Presentation is loading. Please wait.

Presentation is loading. Please wait.

CSCE 641 Computer Graphics: Image Mosaicing Jinxiang Chai.

Similar presentations


Presentation on theme: "CSCE 641 Computer Graphics: Image Mosaicing Jinxiang Chai."— Presentation transcript:

1 CSCE 641 Computer Graphics: Image Mosaicing Jinxiang Chai

2 Outline Image registration - How to break assumptions? 3D-2D registration Image mosaicing

3 Readings Section 9.1, 9.2.3, and 9.3 in Szeliski book (required readings) Szeliski and Shum's siggraph97 paper on image mosaicing (suggested readings) image mosaicing - recognizing panoramas [website] (suggested readings)recognizing panoramaswebsite

4 virtual wide-angle camera Mosaics: Stitching Image Together

5 Mosaic Procedure Basic Procedure

6 Mosaic Procedure Basic Procedure –Take a sequence of images from the same position Rotate the camera about its optical center

7 Mosaic Procedure Basic Procedure –Take a sequence of images from the same position Rotate the camera about its optical center –Compute transformation between second image and first

8 Mosaic Procedure Basic Procedure –Take a sequence of images from the same position Rotate the camera about its optical center –Compute transformation between second image and first –Transform the second image to overlap with the first

9 Mosaic Procedure Basic Procedure –Take a sequence of images from the same position Rotate the camera about its optical center –Compute transformation between second image and first –Transform the second image to overlap with the first –Blend the two together to create a mosaic

10 Mosaic Procedure Basic Procedure –Take a sequence of images from the same position Rotate the camera about its optical center –Compute transformation between second image and first –Transform the second image to overlap with the first –Blend the two together to create a mosaic –If there are more images, repeat

11 Image Mosaic Is a pencil of rays contains all views (i.e., a 360-degree image Wait! No companies make such a camera! - real cameras often have a small field of view. real camera Any ideas?

12 Image Mosaic Is a pencil of rays contains all views (i.e., a 360-degree image Wait! No companies make such a camera! - Yes, stitching all the images into a synthetic 360-degree image real camera Any ideas?

13 Image Mosaic Is a pencil of rays contains all views (i.e., a 360-degree image Wait! No companies make such a camera! - Yes, stitching all the images into a synthetic 360-degree image real camera Any ideas? Synthetic 360-degree camera

14 Image Mosaic Is a pencil of rays contains all views (i.e., a 360-degree image Wait! No companies make such a camera! - Yes, stitching all the images into a synthetic 360-degree image real camera Any ideas? Synthetic 180-degree camera

15 Image Mosaic real camera synthetic camera Can generate any synthetic camera view as long as it has the same center of projection! Is a pencil of rays contains all views

16 Image Re-projection mosaic PP The mosaic has a natural interpretation in 3D –The images are reprojected onto a common plane –The mosaic is formed on this plane –Mosaic is a synthetic wide-angle camera

17 Issues in Image Mosaic mosaic PP How to relate two images from the same camera center? - image registration How to re-project images to a common plane? - image warping ? ?

18 Image Mosaicing Geometric relationship between images

19 Image Mosaicing Geometric relationship between images –Use 8-parameter projective transformation matrix

20 Image Mosaicing Geometric relationship between images –Use 8-parameter projective transformation matrix –Use a 3D rotation model (one R per image) Derive it by yourself!

21 Image Mosaicing Geometric relationship between images –Use 8-parameter projective transformation matrix –Use a 3D rotation model (one R per image) Register all pairwise overlapping images –Feature-based registration –Pixel-based registration Chain together inter-frame rotations

22 Image Stitching How to stitch two images I 1 and I 2 ? I1I1 I2I2

23 Image Stitching How to stitch two images I 1 and I 2 ? I1I1 I2I2 - warp the I2 to I1 based on the estimated projective transformation - blend the overlapping regions

24 Image Stitching How to stitch multiple pictures? - Stitch pairs together, blend, then crop

25 Image Stitching A big image stitched from 5 small images

26 Panoramas What if you want a 360  field of view? mosaic Projection Cylinder

27 Cylindrical Panoramas Steps –Re-project each image onto a cylinder –Blend –Output the resulting mosaic

28 Cylindrical Projection –Map 3D point (X,Y,Z) onto cylinder X Y Z unit cylinder

29 Cylindrical Projection –Map 3D point (X,Y,Z) onto cylinder X Y Z unwrapped cylinder –Convert to cylindrical coordinates unit cylinder

30 Cylindrical Projection –Map 3D point (X,Y,Z) onto cylinder X Y Z unwrapped cylinder –Convert to cylindrical coordinates cylindrical image –Convert to cylindrical image coordinates s defines size of the final image unit cylinder

31 Cylindrical Panoramas mosaic Projection Cylinder A B Cannot map point A to Point B without knowing (X,Y,Z)

32 Cylindrical Panoramas mosaic Projection Cylinder B C A But we can map point C (images) to Point B. - the coordinates of C are [x,y,f] - for details, see required readings

33 Cylindrical Panoramas f = 180 (pixels) Map image to cylindrical or spherical coordinates –need known focal length Image 384x300f = 380f = 280

34 Cylindrical Panorama 3D rotation registration of four images taken with a handheld camera.

35 Cylindrical Panorama

36 Recognizing Panoramas A fully automatic 2D image stitcher system Input Images Output panorama #1

37 Recognizing Panoramas A fully automatic 2D image stitcher system Input Images Output panorama #2

38 Recognizing Panoramas A fully automatic 2D image stitcher system Input Images Output panorama #3

39 Recognizing Panoramas A fully automatic 2D image stitcher system Input Images - How to recognize which images can be used for panoramas? - How to stitch them automatically?

40 Recognizing Panoramas A fully automatic 2D image stitcher system

41 Recognizing Panoramas A fully automatic 2D image stitcher system - Image matching with SIFT featuresSIFT - For every image, find the M best images with RANSAC - Form a graph and find connected component in the graph - Stitching and blending.

42 Recognizing Panoramas

43

44

45

46

47

48 Summary: Recognizing Panoramas

49 Outline Image registration - How to break assumptions? (cont.) 3D-2D registration Image mosaicing


Download ppt "CSCE 641 Computer Graphics: Image Mosaicing Jinxiang Chai."

Similar presentations


Ads by Google