the main source of stereo media is to use 3D video camera to produce content and, therefore, the past 2D video content cannot be rendered for providing stereoscopic visual experience. In this thesis, the objective is to convert traditional images to stereoscopic video data automatically. During the conversion process, it is one of critical issues to properly separate foreground objects from an image. In the literature, there are two traditional ways to segment moving objects, including “background subtraction” and “frame difference”, but both ways have the limitation of fixed background and are not effective solution for dynamic background situations. This study uses the motion estimation to compute motion vectors in an image, and then solves the foreground segmentation problem of dynamic background according to the vector size, orientation and distribution of motion vectors. Finally, perform parallax processing, mask filtering, and gap filling to generate stereoscopic images.
目 錄
摘 要.................................I
Abstract.............................II
目 錄.................................III
圖目錄................................V
Chapter 1............................1
1.1立體感受與影像獲取....................2
1.2立體顯示技術.........................3
1.3深度估測線索.........................6
1.4深度估測技術.........................6
1.5問題定義............................9
Chapter 2............................10
2.1影像取樣............................10
2.2物件分割............................11
2.2.1背景差分法........................11
2.2.2畫面差分法........................12
2.2.3運動估計法........................12
2.3遮罩濾波............................13
2.4視差處理............................14
2.5空洞填補............................16
Chapter 3.............................18
3.1 三步搜尋法(Three Step Search).......19
3.2新三步搜尋法(New Three Step Search)...20
3.3四步搜尋法(Four Step Search)..........21
3.4 鑽石型搜尋法(Diamond Search).........22
3.5 六角形搜尋法(Hexagon-based Search)...24
Chapter 4..............................25
4.1 非自然影像...........................25
4.2 自然影像.............................40
Chapter 5..............................49
參考文獻.................................50
圖目錄
圖1、人眼視覺系統..................2
圖2、影像格式..................3
圖3、分時式顯示原理..................4
圖4、偏光式顯示原理..................5
圖5、裸眼式顯示原理..................5
圖6、機器訓練演算法..................7
圖7、搭配消失點之梯度圖..................8
圖8、運動向量圖..................8
圖9、立體視訊生成方塊圖..................10
圖10、影像取樣示意圖..................11
圖11、運動估測二值圖..................13
圖12、開運算後結果..................14
圖13、閉運算後結果..................14
圖14、人眼與角度關係圖..................15
圖15、左右眼視圖產生示意圖..................15
圖16、空洞修補..................16
圖17、邊界空洞填補..................17
圖18、運動向量圖..................18
圖19、三步搜尋法..................20
圖20、新三步搜尋法..................21
圖21、四步搜尋法..................22
圖22、鑽石型搜尋法..................23
圖23、周圍四點搜尋..................23
圖24、六角形搜尋法..................24
圖25、非自然測試影像..................26
圖26、非自然影像向量圖..................26
圖27、非自然影像二值圖..................27
圖28、非自然影像左圖..................28
圖29、非自然影像右圖..................29
圖30、非自然影像全域立體圖..................29
圖31、非自然影像三步立體圖..................30
圖32、非自然測試影像..................30
圖33、非自然影像向量圖..................31
圖34、非自然影像二值圖..................31
圖35、非自然影像左圖..................32
圖36、非自然影像右圖..................33
圖37、非自然影像全域立體圖..................33
圖38、非自然影像三步立體圖..................34
圖39、非自然測試影像..................35
圖40、非自然影像向量圖..................35
圖41、非自然影像二值圖..................36
圖42、非自然影像左圖..................37
圖43、非自然影像右圖..................38
圖44、非自然影像全域立體圖..................38
圖45、非自然影像三步立體圖..................39
圖46、非自然測試影像效能比較..................39
圖47、自然測試影像..................40
圖48、自然影像向量圖..................40
圖49、自然影像二值圖..................41
圖50、自然影像左圖..................42
圖51、自然影像右圖..................43
圖52、自然影像全域立體圖..................43
圖53、自然影像三步立體圖..................44
圖54、自然影像測試圖..................44
圖55、自然影像向量圖..................45
圖56、自然影像二值圖..................45
圖57、自然影像左圖..................46
圖58、自然影像右圖..................47
圖59、自然影像全域立體圖..................47
圖60、自然影像三步立體圖..................48
圖61、自然測試影像效能比較..................48

[1]Haifeng Lu, Tianxu Zhang and Cong Yao,“A Novel Automatic Motion Segmentation Method Based on Optical Flow,”International Conference on Multimedia Technology, pp. 1-5, 2010.
[2]Fung Hsuan Cheng and Chien-Chih Chen“3D Stereo Imaging Technology From 2D Image Sequence,”3rd International Conference on Innovative Computing Information and Control, pp. 436-439, June 2008.
[3]Arnell. F and Petersson. L,“Fast Object Segmentation from a Moving Camera,”IEEE Proceedings Intelligent Vehicles Symposium, pp. 136-141, 2005.
[4]Chul-Ho Choi, Byong-Heon Kwon and Myung-Ryul Choi, “A Real-Time Field-Sequential Stereoscopic Image Converter,” IEEE Trans. Consumer Electronics, pp. 903-910, Aug. 2004.
[5]Liang Zhang and Wa James Tam,“Stereoscopic Image Generation Based on Depth Images for 3D TV,”IEEE Transactions on Broadcasting, vol. 51, no.2, pp.191-199, June 2005.
[6]Donghyun Kim, Dongbo Min and Kwanghoon Sohn, “Stereoscopic Video Generation Method Using Motion Analysis,” IEEE 3DTV Conference, pp.1-4, May 2007.
[7]P. Harman, J. Flack, S. Fox and M. Dowley, “Rapid 2D to 3D Conversion,” Proc. SPIE, Vol. 4660, pp. 78-86, 2002.
[8]S. Battiato, S. Curti, M. La Cascia, E. Scordato and M.Tortora,“Depth Map Generation By Image Classification,”Proc. of SPIE IS&T/SPIE’s 16th Annual Symposium on Electronic Imaging, pp. 95-104, 2004.
[9]Chao-Chung Cheng, Chung-Te Li and Liang-Gee Chen,“A 2D-to-3D Conversion System Using Edge Information,”Digest of Technical Papers International Conference on Consumer Electronics (ICCE), pp. 377-378, 2010.
[10]Chung-Te Li, Yen-Chieh Lai and Chien Wu,“Perceptual multi-cues 2D-to-3D Conversion System,” IEEE Visual Communication and Image Processing (VCIP), 2011.
[11]賴文能、錢韋志,淺談2D至3D轉換技術,國立中正大學電機工程研究所,影像與識別,vol. 16, no. 2, pp. 61-75, 2010.
[12]劉楷哲、吳其霖、黃偉豪、陳信榮、李錕、羅豐祥,基於3D顯示器之即時3D內容合成技術,影像與識別,vol. 16, no. 2, pp. 76-95, 2010.
[13]R. Li, B. Zeng and M. L. Liou, “A New Three-Step Search Algorithm for Block Motion Estimation,” IEEE Trans. Circuits System Video Technology, vol. 4, no. 4, pp. 438-442, Aug. 1994.
[14]L. M. Po and W. C. Ma, “A Novel Four-Step Search Algorithm for Fast Block Motion Estimation,” IEEE Trans. Circuits System Video Technology, vol. 6, no. 3, pp. 313-317, June 1996.
[15]Shan Zhu and Kai-Kuang Ma, “A New Diamond Search Algorithm for Fast Block-matching Motion Estimation,”IEEE Transactions on Image Processing, vol. 9, no. 2, pp. 287-290, Feb. 2000.
[16]Ce Zhu, Xiao Lin and Lap-Pui Chau,“Hexagon-based Search Pattern for Fast Block Motion Estimation,” IEEE Trans. Circuits System Video Technology, vol. 12, no. 5, pp. 349-355, May 2002.
[17]Lipton, et al., Stereoscopic Television System with Field Storage for Sequential Display of Right and Left Images, U.S. Patent, no. 4562463, 1985.

QRCODE
 
 
 
 
 
                                                                                                                                                                                                                                                                                                                                                                                                               
第一頁 上一頁 下一頁 最後一頁 top