Surface reconstruction of sea-ice through stereo - initial steps Rohith MV Gowri Somanath VIMS Lab.

Slides:



Advertisements
Similar presentations
Bayesian Belief Propagation
Advertisements

Active Shape Models Suppose we have a statistical shape model –Trained from sets of examples How do we use it to interpret new images? Use an “Active Shape.
The fundamental matrix F
CSE473/573 – Stereo and Multiple View Geometry
Efficient High-Resolution Stereo Matching using Local Plane Sweeps Sudipta N. Sinha, Daniel Scharstein, Richard CVPR 2014 Yongho Shin.
Stereo Vision Reading: Chapter 11
Recap from Previous Lecture Tone Mapping – Preserve local contrast or detail at the expense of large scale contrast. – Changing the brightness within.
Stereo.
December 5, 2013Computer Vision Lecture 20: Hidden Markov Models/Depth 1 Stereo Vision Due to the limited resolution of images, increasing the baseline.
Boundary matting for view synthesis Samuel W. Hasinoff Sing Bing Kang Richard Szeliski Computer Vision and Image Understanding 103 (2006) 22–32.
Last Time Pinhole camera model, projection
Stanford CS223B Computer Vision, Winter 2005 Lecture 6: Stereo 2 Sebastian Thrun, Stanford Rick Szeliski, Microsoft Hendrik Dahlkamp and Dan Morris, Stanford.
CS6670: Computer Vision Noah Snavely Lecture 17: Stereo
CSci 6971: Image Registration Lecture 4: First Examples January 23, 2004 Prof. Chuck Stewart, RPI Dr. Luis Ibanez, Kitware Prof. Chuck Stewart, RPI Dr.
Multiple View Geometry : Computational Photography Alexei Efros, CMU, Fall 2005 © Martin Quinn …with a lot of slides stolen from Steve Seitz and.
Direct Methods for Visual Scene Reconstruction Paper by Richard Szeliski & Sing Bing Kang Presented by Kristin Branson November 7, 2002.
The plan for today Camera matrix
Computer Vision Optical Flow Marc Pollefeys COMP 256 Some slides and illustrations from L. Van Gool, T. Darell, B. Horn, Y. Weiss, P. Anandan, M. Black,
CS 223b 1 More on stereo and correspondence. CS 223b 2 =?f g Mostpopular For each window, match to closest window on epipolar line in other image. (slides.
3D from multiple views : Rendering and Image Processing Alexei Efros …with a lot of slides stolen from Steve Seitz and Jianbo Shi.
CSCE 641 Computer Graphics: Image-based Modeling Jinxiang Chai.
Summer Project Presentation Presented by:Mehmet Eser Advisors : Dr. Bahram Parvin Associate Prof. George Bebis.
Aleixo Cambeiro Barreiro 광주과학기술원 컴퓨터 비전 연구실
Announcements PS3 Due Thursday PS4 Available today, due 4/17. Quiz 2 4/24.
Stereo Guest Lecture by Li Zhang
Project 1 artifact winners Project 2 questions Project 2 extra signup slots –Can take a second slot if you’d like Announcements.
Stereo Matching & Energy Minimization Vision for Graphics CSE 590SS, Winter 2001 Richard Szeliski.
Viewpoint Tracking for 3D Display Systems A look at the system proposed by Yusuf Bediz, Gözde Bozdağı Akar.
Multiple View Geometry : Computational Photography Alexei Efros, CMU, Fall 2006 © Martin Quinn …with a lot of slides stolen from Steve Seitz and.
Accurate, Dense and Robust Multi-View Stereopsis Yasutaka Furukawa and Jean Ponce Presented by Rahul Garg and Ryan Kaminsky.
Manhattan-world Stereo Y. Furukawa, B. Curless, S. M. Seitz, and R. Szeliski 2009 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp.
3-D Scene u u’u’ Study the mathematical relations between corresponding image points. “Corresponding” means originated from the same 3D point. Objective.
Computer Vision Spring ,-685 Instructor: S. Narasimhan WH 5409 T-R 10:30am – 11:50am Lecture #15.
Computer vision: models, learning and inference
Camera Calibration & Stereo Reconstruction Jinxiang Chai.
Introduction Belief propagation: known to produce accurate results for stereo processing/ motion estimation High storage requirements limit the use of.
Symmetric Architecture Modeling with a Single Image
Stereo Matching Information Permeability For Stereo Matching – Cevahir Cigla and A.Aydın Alatan – Signal Processing: Image Communication, 2013 Radiometric.
Gwangju Institute of Science and Technology Intelligent Design and Graphics Laboratory Multi-scale tensor voting for feature extraction from unstructured.
Dobrina Boltcheva, Mariette Yvinec, Jean-Daniel Boissonnat INRIA – Sophia Antipolis, France 1. Initialization Use the.
Mutual Information-based Stereo Matching Combined with SIFT Descriptor in Log-chromaticity Color Space Yong Seok Heo, Kyoung Mu Lee, and Sang Uk Lee.
1/20 Obtaining Shape from Scanning Electron Microscope Using Hopfield Neural Network Yuji Iwahori 1, Haruki Kawanaka 1, Shinji Fukui 2 and Kenji Funahashi.
Lecture 12 Stereo Reconstruction II Lecture 12 Stereo Reconstruction II Mata kuliah: T Computer Vision Tahun: 2010.
Perception Introduction Pattern Recognition Image Formation
Rohith MV, Gowri Somanath, Chandra Kambhamettu Video/Image Modeling and Synthesis(VIMS) Lab, Dept. of Computer and Information Sciences Cathleen Geiger.
Recap from Monday Image Warping – Coordinate transforms – Linear transforms expressed in matrix form – Inverse transforms useful when synthesizing images.
Object Stereo- Joint Stereo Matching and Object Segmentation Computer Vision and Pattern Recognition (CVPR), 2011 IEEE Conference on Michael Bleyer Vienna.
A Non-local Cost Aggregation Method for Stereo Matching
Stereo Vision Reading: Chapter 11 Stereo matching computes depth from two or more images Subproblems: –Calibrating camera positions. –Finding all corresponding.
Scientific Writing Abstract Writing. Why ? Most important part of the paper Number of Readers ! Make people read your work. Sell your work. Make your.
December 4, 2014Computer Vision Lecture 22: Depth 1 Stereo Vision Comparing the similar triangles PMC l and p l LC l, we get: Similarly, for PNC r and.
Stereo Readings Szeliski, Chapter 11 (through 11.5) Single image stereogram, by Niklas EenNiklas Een.
Feature-Based Stereo Matching Using Graph Cuts Gorkem Saygili, Laurens van der Maaten, Emile A. Hendriks ASCI Conference 2011.
Computer Vision, Robert Pless
A Region Based Stereo Matching Algorithm Using Cooperative Optimization Zeng-Fu Wang, Zhi-Gang Zheng University of Science and Technology of China Computer.
Lec 22: Stereo CS4670 / 5670: Computer Vision Kavita Bala.
Computer Vision Lecture #10 Hossam Abdelmunim 1 & Aly A. Farag 2 1 Computer & Systems Engineering Department, Ain Shams University, Cairo, Egypt 2 Electerical.
Validation of a high-resolution (400m) SAR motion tracking system near the APLIS’07 Ice Camp M. Thomas, C. A. Geiger and C. Kambhamettu.
Expectation-Maximization (EM) Case Studies
Bahadir K. Gunturk1 Phase Correlation Bahadir K. Gunturk2 Phase Correlation Take cross correlation Take inverse Fourier transform  Location of the impulse.
Lecture 16: Stereo CS4670 / 5670: Computer Vision Noah Snavely Single image stereogram, by Niklas EenNiklas Een.
Non-Ideal Iris Segmentation Using Graph Cuts
Tal Amir Advanced Topics in Computer Vision May 29 th, 2015 COUPLED MOTION- LIGHTING ANALYSIS.
Stereo CS4670 / 5670: Computer Vision Noah Snavely Single image stereogram, by Niklas EenNiklas Een.
A Plane-Based Approach to Mondrian Stereo Matching
Semi-Global Matching with self-adjusting penalties
CS4670 / 5670: Computer Vision Kavita Bala Lec 27: Stereo.
Announcements Midterms graded (handed back at end of lecture)
Finite Element Surface-Based Stereo 3D Reconstruction
Liyuan Li, Jerry Kah Eng Hoe, Xinguo Yu, Li Dong, and Xinqi Chu
Presentation transcript:

Surface reconstruction of sea-ice through stereo - initial steps Rohith MV Gowri Somanath VIMS Lab

Sea ice IntroductionStereo on Ice ImagesOur AlgorithmResultsConclusion

Overview Introduction Need for reconstruction Previous approaches Camera system and field trip Stereo on ice images Our algorithm Results Conclusion IntroductionStereo on Ice ImagesOur AlgorithmResultsConclusion

Need for reconstruction “The feasibility of using snow surface roughness to infer ice thickness and ice bottom roughness is promising….” “…the goal of a circumpolar high resolution data set of Antarctic sea ice and snow thickness distributions has not yet been achieved …” “…crucial for future validation of satellite observations, climate models, and for assimilation into forecast models…” Ref: Workshop on Antarctic Sea Ice Thickness, 2006; Annals of Glaciology IntroductionStereo on Ice ImagesOur AlgorithmResultsConclusion

Previous methods – LIDAR Echelmeyer, K.A., V.B. Valentine, and S.L. Zirnheld, (2002, updated 2004): Airborne surface profiling of Alaskan glaciers. Boulder, CO: National Snow and Ice Data Center. Digital media. Dalå, N. S., R. Forsberg, K. Keller, H. Skourup, L. Stenseng, S. M.Hvidegaard, (2004): Airborne LIDAR measurements of sea ice north of Greenland and Ellesmere Island 2004, GreenICe/SITHOS/CryoGreen/A 76 Projects, Final Report, pp 73. IntroductionStereo on Ice ImagesOur AlgorithmResultsConclusion

Camera system IntroductionStereo on Ice ImagesOur AlgorithmResultsConclusion

Field trip IntroductionStereo on Ice ImagesOur AlgorithmResultsConclusion

Samples IntroductionStereo on Ice ImagesOur AlgorithmResultsConclusion

IntroductionStereo on Ice ImagesOur AlgorithmResultsConclusion

Features in data Smoothly changing disparity No edge Low color variation IntroductionStereo on Ice ImagesOur AlgorithmResultsConclusion

Features in data Specular Highlights IntroductionStereo on Ice ImagesOur AlgorithmResultsConclusion

Stereo Disparity (d) Edge based matching(c) Non-Linear Diffusion(b) Membrane Diffusion IntroductionStereo on Ice ImagesOur AlgorithmResultsConclusion

Diffusion 1 IntroductionStereo on Ice ImagesOur AlgorithmResultsConclusion

Diffusion 10 IntroductionStereo on Ice ImagesOur AlgorithmResultsConclusion

Diffusion 20 IntroductionStereo on Ice ImagesOur AlgorithmResultsConclusion

Diffusion 50 IntroductionStereo on Ice ImagesOur AlgorithmResultsConclusion

Diffusion 80 IntroductionStereo on Ice ImagesOur AlgorithmResultsConclusion

Diffusion 120 IntroductionStereo on Ice ImagesOur AlgorithmResultsConclusion

Diffusion 150 IntroductionStereo on Ice ImagesOur AlgorithmResultsConclusion

Diffusion 200 IntroductionStereo on Ice ImagesOur AlgorithmResultsConclusion

Diffusion 250 IntroductionStereo on Ice ImagesOur AlgorithmResultsConclusion

Diffusion 300 IntroductionStereo on Ice ImagesOur AlgorithmResultsConclusion

Classification Unambiguous Low Variance Occluded IntroductionStereo on Ice ImagesOur AlgorithmResultsConclusion

Algorithm for Classification IntroductionStereo on Ice ImagesOur AlgorithmResultsConclusion

How to fill Low Variance areas? Don’t have any unambiguous information about the depth at those pixels Interpolate from Boundary True Map Surface IntroductionStereo on Ice ImagesOur AlgorithmResultsConclusion

Interpolation 63 Sampled VerticesTrue Map IntroductionStereo on Ice ImagesOur AlgorithmResultsConclusion

How to Interpolate? Given n points on the boundary Triangulate… Which Triangulation? Delaunay Triangulation True Map 61 faces IntroductionStereo on Ice ImagesOur AlgorithmResultsConclusion

Subdivide Loop Subdivision True Map 244 faces IntroductionStereo on Ice ImagesOur AlgorithmResultsConclusion

Subdivide True Map 3904 faces 976 faces IntroductionStereo on Ice ImagesOur AlgorithmResultsConclusion

What if…? True Map 104 faces 225 faces 425 faces 244 faces subdivision IntroductionStereo on Ice ImagesOur AlgorithmResultsConclusion

Towards Algorithm Don’t know vertices…Don’t know edges Given Vertices…What are the best edges? Delaunay Triangulation Outline Scatter Points Triangulate Move Points Repeat… IntroductionStereo on Ice ImagesOur AlgorithmResultsConclusion

Unstructured Triangulation Algorithm IntroductionStereo on Ice ImagesOur AlgorithmResultsConclusion

Advantages Very simple Quality of Triangles is high Errors in Interpolation are low Can handle concave shapes and regions with holes IntroductionStereo on Ice ImagesOur AlgorithmResultsConclusion

Negatives Uses Delaunay to triangulate every iteration May become unstable with wrong choice of parameters (very rare) May not converge IntroductionStereo on Ice ImagesOur AlgorithmResultsConclusion

Finite Element Method Courtesy : A Pragmatic Introduction to the Finite Element Method for Thermal and Stress Analysis, Petr Krysl IntroductionStereo on Ice ImagesOur AlgorithmResultsConclusion

Finite Element Method Courtesy : A Pragmatic Introduction to the Finite Element Method for Thermal and Stress Analysis, Petr Krysl IntroductionStereo on Ice ImagesOur AlgorithmResultsConclusion

Finite Element Method Courtesy : IntroductionStereo on Ice ImagesOur AlgorithmResultsConclusion

True surface True map 63 samples on boundary Interpolation with Delaunay Delaunay Triangulation (61 faces)Delaunay + Loop Subdivision (244 faces) Interpolation of Delaunay + Loop Subdivision Unstructured triangulation From [1] Interpolation with Unstructured triangulation IntroductionStereo on Ice ImagesOur AlgorithmResultsConclusion

Result Ambiguous Unambiguous disparity Triangulation Final disparity IntroductionStereo on Ice ImagesOur AlgorithmResultsConclusion

Comparison (c) Non-Linear Diffusion (b) Membrane Diffusion IntroductionStereo on Ice ImagesOur AlgorithmResultsConclusion (e) Ground Truth

More results IntroductionStereo on Ice ImagesOur AlgorithmResultsConclusion

More results IntroductionStereo on Ice ImagesOur AlgorithmResultsConclusion

Conclusions In areas containing very low color variation, interpolation gives better results than image matching Heuristic for classifying image regions Efficient methods for interpolation using triangulation and FEM IntroductionStereo on Ice ImagesOur AlgorithmResultsConclusion

Future Directions Include disparity variance in factors for classification Change the differential equation to model developable surfaces IntroductionStereo on Ice ImagesOur AlgorithmResultsConclusion

Publications Towards Estimation of Dense Disparities from Stereo Images Containing Large Textureless Regions. Rohith MV, Gowri Somanath, Chandra Kambhamettu, Cathleen Geiger. 19 th International Conference on Pattern Recognition. December Tampa, USA Reconstruction Of Snow And Ice Surfaces Using Multiple View Vision Techniques. Gowri Somanath, Rohith MV, Cathleen Geiger, Chandra Kambhamettu. 65 th Eastern Snow Conference, May 2008, Vermont, USA. IntroductionStereo on Ice ImagesOur AlgorithmResultsConclusion

Bibliography Daniel Scharstein, Richard Szeliski. A Taxonomy and Evaluation of Dense Two-Frame Stereo Correspondence Algorithms. IJCV D. Scharstein, R. Szeliski, Stereo matching with Non-linear Diffusion. Computer Science TR , Cornell University, Mar D. Scharstein, R. Szeliski. Stereo Matching with Non-linear diffusion. CVPR. June Jochen Alberty, Carsten Carstensen, Stefan Funken, Remarks Around 50 Lines of MATLAB: Short Finite Element Implementation, Numerical Algorithms,Volume 20, P. Persson, G.Strang. A simple mesh generator in Matlab. SIAM Review, Volume 46 (2), June IntroductionStereo on Ice ImagesOur AlgorithmResultsConclusion

Acknowledgements Dr. Chandra Kambhamettu Dr. Cathleen Geiger This work was made possible by National Science Foundation (NSF) Office of Polar Program grants, ANT and ARC IntroductionStereo on Ice ImagesOur AlgorithmResultsConclusion