New strategies for generating panoramic images for imperfect image series 8

When the seam-editing mode is selected, the shown panorama in the display window is dimmed out and thefocusing overlapping region is highlighted.. An example of seam-editing mode isshown

Input images (#1)

Input images (#2)

Figure A.1: Two sets of input images used in Chapter4

CHAPTER A Intermediate Result of Seam Driven Stitching Approach

CHAPTER A Intermediate Result of Seam Driven Stitching Approach

Figure B.1: The control dialog for the interactive panorama correction tools.

CHAPTER B Interface of the Interactive Panorama Correction Tool

Figure B.2: An example of the display window in seam-editing mode (a) The posite image with seam (b) The content of the left side image of the overlappingregion (c)The content of the right side image of the overlapping region

com-up accompanied with the initial panorama in the result display window From thefigure we can see that there are three main modes for the post editing: observe,seam editing and warping The user can switch among these modes by simplyclicking the corresponding radio buttons

In observe mode, the user is enabled to drag and scale the whole panorama

to a desired position This helps user to find minor artifacts in the monitoringwindow The seam editing mode and the warping mode are corresponding tothe two introduced tools described in Chapter 5 When the seam-editing mode

is selected, the shown panorama in the display window is dimmed out and thefocusing overlapping region is highlighted An example of seam-editing mode isshown in Figure B.2, we can see that the cursor is changed into a brush and can

be used to mark the overlapping region by holding the left button of the mousedown After the mouse button is released, the drawn stroke is considered as theinput mark-up and the overlapping region is updated using the method described

(a) (b) Figure B.3: An example of the display window in warp mode (a) The panoramabefore warping There exists noticeable artifacts The greed arrow indicates themoving direction of the brush (b) The result after warping.

in Section5.2.2 At the same time, in seam-editing mode, when the user clicks theright button of the mouse, the overlapping region will be toggled among the leftinput image, right input image and the composite result as shown in B.2(b) and(c)

When the warp mode is selected, the focusing object is no more the overlappingregion but the input image Thus, the whole focusing image is highlighted with itsneighboring overlapping region FigureB.3shows an example of the warp mode.The focusing image can be changed by clicking the “prev” and “next” button onthe dialog Similar as the seam-editing tool, a circular brush is used to warp theimages The user first adjusts the brush into a desired size Then by holding theleft button of the mouse down, the user can push the region inside the brush to

a desired position The snapping described in Section5.2.3is triggered when thebrush has intersected with the cutting seam

CHAPTER B Interface of the Interactive Panorama Correction Tool

Appendix C

Input Images for All Examples used

in This Thesis

Figure C.1: Input image series for Figure3.9

Figure C.2: Input image series for Figure3.10

CHAPTER C Input Images for All Examples used in This Thesis

Figure C.3: Input image series for Figure3.11

Figure C.4: Input image series for Figure4.1

Figure C.5: Input image series for Figure5.6

Figure C.6: Input image series for Figure5.7

Figure C.7: Input image series for Figure5.8

CHAPTER C Input Images for All Examples used in This Thesis

soft.com/en-[3] A Agarwala, M Dontcheva, M Agrawala, S Drucker, A Colburn, B Curless,

D Salesin, , and M Cohen Interactive digital photo-montage ACM tions on Graphics (TOG), 23(3):294–302, 2004 1.2, 2.1, 2.4.1, 2.4.2, 2.4.2, 2.4.2,2.5,3.5,5.2.1

Transac-[4] P Anandan, P.J Burt, K Dana, M.W Hansen, and G.S van der Wal Real-timescene stabilization and mosaic construction 1994 2.1

[5] Rakesh Kumar Anandan, Rakesh Kumar, P An, Michal Irani, James Bergen,and Keith Hanna Representation of scenes from collections of images In InIEEE Workshop on Representations of Visual Scenes, pages 10–17, 1995 2.1

[6] Shai Avidan and Ariel Shamir Seam carving for content-aware image resizing.ACM Transactions on Graphics (SIGGRAPH), 26(3):10, 2007 2.5

[7] Fady Badra, Ala Qumsieh, and Gregory Dudek Rotation and zooming inimage mosaicing In IEEE Workshop on Applications of Computer Vision (WACV),pages 50–55, 1998 2.2

[8] Connelly Barnes, Eli Shechtman, Adam Finkelstein, and Dan B Goldman.PatchMatch: A randomized correspondence algorithm for structural imageediting ACM Transactions on Graphics (SIGGRAPH), 28(3) 2.5

[9] Adrien Bartoli, Mathieu Perriollat, and Sylvie Chambon Generalized plate spline warps Internal Journal of Computer Vision (IJCV), 88(1):85–110,

[11] Adam Baumberg Reliable feature matching across widely separated views

In IEEE International Conference on Computer Vistion and Pattern Recognision(CVPR), pages 774–781, 2000 2.2

[12] James R Bergen, P Anandan, Keith J Hanna, and Rajesh Hingorani archical model-based motion estimation In European Conference on ComputerVision (ECCV), pages 237–252, 1992 2.1

Hier-[13] Yuri Boykov, Olga Veksler, and Ramin Zabih Fast approximate energy imization via graph cuts IEEE Transactions on Pattern Analysis and MachineIntelligence (TPAMI), 23:1222–1239, 2001 2.4.2,2.5,4.2.2

[14] Lisa Gottesfeld Brown A survey of image registration techniques ACMComputing Surveys, 24:325–376, 1992 2.1

[15] M Brown and D Lowe Recognising panoramas In International Conference

on Computer Vision (ICCV), 2003 1.2,2.1,2.2,2.3.2,3.5,5.2.1

[16] Matthew Brown and David G Lowe Automatic panoramic image stitchingusing invariant features Internal Journal of Computer Vision (IJCV), 74:59–73,

[20] Shenchang Eric Chen Quicktime vr - an image-based approach to virtualenvironment navigation In ACM SIGGRAPH, pages 26–33, 1995 1.2,2.1,2.2,2.3.3

[21] Jason J Corso and Gregory D Hager Coherent regions for concise and stableimage description In IEEE International Conference on Computer Vistion andPattern Recognision (CVPR), pages 184–190, 2005 2.2

[24] Martin A Fischler and Robert C Bolles Random sample consensus: a

paradig-m for paradig-model fitting with applications to iparadig-mage analysis and autoparadig-mated tography Commun ACM, 24:381–395, June 1981 2.2.2,3.3.1,4.2,5.2.1

car-[25] William T Freeman and Edward H Adelson The design and use of steerablefilters IEEE Transaction on Pattern Analysis and Machine Intelligence, 13(9):891–

[32] Tat jen Cham and Roberto Cipolla A statistical framework for long-rangefeature matching in uncalibrated image mosaicing In IEEE International Con-ference on Computer Vistion and Pattern Recognision (CVPR), pages 442–447 IEEEComputer Society, 1998 2.2

[33] J Jia and C.-K Tang Image stitching using structure deformation IEEETransactions on Pattern Analysis and Machine Intelligence (TPAMI), 30(4):617–

[36] Anat Levin, Assaf Zomet, Shmuel Peleg, and Yair Weiss Seamless imagestitching in the gradient domain In European Conference on Computer Vision(ECCV), volume 3024 of Lecture Notes in Computer Science, pages 377–389.Springer, 2004 1.2,2.1,2.4.1,6.2

stabi-[42] David G Lowe Distinctive image features from scale-invariant keypoints.Internal Journal of Computer Vision (IJCV), 60:91–110, November 2004.2.2,2.2.1,3.3.1,4.2,5.2.1

[43] Bruce D Lucas and Takeo Kanade An iterative image registration techniquewith an application to stereo vision pages 674–679, 1981 2.1

[44] Steve Mann and Rosalind W Picard Virtual bellows: Constructing highquality stills from video In In IEEE International Conference on Image Processing,pages 363–367, 1994 1.2,2.1,2.2

[45] Philip F McLauchlan and Allan Jaenicke Image mosaicing using sequentialbundle adjustment Image and Vision Computing, pages 751–759, 1999 2.1,2.2

Inter-[53] Michael Rubinstein, Ariel Shamir, and Shai Avidan Improved seam carvingfor video retargeting ACM Transactions on Graphics SIGGRAPH), 27(3):1–9,

2008 2.5

[54] Michael Rubinstein, Ariel Shamir, and Shai Avidan Multi-operator mediaretargeting ACM Transactions on Graphics (SIGGRAPH), 28(3):1–11, 2009 2.5

[55] Harpreet Sawhney and Rakesh Kumar True multi-image alignment and itsapplication to mosaicing and lens distortion correction IEEE Transactions onPattern Analysis and Machine Intelligence (TPAMI), 1999 2.1

[56] Scott Schaefer, Travis McPhail, and Joe D Warren Image deformation usingmoving least squares ACM Transactions on Graphics (TOG), 25(3):533–540,

2006 2.5

[57] Frederik Schaffalitzky and Andrew Zisserman Multi-view matching for ordered image sets, or “how do i organize my holiday snaps?” In EuropeanConference on Computer Vision (ECCV), pages 414–431, 2002 2.2

un-[58] H.-Y Shum and R Szeliski Construction of panoramic image mosaics withglobal and local alignment Internal Journal of Computer Vision (IJCV), 36(2):101–

[61] R Szeliski Image alignment and stitching: a tutorial Foundations and Trends

in Computer Graphics and Vision, 2:1–104, 2006 1.2,2.1,2.1,5.2.1

[62] Richard Szeliski Image mosaicing for tele-reality applications In IEEE

[67] Laura Teodosio and Walter Bender Salient video stills: Content and contextpreserved In ACM Multimedia, pages 39–46, 1993 2.1

[68] Matthew Uyttendaele, Ashley Eden, and Richard Szeliski Eliminating ing and exposure artifacts in image mosaics In IEEE International Conference

ghost-on Computer Vistighost-on and Pattern Recognisighost-on (CVPR), pages 509–516, 2001 2.1[69] Yu-Shuen Wang, Chiew-Lan Tai, Olga Sorkine, and Tong-Yee Lee Optimizedscale-and-stretch for image resizing ACM Transactions on Graphics (Proceedings

of ACM SIGGRAPH ASIA), 27(5), 2008 3.4.1,3.4.1

[70] Heung yeung Shum and Richard Szeliski Construction of panoramic image

[72] Barbara Zitov´a and Jan Flusser Image registration methods: a survey Imageand Vision Computing, 21:977–1000, 2003 2.1

[73] I Zoghlami, O Faugeras, and R Deriche Using geometric corners to build a2d mosaic from a set of images In IEEE International Conference on ComputerVistion and Pattern Recognision (CVPR), pages 420–425, 1997 2.2