Supplementary Material for A Locally Linear Regression Model for Boundary Preserving Regularization in Stereo Matching

Size: px
Start display at page:

Download "Supplementary Material for A Locally Linear Regression Model for Boundary Preserving Regularization in Stereo Matching"

Transcription

1 Supplementary Material for A Locally Linear Regression Model for Boundary Preserving Regularization in Stereo Matching Shengqi Zhu 1, Li Zhang 1, and Hailin Jin 2 1 University of Wisconsin - Madison 2 Adobe Systems Incorporated {sqzhu,lizhang}@cs.wisc.edu hljin@adobe.com In this supplementary material, we provide: (1) details in the upper bound calculation and (2) additional results on static images. A Details in Evaluating the Upper bound in Section 4.2 In our paper, we formulate our stereo matching as a global optimization as: Φ(D) = Φ 0 (D) + β 1 Φ 1 (D) + β 2 Φ 2 (D), (A.1) This section provides the details of finding upper bounds for the terms in Eq. (A.1). A.1 Φ 0 : Finding Upper bound for the Relaxed Data Term Recall that in our paper, Φ 0 is the data term. We calculate the upper bound of its relaxed version in two steps, as illustrated in Supplementary Figure 1. Step 1: Approximate the relaxed data term (red dashed line, which is piecewiselinear) by calculating quadratic B-splines (blue solid line); Step 2: Calculate the tightest upper bound Φ 0 (lowest green line) for the quadratic B-splines. Step 1: B-spline Approximation. To construct the quadratic B-spline approximation, we convolve the piecewise linear function with a box filter (of support [ 0.5, 0.5]). From the B-spline theory, the resulting B-spline approximation has the following property. Let L be the number of disparity levels. For each l {1, 2,, L}, if the three points on the relaxed data term at l 1, l, and l + 1 are on a straight line, the B-spline remains to be the the same line in the interval of [l 0.5, l + 0.5]. If the three points form a corner, the B-spline over [l 0.5, l+0.5] is a parabola that connects the two points on the original function at l 0.5 and l with first order continuity. In summary, within each interval [l 0.5, l + 0.5], the B-spline approximation of the relaxed data term is a parabola defined as β i (d) = p l d 2 + q l d + r l (A.2) where p l, q l, and r l are coefficients that vary with each pixel i; we drop the subscript i to simplify notation. This simple box-filtering treatment enables us to use unconstrained quadratic optimization to solve the stereo problem.

2 2 Shengqi Zhu, Li Zhang, and Hailin Jin 11 x Supplementary Fig. 1. (Same as Figure 1(d) in the paper.) Relationship among relaxed data term (dashed red), its B-spline approximation (solid blue), and the upper bound for the B-spline approximation (solid green) at the current disparity estimation (black circle). We search for the upper bound parabola with the lowest curvature to approximate the relaxed data term in our optimization. Step 2: Upper bound for the Quadratic B-spline. Precisely, let D be the current disparity map and d i be the current disparity value at pixel i (black circle in Supplementary Figure 1). We first evaluate the value β i (d i ) and the derivative β i (d i ) of the B-spline approximation at d i. The current quadratic upper bound β i ( ) (solid green line) is set to be a parabola that has the same value and derivative as β i ( ) at d i, using the following form β i (d) = τ i (d d i ) 2 + β i(d i ) (d d i ) + β i (d i ) (A.3) where τ i 0 is the only free parameter. When τ i, β i ( ) will have infinite curvature and surely be an exact upper bound for β i ( ). Therefore, our goal is to find the smallest τ i so that the upper bound curve β i ( ) has low curvature and matches as closely as possible to the β i ( ) curve. This is the same as finding, for each l, the smallest value for τ i so that β i ( ) β i ( ) over [l 0.5, l + 0.5] and then taking the maximum of these L smallest values. The procedure has a complexity of O(L) for each pixel. A.2 Φ 1 : Finding Upper bound for the First Order Term Φ 1 Here we show how to get Eq (14). From Eq (8), we have Φ 1 (D, A) = v ij (d 2 j + at i f ifi Ta i 2a T i f jd j ) + a T i Λa i i j = D T diag(v j )D + (a T i F ia i 2a T i G id), i (A.4) where diag(v j ) is a diagonal matrix with elements v j ; v j, F i, and G i are defined as: v j = v ij i F i = Λ + v ij f i fi T (A.5) j G i = [v i1 f 1, v i2 f 2,, v in f N ]. Fixing D, Eq. (A.4) is minimized when a i = F 1 i G i D. (A.6)

3 Supp: A Local Regression Model for Stereo Matching Regularization 3 where Substituting Eq. (A.6) into Eq. (A.4), we have Φ 1 (D) = D T H 1 D, H 1 = diag(v j ) i B Additional Results B.1 More Comparisons with [11] (A.7) G T i F 1 i G i. (A.8) We provide one more example to compare with [11], which uses pair-wise weighted L 1 penalty as regularization. As shown in Supplementary Figure 2, our method better preserves sharp boundaries and avoids producing isolated small regions. (a) Left Image (c) [11] with color bandwidth = 4 (e) [11] with color bandwidth = 2 (g) Our method (b) Ground Truth (d) Bad Pixel Rate 13.57% (non-occ) (f) Bad Pixel Rate 11.50% (non-occ) (h) Bad Pixel Rate 8.04% (non-occ) Supplementary Fig. 2. Comparison between our method and [11] on Art dataset. The weighted L 1 penalty used in [11] makes its results sensitively depend on the color bandwidth parameter. A large color bandwidth yields over-smoothing across discontinuity boundaries ((c) and (d)) and a small color bandwidth results in many isolated areas in the background due to texture ((e) and (f)). Our method ((g) and (h)) better preserves boundaries and avoids producing isolated small regions. B.2 Middlebury Results In this section, we report the results of our algorithm on all 31 Middlebury datasets, including the 4 benchmarks, the 2005 and the 2006 datasets. We set the same parameter as in Section 6.3 (β 1 = β 2 = , λ = 10 3 ) for all the experiments. The error rates are for the non-occluded bad pixels.

4 4 Shengqi Zhu, Li Zhang, and Hailin Jin Tsukuba, 2001 groundtruth our result bad pixel [1.05%] Venus, 2001 groundtruth our result bad pixel [0.29%] Teddy, 2003 groundtruth our result bad pixel [4.56%] Cones, 2003 groundtruth our result bad pixel [2.17%] Art, 2005 groundtruth our result bad pixel [8.04%]

5 Supp: A Local Regression Model for Stereo Matching Regularization 5 Books, 2005 groundtruth our result bad pixel [9.08%] Dolls, 2005 groundtruth our result bad pixel [3.48%] Laundry, 2005 groundtruth our result bad pixel [9.76%] Moebius, 2005 groundtruth our result bad pixel [9.60%] Reindeer, 2005 groundtruth our result bad pixel [4.01%]

6 6 Shengqi Zhu, Li Zhang, and Hailin Jin Aloe, 2006 groundtruth our result bad pixel [4.89%] Baby1, 2006 groundtruth our result bad pixel [2.85%] Baby2, 2006 groundtruth our result bad pixel [3.82%] Baby3, 2006 groundtruth our result bad pixel [2.70%] Bowling1, 2006 groundtruth our result bad pixel [14.27%]

7 Supp: A Local Regression Model for Stereo Matching Regularization 7 Bowling2, 2006 groundtruth our result bad pixel [4.59%] Cloth1, 2006 groundtruth our result bad pixel [0.91%] Cloth2, 2006 groundtruth our result bad pixel [2.13%] Cloth3, 2006 groundtruth our result bad pixel [1.29%] Cloth4, 2006 groundtruth our result bad pixel [1.19%]

8 8 Shengqi Zhu, Li Zhang, and Hailin Jin Flowerpots, 2006 groundtruth our result bad pixel [11.78%] Lampshade1, 2006 groundtruth our result bad pixel [10.01%] Lampshade2, 2006 groundtruth our result bad pixel [3.56%] Midd1, 2006 groundtruth our result bad pixel [46.52%] Midd2, 2006 groundtruth our result bad pixel [43.85%]

9 Supp: A Local Regression Model for Stereo Matching Regularization 9 Monopoly, 2006 groundtruth our result bad pixel [36.80%] Plastic, 2006 groundtruth our result bad pixel [18.28%] Rocks1, 2006 groundtruth our result bad pixel [1.99%] Rocks2, 2006 groundtruth our result bad pixel [1.10%] Wood1, 2006 groundtruth our result bad pixel [3.34%]

10 10 Shengqi Zhu, Li Zhang, and Hailin Jin Wood2, 2006 groundtruth our result bad pixel [0.39%]

A Locally Linear Regression Model for Boundary Preserving Regularization in Stereo Matching

A Locally Linear Regression Model for Boundary Preserving Regularization in Stereo Matching A Locally Linear Regression Model for Boundary Preserving Regularization in Stereo Matching Shengqi Zhu 1, Li Zhang 1, and Hailin Jin 2 1 University of Wisconsin - Madison 2 Adobe Systems Incorporated

More information

Supplementary Material for ECCV 2012 Paper: Extracting 3D Scene-consistent Object Proposals and Depth from Stereo Images

Supplementary Material for ECCV 2012 Paper: Extracting 3D Scene-consistent Object Proposals and Depth from Stereo Images Supplementary Material for ECCV 2012 Paper: Extracting 3D Scene-consistent Object Proposals and Depth from Stereo Images Michael Bleyer 1, Christoph Rhemann 1,2, and Carsten Rother 2 1 Vienna University

More information

FILTER BASED ALPHA MATTING FOR DEPTH IMAGE BASED RENDERING. Naoki Kodera, Norishige Fukushima and Yutaka Ishibashi

FILTER BASED ALPHA MATTING FOR DEPTH IMAGE BASED RENDERING. Naoki Kodera, Norishige Fukushima and Yutaka Ishibashi FILTER BASED ALPHA MATTING FOR DEPTH IMAGE BASED RENDERING Naoki Kodera, Norishige Fukushima and Yutaka Ishibashi Graduate School of Engineering, Nagoya Institute of Technology ABSTRACT In this paper,

More information

Fast Stereo Matching of Feature Links

Fast Stereo Matching of Feature Links Fast Stereo Matching of Feature Links 011.05.19 Chang-il, Kim Introduction Stereo matching? interesting topics of computer vision researches To determine a disparity between stereo images A fundamental

More information

Correctness Prediction, Accuracy Improvement and Generalization of Stereo Matching using Supervised Learning

Correctness Prediction, Accuracy Improvement and Generalization of Stereo Matching using Supervised Learning International Journal of Computer Vision manuscript No. (will be inserted by the editor) Correctness Prediction, Accuracy Improvement and Generalization of Stereo Matching using Supervised Learning Aristotle

More information

Subpixel accurate refinement of disparity maps using stereo correspondences

Subpixel accurate refinement of disparity maps using stereo correspondences Subpixel accurate refinement of disparity maps using stereo correspondences Matthias Demant Lehrstuhl für Mustererkennung, Universität Freiburg Outline 1 Introduction and Overview 2 Refining the Cost Volume

More information

MGM: A Significantly More Global Matching for Stereovision

MGM: A Significantly More Global Matching for Stereovision G. FACCIOLO, C. DE FRANCHIS, E. MEINHARDT: MORE GLOBAL MATCHING 1 MGM: A Significantly More Global Matching for Stereovision Gabriele Facciolo 1 facciolo@cmla.ens-cachan.fr Carlo de Franchis 2 carlo.de-franchis@ens-cachan.fr

More information

Filter Flow: Supplemental Material

Filter Flow: Supplemental Material Filter Flow: Supplemental Material Steven M. Seitz University of Washington Simon Baker Microsoft Research We include larger images and a number of additional results obtained using Filter Flow [5]. 1

More information

Lecture 2.2 Cubic Splines

Lecture 2.2 Cubic Splines Lecture. Cubic Splines Cubic Spline The equation for a single parametric cubic spline segment is given by 4 i t Bit t t t i (..) where t and t are the parameter values at the beginning and end of the segment.

More information

Segment-Tree based Cost Aggregation for Stereo Matching

Segment-Tree based Cost Aggregation for Stereo Matching Segment-Tree based Cost Aggregation for Stereo Matching Xing Mei 1, Xun Sun 2, Weiming Dong 1, Haitao Wang 2, Xiaopeng Zhang 1 1 NLPR, Institute of Automation, Chinese Academy of Sciences, Beijing, China

More information

Performance of Stereo Methods in Cluttered Scenes

Performance of Stereo Methods in Cluttered Scenes Performance of Stereo Methods in Cluttered Scenes Fahim Mannan and Michael S. Langer School of Computer Science McGill University Montreal, Quebec H3A 2A7, Canada { fmannan, langer}@cim.mcgill.ca Abstract

More information

Robert Collins CSE486, Penn State. Lecture 09: Stereo Algorithms

Robert Collins CSE486, Penn State. Lecture 09: Stereo Algorithms Lecture 09: Stereo Algorithms left camera located at (0,0,0) Recall: Simple Stereo System Y y Image coords of point (X,Y,Z) Left Camera: x T x z (, ) y Z (, ) x (X,Y,Z) z X right camera located at (T x,0,0)

More information

Data Term. Michael Bleyer LVA Stereo Vision

Data Term. Michael Bleyer LVA Stereo Vision Data Term Michael Bleyer LVA Stereo Vision What happened last time? We have looked at our energy function: E ( D) = m( p, dp) + p I < p, q > N s( p, q) We have learned about an optimization algorithm that

More information

Supplementary Material: Piecewise Planar and Compact Floorplan Reconstruction from Images

Supplementary Material: Piecewise Planar and Compact Floorplan Reconstruction from Images Supplementary Material: Piecewise Planar and Compact Floorplan Reconstruction from Images Ricardo Cabral Carnegie Mellon University rscabral@cmu.edu Yasutaka Furukawa Washington University in St. Louis

More information

SEMI-GLOBAL MATCHING WITH SELF-ADJUSTING PENALTIES

SEMI-GLOBAL MATCHING WITH SELF-ADJUSTING PENALTIES SEMI-GLOBAL MATCHING WITH SELF-ADJUSTING PENALTIES E. Karkalou*, C. Stentoumis, G. Karras (ellikarkalou at gmail.com, cstent at mail.ntua.gr, gkarras at central.ntua.gr) Laboratory of Photogrammetry, National

More information

The Naked Truth about Cost Functions for Stereo Matching

The Naked Truth about Cost Functions for Stereo Matching The Naked Truth about Cost Functions for Stereo Matching Simon Hermann and Reinhard Klette The.enpeda.. Project, The University of Auckland Auckland, New Zealand Abstract. This paper reports about the

More information

Lecture VIII. Global Approximation Methods: I

Lecture VIII. Global Approximation Methods: I Lecture VIII Global Approximation Methods: I Gianluca Violante New York University Quantitative Macroeconomics G. Violante, Global Methods p. 1 /29 Global function approximation Global methods: function

More information

Almost Curvature Continuous Fitting of B-Spline Surfaces

Almost Curvature Continuous Fitting of B-Spline Surfaces Journal for Geometry and Graphics Volume 2 (1998), No. 1, 33 43 Almost Curvature Continuous Fitting of B-Spline Surfaces Márta Szilvási-Nagy Department of Geometry, Mathematical Institute, Technical University

More information

Point, Line Segment, and Region-Based Stereo Matching for Mobile Robotics

Point, Line Segment, and Region-Based Stereo Matching for Mobile Robotics Point, Line Segment, and Region-Based Stereo Matching for Mobile Robotics A thesis presented by Brian P. McKinnon to The Department of Computer Science in partial fulfillment of the requirements for the

More information

Section 9.3 Graphing Quadratic Functions

Section 9.3 Graphing Quadratic Functions Section 9.3 Graphing Quadratic Functions A Quadratic Function is an equation that can be written in the following Standard Form., where a 0. Every quadratic function has a U-shaped graph called a. If the

More information

Stereo Matching with Reliable Disparity Propagation

Stereo Matching with Reliable Disparity Propagation Stereo Matching with Reliable Disparity Propagation Xun Sun, Xing Mei, Shaohui Jiao, Mingcai Zhou, Haitao Wang Samsung Advanced Institute of Technology, China Lab Beijing, China xunshine.sun,xing.mei,sh.jiao,mingcai.zhou,ht.wang@samsung.com

More information

Input. Output. Problem Definition. Rectified stereo image pair All correspondences lie in same scan lines

Input. Output. Problem Definition. Rectified stereo image pair All correspondences lie in same scan lines Problem Definition 3 Input Rectified stereo image pair All correspondences lie in same scan lines Output Disparity map of the reference view Foreground: large disparity Background: small disparity Matching

More information

Improved depth map estimation in Stereo Vision

Improved depth map estimation in Stereo Vision Improved depth map estimation in Stereo Vision Hajer Fradi and and Jean-Luc Dugelay EURECOM, Sophia Antipolis, France ABSTRACT In this paper, we present a new approach for dense stereo matching which is

More information

Computer Graphics / Animation

Computer Graphics / Animation Computer Graphics / Animation Artificial object represented by the number of points in space and time (for moving, animated objects). Essential point: How do you interpolate these points in space and time?

More information

Computational Physics PHYS 420

Computational Physics PHYS 420 Computational Physics PHYS 420 Dr Richard H. Cyburt Assistant Professor of Physics My office: 402c in the Science Building My phone: (304) 384-6006 My email: rcyburt@concord.edu My webpage: www.concord.edu/rcyburt

More information

Unit 1 Day 4 Notes Piecewise Functions

Unit 1 Day 4 Notes Piecewise Functions AFM Unit 1 Day 4 Notes Piecewise Functions Name Date We have seen many graphs that are expressed as single equations and are continuous over a domain of the Real numbers. We have also seen the "discrete"

More information

February 2017 (1/20) 2 Piecewise Polynomial Interpolation 2.2 (Natural) Cubic Splines. MA378/531 Numerical Analysis II ( NA2 )

February 2017 (1/20) 2 Piecewise Polynomial Interpolation 2.2 (Natural) Cubic Splines. MA378/531 Numerical Analysis II ( NA2 ) f f f f f (/2).9.8.7.6.5.4.3.2. S Knots.7.6.5.4.3.2. 5 5.2.8.6.4.2 S Knots.2 5 5.9.8.7.6.5.4.3.2..9.8.7.6.5.4.3.2. S Knots 5 5 S Knots 5 5 5 5.35.3.25.2.5..5 5 5.6.5.4.3.2. 5 5 4 x 3 3.5 3 2.5 2.5.5 5

More information

Stereo Vision in Structured Environments by Consistent Semi-Global Matching

Stereo Vision in Structured Environments by Consistent Semi-Global Matching Stereo Vision in Structured Environments by Consistent Semi-Global Matching Heiko Hirschmüller Institute of Robotics and Mechatronics Oberpfaffenhofen German Aerospace Center (DLR), 82230 Wessling, Germany.

More information

Lecture IV Bézier Curves

Lecture IV Bézier Curves Lecture IV Bézier Curves Why Curves? Why Curves? Why Curves? Why Curves? Why Curves? Linear (flat) Curved Easier More pieces Looks ugly Complicated Fewer pieces Looks smooth What is a curve? Intuitively:

More information

Bayesian Background Estimation

Bayesian Background Estimation Bayesian Background Estimation mum knot spacing was determined by the width of the signal structure that one wishes to exclude from the background curve. This paper extends the earlier work in two important

More information

Boundary descriptors. Representation REPRESENTATION & DESCRIPTION. Descriptors. Moore boundary tracking

Boundary descriptors. Representation REPRESENTATION & DESCRIPTION. Descriptors. Moore boundary tracking Representation REPRESENTATION & DESCRIPTION After image segmentation the resulting collection of regions is usually represented and described in a form suitable for higher level processing. Most important

More information

PatchMatch Stereo - Stereo Matching with Slanted Support Windows

PatchMatch Stereo - Stereo Matching with Slanted Support Windows M. BLEYER, C. RHEMANN, C. ROTHER: PATCHMATCH STEREO 1 PatchMatch Stereo - Stereo Matching with Slanted Support Windows Michael Bleyer 1 bleyer@ims.tuwien.ac.at Christoph Rhemann 1 rhemann@ims.tuwien.ac.at

More information

Comparison of stereo inspired optical flow estimation techniques

Comparison of stereo inspired optical flow estimation techniques Comparison of stereo inspired optical flow estimation techniques ABSTRACT The similarity of the correspondence problems in optical flow estimation and disparity estimation techniques enables methods to

More information

Rational Bezier Curves

Rational Bezier Curves Rational Bezier Curves Use of homogeneous coordinates Rational spline curve: define a curve in one higher dimension space, project it down on the homogenizing variable Mathematical formulation: n P(u)

More information

An introduction to interpolation and splines

An introduction to interpolation and splines An introduction to interpolation and splines Kenneth H. Carpenter, EECE KSU November 22, 1999 revised November 20, 2001, April 24, 2002, April 14, 2004 1 Introduction Suppose one wishes to draw a curve

More information

Multi-Label Moves for Multi-Label Energies

Multi-Label Moves for Multi-Label Energies Multi-Label Moves for Multi-Label Energies Olga Veksler University of Western Ontario some work is joint with Olivier Juan, Xiaoqing Liu, Yu Liu Outline Review multi-label optimization with graph cuts

More information

Curve fitting using linear models

Curve fitting using linear models Curve fitting using linear models Rasmus Waagepetersen Department of Mathematics Aalborg University Denmark September 28, 2012 1 / 12 Outline for today linear models and basis functions polynomial regression

More information

Lecture 9. Curve fitting. Interpolation. Lecture in Numerical Methods from 28. April 2015 UVT. Lecture 9. Numerical. Interpolation his o

Lecture 9. Curve fitting. Interpolation. Lecture in Numerical Methods from 28. April 2015 UVT. Lecture 9. Numerical. Interpolation his o Curve fitting. Lecture in Methods from 28. April 2015 to ity Interpolation FIGURE A S Splines Piecewise relat UVT Agenda of today s lecture 1 Interpolation Idea 2 3 4 5 6 Splines Piecewise Interpolation

More information

COMBINING DEPTH INFORMATION AND LOCAL EDGE DETECTION FOR STEREO IMAGE ENHANCEMENT

COMBINING DEPTH INFORMATION AND LOCAL EDGE DETECTION FOR STEREO IMAGE ENHANCEMENT 20th European Signal Processing Conference (EUSIPCO 2012) Bucharest, Romania, August 27-31, 2012 COMBINING DEPTH INFORMATION AND LOCAL EDGE DETECTION FOR STEREO IMAGE ENHANCEMENT Walid Hachicha 1 Azeddine

More information

FMA901F: Machine Learning Lecture 3: Linear Models for Regression. Cristian Sminchisescu

FMA901F: Machine Learning Lecture 3: Linear Models for Regression. Cristian Sminchisescu FMA901F: Machine Learning Lecture 3: Linear Models for Regression Cristian Sminchisescu Machine Learning: Frequentist vs. Bayesian In the frequentist setting, we seek a fixed parameter (vector), with value(s)

More information

Graph Cut based Continuous Stereo Matching using Locally Shared Labels

Graph Cut based Continuous Stereo Matching using Locally Shared Labels Graph Cut based Continuous Stereo Matching using Locally Shared Labels Tatsunori Taniai University of Tokyo, Japan taniai@iis.u-tokyo.ac.jp Yasuyuki Matsushita Microsoft Research Asia, China yasumat@microsoft.com

More information

segments. The geometrical relationship of adjacent planes such as parallelism and intersection is employed for determination of whether two planes sha

segments. The geometrical relationship of adjacent planes such as parallelism and intersection is employed for determination of whether two planes sha A New Segment-based Stereo Matching using Graph Cuts Daolei Wang National University of Singapore EA #04-06, Department of Mechanical Engineering Control and Mechatronics Laboratory, 10 Kent Ridge Crescent

More information

SUPPLEMENTARY FILE S1: 3D AIRWAY TUBE RECONSTRUCTION AND CELL-BASED MECHANICAL MODEL. RELATED TO FIGURE 1, FIGURE 7, AND STAR METHODS.

SUPPLEMENTARY FILE S1: 3D AIRWAY TUBE RECONSTRUCTION AND CELL-BASED MECHANICAL MODEL. RELATED TO FIGURE 1, FIGURE 7, AND STAR METHODS. SUPPLEMENTARY FILE S1: 3D AIRWAY TUBE RECONSTRUCTION AND CELL-BASED MECHANICAL MODEL. RELATED TO FIGURE 1, FIGURE 7, AND STAR METHODS. 1. 3D AIRWAY TUBE RECONSTRUCTION. RELATED TO FIGURE 1 AND STAR METHODS

More information

Lecture 8. Divided Differences,Least-Squares Approximations. Ceng375 Numerical Computations at December 9, 2010

Lecture 8. Divided Differences,Least-Squares Approximations. Ceng375 Numerical Computations at December 9, 2010 Lecture 8, Ceng375 Numerical Computations at December 9, 2010 Computer Engineering Department Çankaya University 8.1 Contents 1 2 3 8.2 : These provide a more efficient way to construct an interpolating

More information

3D RECONSTRUCTION FROM STEREO/ RANGE IMAGES

3D RECONSTRUCTION FROM STEREO/ RANGE IMAGES University of Kentucky UKnowledge University of Kentucky Master's Theses Graduate School 2007 3D RECONSTRUCTION FROM STEREO/ RANGE IMAGES Qingxiong Yang University of Kentucky, qyang2@uky.edu Click here

More information

Introduction to Optimization

Introduction to Optimization Introduction to Optimization Second Order Optimization Methods Marc Toussaint U Stuttgart Planned Outline Gradient-based optimization (1st order methods) plain grad., steepest descent, conjugate grad.,

More information

LETTER Local and Nonlocal Color Line Models for Image Matting

LETTER Local and Nonlocal Color Line Models for Image Matting 1814 IEICE TRANS. FUNDAMENTALS, VOL.E97 A, NO.8 AUGUST 2014 LETTER Local and Nonlocal Color Line Models for Image Matting Byoung-Kwang KIM a), Meiguang JIN, Nonmembers, and Woo-Jin SONG, Member SUMMARY

More information

Segmentation Based Stereo. Michael Bleyer LVA Stereo Vision

Segmentation Based Stereo. Michael Bleyer LVA Stereo Vision Segmentation Based Stereo Michael Bleyer LVA Stereo Vision What happened last time? Once again, we have looked at our energy function: E ( D) = m( p, dp) + p I < p, q > We have investigated the matching

More information

Fundamentals of Stereo Vision Michael Bleyer LVA Stereo Vision

Fundamentals of Stereo Vision Michael Bleyer LVA Stereo Vision Fundamentals of Stereo Vision Michael Bleyer LVA Stereo Vision What Happened Last Time? Human 3D perception (3D cinema) Computational stereo Intuitive explanation of what is meant by disparity Stereo matching

More information

Learning and Inferring Depth from Monocular Images. Jiyan Pan April 1, 2009

Learning and Inferring Depth from Monocular Images. Jiyan Pan April 1, 2009 Learning and Inferring Depth from Monocular Images Jiyan Pan April 1, 2009 Traditional ways of inferring depth Binocular disparity Structure from motion Defocus Given a single monocular image, how to infer

More information

Systems of Equations and Inequalities. Copyright Cengage Learning. All rights reserved.

Systems of Equations and Inequalities. Copyright Cengage Learning. All rights reserved. 5 Systems of Equations and Inequalities Copyright Cengage Learning. All rights reserved. 5.5 Systems of Inequalities Copyright Cengage Learning. All rights reserved. Objectives Graphing an Inequality Systems

More information

A C 2 Four-Point Subdivision Scheme with Fourth Order Accuracy and its Extensions

A C 2 Four-Point Subdivision Scheme with Fourth Order Accuracy and its Extensions A C 2 Four-Point Subdivision Scheme with Fourth Order Accuracy and its Extensions Nira Dyn School of Mathematical Sciences Tel Aviv University Michael S. Floater Department of Informatics University of

More information

WITH RECENT advances in 3-D display technology,

WITH RECENT advances in 3-D display technology, 784 IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY, VOL. 23, NO. 5, MAY 2013 Efficient Disparity Estimation Using Hierarchical Bilateral Disparity Structure Based Graph Cut Algorithm With

More information

A popular method for moving beyond linearity. 2. Basis expansion and regularization 1. Examples of transformations. Piecewise-polynomials and splines

A popular method for moving beyond linearity. 2. Basis expansion and regularization 1. Examples of transformations. Piecewise-polynomials and splines A popular method for moving beyond linearity 2. Basis expansion and regularization 1 Idea: Augment the vector inputs x with additional variables which are transformation of x use linear models in this

More information

Consider functions such that then satisfies these properties: So is represented by the cubic polynomials on on and on.

Consider functions such that then satisfies these properties: So is represented by the cubic polynomials on on and on. 1 of 9 3/1/2006 2:28 PM ne previo Next: Trigonometric Interpolation Up: Spline Interpolation Previous: Piecewise Linear Case Cubic Splines A piece-wise technique which is very popular. Recall the philosophy

More information

Unit: Quadratic Functions

Unit: Quadratic Functions Unit: Quadratic Functions Learning increases when you have a goal to work towards. Use this checklist as guide to track how well you are grasping the material. In the center column, rate your understand

More information

In Defense of 3D-Label Stereo

In Defense of 3D-Label Stereo 2013 IEEE Conference on Computer Vision and Pattern Recognition In Defense of 3D-Label Stereo Carl Olsson Johannes Ulén Centre for Mathematical Sciences Lund University, Sweden calle@maths.lth.se ulen@maths.lth.se

More information

See the course website for important information about collaboration and late policies, as well as where and when to turn in assignments.

See the course website for important information about collaboration and late policies, as well as where and when to turn in assignments. COS Homework # Due Tuesday, February rd See the course website for important information about collaboration and late policies, as well as where and when to turn in assignments. Data files The questions

More information

Mar. 20 Math 2335 sec 001 Spring 2014

Mar. 20 Math 2335 sec 001 Spring 2014 Mar. 20 Math 2335 sec 001 Spring 2014 Chebyshev Polynomials Definition: For an integer n 0 define the function ( ) T n (x) = cos n cos 1 (x), 1 x 1. It can be shown that T n is a polynomial of degree n.

More information

Supplemental Material: A Dataset and Evaluation Methodology for Depth Estimation on 4D Light Fields

Supplemental Material: A Dataset and Evaluation Methodology for Depth Estimation on 4D Light Fields Supplemental Material: A Dataset and Evaluation Methodology for Depth Estimation on 4D Light Fields Katrin Honauer 1, Ole Johannsen 2, Daniel Kondermann 1, Bastian Goldluecke 2 1 HCI, Heidelberg University

More information

Unit 3, Lesson 3.1 Creating and Graphing Equations Using Standard Form

Unit 3, Lesson 3.1 Creating and Graphing Equations Using Standard Form Unit 3, Lesson 3.1 Creating and Graphing Equations Using Standard Form Imagine the path of a basketball as it leaves a player s hand and swooshes through the net. Or, imagine the path of an Olympic diver

More information

Fall CSCI 420: Computer Graphics. 4.2 Splines. Hao Li.

Fall CSCI 420: Computer Graphics. 4.2 Splines. Hao Li. Fall 2014 CSCI 420: Computer Graphics 4.2 Splines Hao Li http://cs420.hao-li.com 1 Roller coaster Next programming assignment involves creating a 3D roller coaster animation We must model the 3D curve

More information

SPARSE RECONSTRUCTION FOR DISPARITY MAPS USING COMBINED WAVELET AND CONTOURLET TRANSFORMS. Lee-kang Liu and Truong Nguyen

SPARSE RECONSTRUCTION FOR DISPARITY MAPS USING COMBINED WAVELET AND CONTOURLET TRANSFORMS. Lee-kang Liu and Truong Nguyen SPARSE RECONSTRUCTION FOR DISPARITY MAPS USING COMBINED WAVELET AND CONTOURLET TRANSFORMS Lee-kang Liu and Truong Nguyen University of California, San Diego Electrical and Computer Engineering http://videoprocessing.ucsd.edu

More information

2009 Conference for Visual Media Production

2009 Conference for Visual Media Production 009 Conference for Visual Media Production COECTIG SHAPESS VAIATIOS I STEEO IMAGE PAIS C. Doutre and P. asiopoulos Departement of Electrical and Computer gineering Uniersity of British Columbia, Vancouer,

More information

Supplementary Material for Zoom and Learn: Generalizing Deep Stereo Matching to Novel Domains

Supplementary Material for Zoom and Learn: Generalizing Deep Stereo Matching to Novel Domains Supplementary Material for Zoom and Learn: Generalizing Deep Stereo Matching to Novel Domains Jiahao Pang 1 Wenxiu Sun 1 Chengxi Yang 1 Jimmy Ren 1 Ruichao Xiao 1 Jin Zeng 1 Liang Lin 1,2 1 SenseTime Research

More information

How to create shapes. Drawing basic shapes. Adobe Photoshop Elements 8 guide

How to create shapes. Drawing basic shapes. Adobe Photoshop Elements 8 guide How to create shapes With the shape tools in Adobe Photoshop Elements, you can draw perfect geometric shapes, regardless of your artistic ability or illustration experience. The first step to drawing shapes

More information

A toolbox of smooths. Simon Wood Mathematical Sciences, University of Bath, U.K.

A toolbox of smooths. Simon Wood Mathematical Sciences, University of Bath, U.K. A toolbo of smooths Simon Wood Mathematical Sciences, University of Bath, U.K. Smooths for semi-parametric GLMs To build adequate semi-parametric GLMs requires that we use functions with appropriate properties.

More information

Central issues in modelling

Central issues in modelling Central issues in modelling Construct families of curves, surfaces and volumes that can represent common objects usefully; are easy to interact with; interaction includes: manual modelling; fitting to

More information

Parameterization. Michael S. Floater. November 10, 2011

Parameterization. Michael S. Floater. November 10, 2011 Parameterization Michael S. Floater November 10, 2011 Triangular meshes are often used to represent surfaces, at least initially, one reason being that meshes are relatively easy to generate from point

More information

A New Approach for Stereo Matching Algorithm with Dynamic Programming

A New Approach for Stereo Matching Algorithm with Dynamic Programming A New Approach for Stereo Matching Algorithm with Dynamic Programming Miss. Priyanka M. Lohot PG Student, Computer Engineering, Shah & Anchor Kutchhi Engineering College, Mumbai University. priya.lohot@gmail.com

More information

Stereo Wrap + Motion. Computer Vision I. CSE252A Lecture 17

Stereo Wrap + Motion. Computer Vision I. CSE252A Lecture 17 Stereo Wrap + Motion CSE252A Lecture 17 Some Issues Ambiguity Window size Window shape Lighting Half occluded regions Problem of Occlusion Stereo Constraints CONSTRAINT BRIEF DESCRIPTION 1-D Epipolar Search

More information

Stereo Correspondence with Occlusions using Graph Cuts

Stereo Correspondence with Occlusions using Graph Cuts Stereo Correspondence with Occlusions using Graph Cuts EE368 Final Project Matt Stevens mslf@stanford.edu Zuozhen Liu zliu2@stanford.edu I. INTRODUCTION AND MOTIVATION Given two stereo images of a scene,

More information

A C 2 Four-Point Subdivision Scheme with Fourth Order Accuracy and its Extensions

A C 2 Four-Point Subdivision Scheme with Fourth Order Accuracy and its Extensions A C 2 Four-Point Subdivision Scheme with Fourth Order Accuracy and its Extensions Nira Dyn Michael S. Floater Kai Hormann Abstract. We present a new four-point subdivision scheme that generates C 2 curves.

More information

What have we leaned so far?

What have we leaned so far? What have we leaned so far? Camera structure Eye structure Project 1: High Dynamic Range Imaging What have we learned so far? Image Filtering Image Warping Camera Projection Model Project 2: Panoramic

More information

Algebra II Quadratic Functions

Algebra II Quadratic Functions 1 Algebra II Quadratic Functions 2014-10-14 www.njctl.org 2 Ta b le o f C o n te n t Key Terms click on the topic to go to that section Explain Characteristics of Quadratic Functions Combining Transformations

More information

A Constant-Space Belief Propagation Algorithm for Stereo Matching

A Constant-Space Belief Propagation Algorithm for Stereo Matching A Constant-Space Belief Propagation Algorithm for Stereo Matching Qingxiong Yang Liang Wang Narendra Ahuja University of Illinois at Urbana Champaign University of Kentucky http://vision.ai.uiuc.edu/ qyang6/

More information

Learning Two-View Stereo Matching

Learning Two-View Stereo Matching Learning Two-View Stereo Matching Jianxiong Xiao Jingni Chen Dit-Yan Yeung Long Quan Department of Computer Science and Engineering The Hong Kong University of Science and Technology The 10th European

More information

INF3320 Computer Graphics and Discrete Geometry

INF3320 Computer Graphics and Discrete Geometry INF3320 Computer Graphics and Discrete Geometry More smooth Curves and Surfaces Christopher Dyken, Michael Floater and Martin Reimers 10.11.2010 Page 1 More smooth Curves and Surfaces Akenine-Möller, Haines

More information

CS 450 Numerical Analysis. Chapter 7: Interpolation

CS 450 Numerical Analysis. Chapter 7: Interpolation Lecture slides based on the textbook Scientific Computing: An Introductory Survey by Michael T. Heath, copyright c 2018 by the Society for Industrial and Applied Mathematics. http://www.siam.org/books/cl80

More information

Curves. Computer Graphics CSE 167 Lecture 11

Curves. Computer Graphics CSE 167 Lecture 11 Curves Computer Graphics CSE 167 Lecture 11 CSE 167: Computer graphics Polynomial Curves Polynomial functions Bézier Curves Drawing Bézier curves Piecewise Bézier curves Based on slides courtesy of Jurgen

More information

Limitations of Matrix Completion via Trace Norm Minimization

Limitations of Matrix Completion via Trace Norm Minimization Limitations of Matrix Completion via Trace Norm Minimization ABSTRACT Xiaoxiao Shi Computer Science Department University of Illinois at Chicago xiaoxiao@cs.uic.edu In recent years, compressive sensing

More information

Sung-Eui Yoon ( 윤성의 )

Sung-Eui Yoon ( 윤성의 ) CS480: Computer Graphics Curves and Surfaces Sung-Eui Yoon ( 윤성의 ) Course URL: http://jupiter.kaist.ac.kr/~sungeui/cg Today s Topics Surface representations Smooth curves Subdivision 2 Smooth Curves and

More information

Global Stereo Matching Leveraged by Sparse Ground Control Points

Global Stereo Matching Leveraged by Sparse Ground Control Points Global Stereo Matching Leveraged by Sparse Ground Control Points Liang Wang Ruigang Yang Center for Visualization and Virtual Environments, University of Kentucky 329 Rose Street, Lexington, KY, 40506-0633,

More information

Curve Representation ME761A Instructor in Charge Prof. J. Ramkumar Department of Mechanical Engineering, IIT Kanpur

Curve Representation ME761A Instructor in Charge Prof. J. Ramkumar Department of Mechanical Engineering, IIT Kanpur Curve Representation ME761A Instructor in Charge Prof. J. Ramkumar Department of Mechanical Engineering, IIT Kanpur Email: jrkumar@iitk.ac.in Curve representation 1. Wireframe models There are three types

More information

Regularization and Markov Random Fields (MRF) CS 664 Spring 2008

Regularization and Markov Random Fields (MRF) CS 664 Spring 2008 Regularization and Markov Random Fields (MRF) CS 664 Spring 2008 Regularization in Low Level Vision Low level vision problems concerned with estimating some quantity at each pixel Visual motion (u(x,y),v(x,y))

More information

MATLAB User Guide for Depth Reconstruction from Sparse Samples

MATLAB User Guide for Depth Reconstruction from Sparse Samples MATLAB User Guide for Depth Reconstruction from Sparse Samples Lee-Kang Liu, Stanley H. Chan and Truong Q. Nguyen This MATLAB user guide presents the instructions of how to use the MATLAB functions accompanied

More information

Fast Approximate Energy Minimization via Graph Cuts

Fast Approximate Energy Minimization via Graph Cuts IEEE Transactions on PAMI, vol. 23, no. 11, pp. 1222-1239 p.1 Fast Approximate Energy Minimization via Graph Cuts Yuri Boykov, Olga Veksler and Ramin Zabih Abstract Many tasks in computer vision involve

More information

Digital Image Processing Fundamentals

Digital Image Processing Fundamentals Ioannis Pitas Digital Image Processing Fundamentals Chapter 7 Shape Description Answers to the Chapter Questions Thessaloniki 1998 Chapter 7: Shape description 7.1 Introduction 1. Why is invariance to

More information

Splines. Parameterization of a Curve. Curve Representations. Roller coaster. What Do We Need From Curves in Computer Graphics? Modeling Complex Shapes

Splines. Parameterization of a Curve. Curve Representations. Roller coaster. What Do We Need From Curves in Computer Graphics? Modeling Complex Shapes CSCI 420 Computer Graphics Lecture 8 Splines Jernej Barbic University of Southern California Hermite Splines Bezier Splines Catmull-Rom Splines Other Cubic Splines [Angel Ch 12.4-12.12] Roller coaster

More information

1. Suppose that the equation F (x, y, z) = 0 implicitly defines each of the three variables x, y, and z as functions of the other two:

1. Suppose that the equation F (x, y, z) = 0 implicitly defines each of the three variables x, y, and z as functions of the other two: Final Solutions. Suppose that the equation F (x, y, z) implicitly defines each of the three variables x, y, and z as functions of the other two: z f(x, y), y g(x, z), x h(y, z). If F is differentiable

More information

9.1: GRAPHING QUADRATICS ALGEBRA 1

9.1: GRAPHING QUADRATICS ALGEBRA 1 9.1: GRAPHING QUADRATICS ALGEBRA 1 OBJECTIVES I will be able to graph quadratics: Given in Standard Form Given in Vertex Form Given in Intercept Form What does the graph of a quadratic look like? https://www.desmos.com/calculator

More information

MA 323 Geometric Modelling Course Notes: Day 14 Properties of Bezier Curves

MA 323 Geometric Modelling Course Notes: Day 14 Properties of Bezier Curves MA 323 Geometric Modelling Course Notes: Day 14 Properties of Bezier Curves David L. Finn In this section, we discuss the geometric properties of Bezier curves. These properties are either implied directly

More information

Supplementary Material: Pixelwise Instance Segmentation with a Dynamically Instantiated Network

Supplementary Material: Pixelwise Instance Segmentation with a Dynamically Instantiated Network Supplementary Material: Pixelwise Instance Segmentation with a Dynamically Instantiated Network Anurag Arnab and Philip H.S. Torr University of Oxford {anurag.arnab, philip.torr}@eng.ox.ac.uk 1. Introduction

More information

Corner Detection using Difference Chain Code as Curvature

Corner Detection using Difference Chain Code as Curvature Third International IEEE Conference on Signal-Image Technologies technologies and Internet-Based System Corner Detection using Difference Chain Code as Curvature Neeta Nain Vijay Laxmi Bhavitavya Bhadviya

More information

Precalculus Notes: Unit 7 Systems of Equations and Matrices

Precalculus Notes: Unit 7 Systems of Equations and Matrices Date: 7.1, 7. Solving Systems of Equations: Graphing, Substitution, Elimination Syllabus Objectives: 8.1 The student will solve a given system of equations or system of inequalities. Solution of a System

More information

CHAPTER 6 Parametric Spline Curves

CHAPTER 6 Parametric Spline Curves CHAPTER 6 Parametric Spline Curves When we introduced splines in Chapter 1 we focused on spline curves, or more precisely, vector valued spline functions. In Chapters 2 and 4 we then established the basic

More information

Last update: May 4, Vision. CMSC 421: Chapter 24. CMSC 421: Chapter 24 1

Last update: May 4, Vision. CMSC 421: Chapter 24. CMSC 421: Chapter 24 1 Last update: May 4, 200 Vision CMSC 42: Chapter 24 CMSC 42: Chapter 24 Outline Perception generally Image formation Early vision 2D D Object recognition CMSC 42: Chapter 24 2 Perception generally Stimulus

More information

Colour Segmentation-based Computation of Dense Optical Flow with Application to Video Object Segmentation

Colour Segmentation-based Computation of Dense Optical Flow with Application to Video Object Segmentation ÖGAI Journal 24/1 11 Colour Segmentation-based Computation of Dense Optical Flow with Application to Video Object Segmentation Michael Bleyer, Margrit Gelautz, Christoph Rhemann Vienna University of Technology

More information

Curve Subdivision in SE(2)

Curve Subdivision in SE(2) Curve Subdivision in SE(2) Jan Hakenberg, ETH Zürich 2018-07-26 Figure: A point in the special Euclidean group SE(2) consists of a position in the plane and a heading. The figure shows two rounds of cubic

More information

COMP3421. Global Lighting Part 2: Radiosity

COMP3421. Global Lighting Part 2: Radiosity COMP3421 Global Lighting Part 2: Radiosity Recap: Global Lighting The lighting equation we looked at earlier only handled direct lighting from sources: We added an ambient fudge term to account for all

More information