feature based image mosaicing satya prakash mallick

21
Feature Based Image Mosaicing Satya Prakash Mallick

Upload: mason-seamons

Post on 11-Dec-2015

219 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Feature Based Image Mosaicing Satya Prakash Mallick

Feature Based Image Mosaicing

Satya Prakash Mallick

Page 2: Feature Based Image Mosaicing Satya Prakash Mallick

IntroductionMosaicing methods can be classified

broadly into

1. Direct Method: Uses information from all pixels. Iteratively updates an estimate of homography so

that a particular cost function is minimized. Sometimes Phase-Correlation is used to estimate

the a few parameters of the homography.

2. Feature Based Method: A few corresponding points are selected on the two images and homography is estimated using these reliable points only.

Page 3: Feature Based Image Mosaicing Satya Prakash Mallick

My Choice?

Feature Based Method: Because1. They are in general more accurate. 2. Can Handle large disparities.3. Convergence: Direct methods, may not

converge to the optimal solution is the presence of local minima.

4. For reliable performance direct methods rely on feature based initialization.

Page 4: Feature Based Image Mosaicing Satya Prakash Mallick

Problems in Image Mosaicing

1. Global Alignment: essentially means recovery of underlying homography.

2. Local Alignment: Correct the local mismatches left after global alignment.

3. Image Blending: a decision has to be made as to what color the overlapping region should take.

4. Image Warping: After calculation of homography, a decision has to be made as to how to warp one image w.r.t the other.

Page 5: Feature Based Image Mosaicing Satya Prakash Mallick

Problems in Image Mosaicing

5. Automatic selection of images to blend.

6. Auto-exposure compensation7. Camera error compensation.

I am looking at the first four problems.

Page 6: Feature Based Image Mosaicing Satya Prakash Mallick

The Algorithm ( Overview )

1. Detect corners in both the images2. Solve for correspondence. 3. Use RANSAC to estimate homography.4. Refine the estimate of homography

using a non-linear method.5. Make a decision on how to blend the

images.6. Warp the one image with respect to the

other taking the blending decision into consideration

Page 7: Feature Based Image Mosaicing Satya Prakash Mallick

Solving for correspondences

Corners were detected in both the images using Harris corner detector. Correspondences are solved using a version of Zhang’s relaxation algorithm:

Matching Through Correlation: Disambiguating Matches Through Relaxation:

Page 8: Feature Based Image Mosaicing Satya Prakash Mallick

Solving for correspondences

Matching through correlation:

Page 9: Feature Based Image Mosaicing Satya Prakash Mallick

Solving for correspondences

Disambiguating matches through relaxation:

Show correspondence results

Page 10: Feature Based Image Mosaicing Satya Prakash Mallick

Estimation of homography

RANSAC was implemented for estimating the homography relating the two scenes. Results show a comparison between using RANSAC and Least squares. The error function being minimized is: 2'2'1 ),(),( iiii HxxdxHxd

Page 11: Feature Based Image Mosaicing Satya Prakash Mallick

Image Blending

Weighted Image Blending:

Page 12: Feature Based Image Mosaicing Satya Prakash Mallick

Image Blending

Lightest Path Cut: The idea is to take the difference of the images in the overlapping region and cut the region along that curve of minimum intensity. The implementation was done using dynamic programming.

Note: Optimal implementation of the algorithm is non-trivial and can be a good problem to look into.

),,min( 1,1,11,1 jijijiijij EEEeE

Page 13: Feature Based Image Mosaicing Satya Prakash Mallick

Results of Blending

Try to find the curve along which the image overlap was cut! ALL THE BEST!

Page 14: Feature Based Image Mosaicing Satya Prakash Mallick

Homography Refinement

In the Weighted image blending example we saw, there was some blurring at the edges of the overlap. The obvious question is:

Are we sure that the homography estimate is good enough or can it be further refined?

Page 15: Feature Based Image Mosaicing Satya Prakash Mallick

Homography Refinement

The Homography was refined using Newton’s non-linear optimization technique. It’s “safe” to use Newton’s method because we are already close to the solution.

The vector X shown is made of inliers got using RANSAC

x1]x2 1,[)(

]2,2....,21,21,1,1......,12,12,11,11[

)(

HHHf

NyNxyxNyNxyxyxX

HfXT

Page 16: Feature Based Image Mosaicing Satya Prakash Mallick

Homography Refinement

Results: The blurring near the edge of overlapping region is almost gone.

Page 17: Feature Based Image Mosaicing Satya Prakash Mallick

Limitations and Mistakes

My biggest mistake: I assumed that, to make a mosaic with many images, it is sufficient to keep on adding images to the main mosaic. The mosaic doesn’t do well, if we try to blend more than four images. So there should have been a global refinement on different homographies

Page 18: Feature Based Image Mosaicing Satya Prakash Mallick

Conclusions

Lessons to be learnt: Non-Linear optimization of homography is not

an overkill. After RANSAC, it is usually very fast ( Most of the time, my results converged after 2-3 iterations only)

It’s a good idea to blend images along the lightest curve along the difference image. However, the image quilting algorithm cannot be directly used in that.

To blend a large number of images, there is “something” more to be done than finding pair-wise homography of images.

Page 19: Feature Based Image Mosaicing Satya Prakash Mallick

Conclusion

Image warping should always be done backward with bilinear interpolation.

Page 20: Feature Based Image Mosaicing Satya Prakash Mallick

References

[Hartley] Hartley, R. & Zisserman, A. (2000) Multiple View Geometry{Cambridge University Press, UK.[Shum] Shum, H. & Szeliski, R. (1998) Construction and refinement of panoramic mosaics with global and local alignment. IEEE Int'l Conf. Computer Vision, pp. 953-958.[Faugeras] Zoghlami, I. & Faugeras,O. & Deriche,R. (1997) Using geometric corners to build a 2d mosaic from as set of images.Computer Vision and Pattern Recognition, pp 421-425.[Zhang] Zhang, Z.& Deriche, R. & Faugeras, O & Luong, Q. A robust technique for matching two uncalibrated images through the recovery of the unknown epipolar geometry (1995) Artificial Intelligence Journal, 78:87-119, October 1995[Harris] Harris, C. & Stephens, M. A combined corner and edge detector.(1998) Proc. of 4th Alvey Vision Conf.,147-151.[Szeliski] Szeliski, R. Image Mosaicing for Tele-Reality Applications.(1994). Digital Equipment Corporation, Cambridge, USA.[Davis] Davis, J. Mosaics of scenes with moving objects.(1998).Computer Vision and Pattern Recognition[Capel] Capel,D & Zisserman,A. Automated mosaicing with super-resolutionzoom.(1998).Computer Vision and Pattern Recognition

Page 21: Feature Based Image Mosaicing Satya Prakash Mallick

THANKS !