跳到主要內容

臺灣博碩士論文加值系統

(18.97.14.87) 您好!臺灣時間:2025/02/17 12:44
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

我願授權國圖
: 
twitterline
研究生:彭俊傑
研究生(外文):Jyun-Jie Peng
論文名稱:基於ORB特徵之全景影像穩定演算法研究
論文名稱(外文):Algorithm Design Based on ORB Feature for Panoramic Video Stabilization
指導教授:賴永康
口試委員:黃朝宗吳崇賓
口試日期:2017-07-20
學位類別:碩士
校院名稱:國立中興大學
系所名稱:電機工程學系所
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:2017
畢業學年度:105
語文別:中文
論文頁數:73
中文關鍵詞:環景影像數位影像穩定
外文關鍵詞:panoramic videodigital video stabilization
相關次數:
  • 被引用被引用:0
  • 點閱點閱:952
  • 評分評分:
  • 下載下載:0
  • 收藏至我的研究室書目清單書目收藏:0
360全景影像在近幾年相當流行,然而,因為透過手持式360全景攝影機拍攝的影像易受攝影機的轉、震動影響,以虛擬實境視角觀看時,易視覺上的晃動導致類似暈車的症狀,因此,消弭影像震動之數位影像穩定技術(Digital Image Stabilization)在近幾年應用非常的頻繁,本論文中提出如何在手持式攝影機時,因人為不可預期的抖動或風力、外力等的影響而造成影像的抖動,以數位影像處理的方式去除不必要的抖動而保留必要的移動。數位影像穩定技術的困難點主要有兩個:(1) 如何從影像序列中有效率的估算出準確、可靠的全域移動向量(Global Motion Vector);(2) 如何從所取得的移動向量在邊界的限制下補償出一平滑的影像運動軌跡,本論文將會解決這些問題。
本論文首先介紹我們提出一個全景影像切割子畫面幀間預測演算法,並套用在數位影像穩定上面,本論文在移動估測技術(motion estimation)上提出將等距柱狀投影(equirectangular map)之球型影像切割出四個較符合二維影像之子畫面,並以ORB (Oriented FAST and Rotated BRIEF)特徵提取子畫面中之特徵點進行匹配,並基於特徵點距離、及RANSAC方法濾除前景移動得到的全域移動向量,接著將二維移動向量轉換為三維旋轉角度,以三維旋轉角度做移動向量加總演算法得到移動補償向量。此演算法可適用在不同的狀況的影像序列如複雜背景移動、大區域低對比、影像前景移動快速及抖動幅度大的影像狀況,使用ITF(Inter-frame transformation fidelity) 客觀的去評估效果可驗證本論文所提出的方法有效的改善數位影像穩定的效能。
本論文亦對系統提出最佳化之方法,包含Feature Buffer、特徵點距離之前景移動濾除、去除高斯金字塔階層、變形區塊之使用,系統運算時間可由1351.511ms改善至124.29ms,改善約10.87倍,本論文提出演算法之處理速度與其他文獻相比亦得到不錯之改善。
360-degree panoramic images are popular in recent years. However, because the image captured by the hand-held 360 degree panoramic camera is susceptible to the camera's turn and vibration effects, the visual sloshing causes a similar motion sickness when viewed in a virtual reality view. Therefore, to eliminate the image of the digital image stabilization technology in recent years, the application is very frequent. This paper suggests how to shake the image due to the unpredictable jitter or wind force, external force, etc. in the hand-held camera, and remove the necessary movement in the way of digital image processing to remove unnecessary jitter. There are two difficulties in the digital image stabilization technique: (1) To estimate an efficient, precise, and reliable global motion vector (GMV). (2) To use the existing GMV to compensate for a smooth motion trajectory within the window shifting allowance boundary.

Based on our proposed algorithm, we present a panoramic image sub-images inter frame prediction algorithm and apply it to digital image stabilization. Our motion estimation technique is proposed to cut out four sub-images that fit two-dimensional images from the equirectangular map image. Matching the feature point which extracted by ORB (Oriented FAST and Rotated BRIEF) method. And the RANSAC method is used to filter the global motion vector obtained by the foreground movement, and then the two-dimensional motion vector is transformed into the three-dimensional rotation angle. This algorithm can be applied in different conditions of the image sequence such as complex background movement, large area low contrast, image foreground movement fast and jitter amplitude large image condition. Using ITF (Inter-frame transformation fidelity) objective evaluation effect can be verified by the method proposed in this paper to effectively improve the stability of digital image performance.

We also propose methods to optimize the system which contain the Feature Buffer, the distance of the foreground movement to filter out, remove the Gaussian pyramid layer and image warping by the warping blocks. System operation time can be improved from 1351.511ms to 124.29ms. Processing time improved by 10.87 times. In this paper, the processing speed of the algorithm is improved compared with other literatures.
第一章 引言1
一、影像穩定技術種類簡介2
二、影像穩定技術發展概述6
三、論文概述8
第二章 全景影像之影像穩定演算法介紹與探討9
一、全景影像之影像格式10
二、全景影像之影像穩定演算法12
第三章 全景影像切割子畫面幀間預測演算法17
一、切割子畫面18
二、基於特徵點追蹤之全域移動向量計算19
(一)、特徵點提取19
(二)、特徵點匹配24
(三)、基於特徵點距離之前景移動濾除25
(四)、基於RANSAC之離群向量濾除26
(五)、二維移動向量之旋轉角度轉換29
三、移動補償估測技術37
(一)、移動向量加總演算法37
四、影像變形還原44
(一)、反旋計算(derotation)44
(二)、變形區塊47
五、算法搭配與效能之比較50
第四章 全景影像穩定系統最佳化與結果討論57
一、系統效能分析58
二、系統最佳化59
(一)、Feature Buffer59
(二)、特徵點距離之前景移動濾除60
(三)、去除高斯金字塔階層61
(四)、變形區塊之使用63
三、系統最佳化結果65
第五章 結論與未來展望68
一、結論68
二、未來展望68
參考文獻69
[1]K.M. Stanney, R.S. Kennedy and J. M. Drexler, “Cybersickness is not simulator sickness,” in Proceedings of the Human Factors and Ergonomics Society annual meeting, vol. 41, no. 2, pp. 1138-1142, Sage CA: Los Angeles, CA: SAGE Publications, October 1997.
[2]N. A. Tsoligkas, S. Xalkiadis, D. Xu and I. French, "A guide to digital image stabilization procedure — An overview," 2011 18th International Conference on Systems, Signals and Image Processing, Sarajevo, 2011, pp. 1-4.

[3]P. Paquis, M. Lonjon and P. Grellier, “Craniovertebral stabilization using a new CCD technique (Sofamor-Danek),” Neurochirurgie, vol. 44, no. 2, pp. 101-104, 1998.
[4]W. H. Cho and K. S. Hong, "A fast CIS still image stabilization method without parallax and moving object problems," in IEEE Transactions on Consumer Electronics, vol. 54, no. 2, pp. 197-205, May 2008.
[5]T. Albrecht, T. Tan, G. A. W. West and T. Ly, "Omnidirectional video stabilisation on a virtual camera using sensor fusion," 2010 11th International Conference on Control Automation Robotics & Vision, Singapore, 2010, pp. 2067-2072.
[6]R. Gregor, M. Lutzeler, M. Pellkofer, K. H. Siedersberger and E. D. Dickmanns, "EMS-Vision: a perceptual system for autonomous vehicles," in IEEE Transactions on Intelligent Transportation Systems, vol. 3, no. 1, pp. 48-59, Mar 2002.
[7]Wikipedia, “Equirectangular projection,” https://en.wikipedia.org/wiki/Equirectangular_projection
[8]PanoTools.org Wiki, “Projections,” http://wiki.panotools.org/Projections
[9]M. Kamali, A. Banno, J. C. Bazin, I. S. Kweon and K. Ikeuchi, “Stabilizing omnidirectional videos using 3D structure and spherical image warping,” in Proceedings IAPR Conference on Machine Vision Applications, pp. 177-180, June 2011.
[10]S. Kasahara, S. Nagai and J. Rekimoto, “First person omnidirectional video: System design and implications for immersive experience,” in Proceedings of the ACM International Conference on Interactive Experiences for TV and Online Video, ACM, pp. 33-42, June 2015.
[11]S. Pathak, A. Moro, A. Yamashita and H. Asama, "A decoupled virtual camera using spherical optical flow," 2016 IEEE International Conference on Image Processing (ICIP), Phoenix, AZ, 2016, pp. 4488-4492.
[12]J. Kopf, “360 video stabilization,” ACM Transactions on Graphics (TOG), 35(6), 195, November 2016.
[13]S. Erturk, "Digital image stabilization with sub-image phase correlation based global motion estimation," in IEEE Transactions on Consumer Electronics, vol. 49, no. 4, pp. 1320-1325, Nov. 2003.
[14]廖云暄,”數位影像穩定之演算法研究及其電路架構設計與實現”,中興大學電機工程學系所學位論文,中華民國一百零一年七月。
[15]S. Battiato, G. Gallo, G. Puglisi and S. Scellato, "SIFT Features Tracking for Video Stabilization," 14th International Conference on Image Analysis and Processing (ICIAP 2007), Modena, 2007, pp. 825-830.
[16]M. Kamali, S. ONO and K. IKEUCHI, “An efficient method for detecting and stabilizing shaky parts of videos in vehicle-mounted cameras,” SEISAN KENKYU, vol. 66, no. 2, pp. 87-94, March 2014.
[17]J. Xu, H. w. Chang, S. Yang and M. Wang, "Fast feature-based video stabilization without accumulative global motion estimation," in IEEE Transactions on Consumer Electronics, vol. 58, no. 3, pp. 993-999, August 2012.
[18]K. Y. Huang, Y. M. Tsai, C. C. Tsai and L. G. Chen, "Video stabilization for vehicular applications using SURF-like descriptor and KD-tree," 2010 IEEE International Conference on Image Processing, Hong Kong, 2010, pp. 3517-3520.
[19]E. Rublee, V. Rabaud, K. Konolige and G. Bradski, "ORB: An efficient alternative to SIFT or SURF," 2011 International Conference on Computer Vision, Barcelona, 2011, pp. 2564-2571.
[20]E. Rosten, R. Porter and T. Drummond, "Faster and Better: A Machine Learning Approach to Corner Detection," in IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 32, no. 1, pp. 105-119, Jan. 2010.
[21]P. L. Rosin, “Measuring corner properties,” Computer Vision and Image Understanding, vol. 73, no. 2, pp. 291-307, February 1999.
[22]M. Calonder, V. Lepetit, C. Strecha and P. Fua, “Brief: Binary robust independent elementary features,” Computer Vision–ECCV 2010, pp 778-792, 2010.
[23]M. A. Fischler and R. C. Bolles, “Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography,” Communications of the ACM, vol. 24, no. 6, pp. 381-395, June 1981.
[24]S. Erturk, "Image sequence stabilisation: motion vector integration (MVI) versus frame position smoothing (FPS)," ISPA 2001. Proceedings of the 2nd International Symposium on Image and Signal Processing and Analysis. In conjunction with 23rd International Conference on Information Technology Interfaces (IEEE Cat., Pula, 2001, pp. 266-271.
[25]R. Anupriya, U. Gomathy, C. Karunya and Y. Anitha, “Video stabilization based on global motion estimation using background feature point classification,” [Online]. Available: http://www.ijrde.com/attachments/article/174/69.pdf.
[26]R. Miyauchi, N. Shiroma and F. Matsuno, "Development of omni-directional image stabilization system using camera posture information," 2007 IEEE International Conference on Robotics and Biomimetics (ROBIO), Sanya, 2007, pp. 920-925.
[27]Nikon, “Nikon _ Imaging Products _ KeyMission 360”
http://imaging.nikon.com/lineup/action/keymission_360/spec.htm
[28]C. Morimoto and R. Chellappa, "Evaluation of image stabilization algorithms," Acoustics, Speech and Signal Processing, 1998. Proceedings of the 1998 IEEE International Conference on, Seattle, WA, 1998, pp. 2789-2792 vol.5.
[29]Y. Wang, Z. Hou, K. Leman and R. Chang, “Real-Time Video Stabilization for Unmanned Aerial Vehicles,” in MVA, pp. 336-339, June 2011.
[30]P. Viswanath, P. Swami, K. Desappan, A. Jain and A. Pathayapurakkal, “Orb in 5 ms: An efficient simd friendly implementation,” in Asian Conference on Computer Vision, Springer, Cham, pp. 675-686, November 2014.
[31]Engineering Blog _ Facebook Code, “Next-generation video encoding techniques for 360 video and VR,”
https://code.facebook.com/posts/1126354007399553/next-generation-video-encoding-techniques-for-360-video-and-vr/
連結至畢業學校之論文網頁點我開啟連結
註: 此連結為研究生畢業學校所提供,不一定有電子全文可供下載,若連結有誤,請點選上方之〝勘誤回報〞功能,我們會盡快修正,謝謝!
QRCODE
 
 
 
 
 
                                                                                                                                                                                                                                                                                                                                                                                                               
第一頁 上一頁 下一頁 最後一頁 top