Download presentation
Presentation is loading. Please wait.
1
Jianke Zhu From Haibin Ling’s ICCV talk Fast Marching Method and Deformation Invariant Features
2
Outline Introduction Fast Marching Method Deformation Invariant Framework Experiments Conclusion and Future Work
3
General Deformation One-to-one, continuous mapping. Intensity values are deformation invariant. (their positions may change)
4
Our Solution A deformation invariant framework Embed images as surfaces in 3D Geodesic distance is made deformation invariant by adjusting an embedding parameter Build deformation invariant descriptors using geodesic distances
5
Related Work Embedding and geodesics Beltrami framework [Sochen&etal98] Bending invariant [Elad&Kimmel03] Articulation invariant [Ling&Jacobs05] Histogram-based descriptors Shape context [Belongie&etal02] SIFT [Lowe04] Spin Image [Lazebnik&etal05, Johnson&Hebert99] Invariant descriptors Scale invariant descriptors [Lindeberg98, Lowe04] Affine invariant [Mikolajczyk&Schmid04, Kadir04, Petrou&Kadyrov04] MSER [Matas&etal02]
6
Outline Introduction Deformation Invariant Framework Intuition through 1D images 2D images Experiments Conclusion and Future Work
7
1D Image Embedding 1D Image I(x) EMBEDDING I(x) ( (1-α)x, αI ) αIαI (1-α)x Aspect weight α : measures the importance of the intensity
8
Geodesic Distance αIαI (1-α)x p q g(p,q) Length of the shortest path along surface
9
Geodesic Distance and α I1I1 I2I2 Geodesic distance becomes deformation invariant for α close to 1 embed
10
Image Embedding & Curve Lengths Depends only on intensity I Deformation Invariant Image I Embedded Surface Curve on Length of Take limit
11
Computing Geodesic Distances Fast Marching [Sethian96] Geodesic level curves Moving front Varying speed p
12
Deformation Invariant Sampling Geodesic Sampling 1. Fast marching: get geodesic level curves with sampling interval Δ 2. Sampling along level curves with Δ p sparse dense Δ Δ Δ Δ Δ
13
Deformation Invariant Sampling Geodesic Level Curves Geodesic Sampling 1. Fast marching: get geodesic level curves with sampling gap Δ 2. Sampling along level curves with Δ p
14
Geodesic Distance & Fast Marching
15
Deformation Invariant Descriptor p q p q Geodesic-Intensity Histogram (GIH) geodesic distance intensity geodesic distance intensity
16
Real Example p q
17
Deformation Invariant Framework Image Embedding ( close to 1) Deformation Invariant Sampling Geodesic Sampling Build Deformation Invariant Descriptors (GIH)
18
Practical Issues Lighting change Affine lighting model Normalize the intensity Interest-Point No special interest-point is required Extreme point (LoG, MSER etc.) is more reliable and effective
19
Invariant vs. Descriminative
20
Outline Introduction Deformation Invariance for Images Experiments Interest-point matching Conclusion and Future Work
21
Data Sets Synthetic Deformation & Lighting Change (8 pairs) Real Deformation (3 pairs)
22
Interest-Points * Courtesy of Mikolajczyk, http://www.robots.ox.ac.uk/~vgg/research/affine/ Interest-point Matching Harris-affine points [Mikolajczyk&Schmid04] * Affine invariant support regions Not required by GIH 200 points per image Ground-truth labeling Automatically for synthetic image pairs Manually for real image pairs
23
Descriptors and Performance Evaluation Descriptors We compared GIH with following descriptors: Steerable filter [Freeman&Adelson91], SIFT [Lowe04], moments [VanGool&etal96], complex filter [Schaffalitzky&Zisserman02], spin image [Lazebnik&etal05] * Performance Evaluation ROC curve: detection rate among top N matches. Detection rate * Courtesy of Mikolajczyk, http://www.robots.ox.ac.uk/~vgg/research/affine/
24
Synthetic Image Pairs
25
Real Image Pairs
26
Study of Interest-Points
27
Outline Introduction Deformation Invariance for Images Experiments Conclusion and Future Work
28
Thank You!
Similar presentations
© 2025 SlidePlayer.com Inc.
All rights reserved.