Object Detection with Superquadrics Presenter: Ilya Afanasyev Facoltà di Ingegneria Trento, 2010 14/12/2010 1/20.

Slides:



Advertisements
Similar presentations
Robotica e Sensor Fusion per i Sistemi Meccatronici
Advertisements

Robotica e Sensor Fusion per i Sistemi Meccatronici
Recognize that to solve certain equations, number systems need to be extended from whole numbers to integers, from integers to rational numbers,
QR Code Recognition Based On Image Processing
Principal Component Analysis Based on L1-Norm Maximization Nojun Kwak IEEE Transactions on Pattern Analysis and Machine Intelligence, 2008.
Introduction to Computer Vision 3D Vision Lecture 4 Calibration CSc80000 Section 2 Spring 2005 Professor Zhigang Zhu, Rm 4439
Developable Surface Fitting to Point Clouds Martin Peternell Computer Aided Geometric Design 21(2004) Reporter: Xingwang Zhang June 19, 2005.
Methods of solving problems in electrostatics Section 3.
Kinematics & Grasping Need to know: Representing mechanism geometry Standard configurations Degrees of freedom Grippers and graspability conditions Goal.
Photo-realistic Rendering and Global Illumination in Computer Graphics Spring 2012 Material Representation K. H. Ko School of Mechatronics Gwangju Institute.
Computer vision: models, learning and inference
3D M otion D etermination U sing µ IMU A nd V isual T racking 14 May 2010 Centre for Micro and Nano Systems The Chinese University of Hong Kong Supervised.
Intelligent Systems Lab. Extrinsic Self Calibration of a Camera and a 3D Laser Range Finder from Natural Scenes Davide Scaramuzza, Ahad Harati, and Roland.
Uncertainty Representation. Gaussian Distribution variance Standard deviation.
Superquadric Recovery in Range Images via Region Growing influenced by Boundary Information Master-Thesis Christian Cea Bastidas.
Registration of two scanned range images using k-d tree accelerated ICP algorithm By Xiaodong Yan Dec
Segmentation into Planar Patches for Recovery of Unmodeled Objects Kok-Lim Low COMP Computer Vision 4/26/2000.
Vision, Video And Virtual Reality 3D Vision Lecture 13 Calibration CSC 59866CD Fall 2004 Zhigang Zhu, NAC 8/203A
Mean Shift A Robust Approach to Feature Space Analysis Kalyan Sunkavalli 04/29/2008 ES251R.
A synthetic camera model to test calibration procedures A four step procedure (last slide) based on an initial position (LookAt) and 13 parameters: ( 
2D/3D Geometric Transformations CS485/685 Computer Vision Dr. George Bebis.
Curves Locus of a point moving with one degree of freedom
Lecture 11: Structure from motion CS6670: Computer Vision Noah Snavely.
The Pinhole Camera Model
Object recognition under varying illumination. Lighting changes objects appearance.
Camera parameters Extrinisic parameters define location and orientation of camera reference frame with respect to world frame Intrinsic parameters define.
כמה מהתעשייה? מבנה הקורס השתנה Computer vision.
Reprojection of 3D points of Superquadrics Curvature caught by Kinect IR-depth sensor to CCD of RGB camera Mariolino De Cecco, Nicolo Biasi, Ilya Afanasyev.
Slide 1 of 40 Automation, Robotics and Mechatronics Lab, SUNY at Buffalo Shape Recovery from Medical Image Data Using Extended Superquadrics Talib Bhabhrawala.
Absolute error. absolute function absolute value.
Dual Evolution for Geometric Reconstruction Huaiping Yang (FSP Project S09202) Johannes Kepler University of Linz 1 st FSP-Meeting in Graz, Nov ,
Computer Graphics: Programming, Problem Solving, and Visual Communication Steve Cunningham California State University Stanislaus and Grinnell College.
Mathematics for Computer Graphics (Appendix A) Won-Ki Jeong.
Chemistry 330 The Mathematics Behind Quantum Mechanics.
Technology and Historical Overview. Introduction to 3d Computer Graphics  3D computer graphics is the science, study, and method of projecting a mathematical.
October 14, 2014Computer Vision Lecture 11: Image Segmentation I 1Contours How should we represent contours? A good contour representation should meet.
Collision handling: detection and response
Intro to 3D Models Angel Angel: Interactive Computer Graphics5E © Addison-Wesley
A Method for Registration of 3D Surfaces ICP Algorithm
Math / Physics 101 GAM 376 Robin Burke Fall 2006.
Generalized Hough Transform
Lecture 08 Detecting Shape Using Hough Transform Lecture 08 Detecting Shape Using Hough Transform Mata kuliah: T Computer Vision Tahun: 2010.
Background Mathematics Aaron Bloomfield CS 445: Introduction to Graphics Fall 2006.
A Flexible New Technique for Camera Calibration Zhengyou Zhang Sung Huh CSPS 643 Individual Presentation 1 February 25,
3D Computer Vision and Video Computing 3D Vision Topic 2 of Part II Calibration CSc I6716 Fall 2009 Zhigang Zhu, City College of New York
CSE 185 Introduction to Computer Vision Feature Matching.
October 16, 2014Computer Vision Lecture 12: Image Segmentation II 1 Hough Transform The Hough transform is a very general technique for feature detection.
3D Computer Vision and Video Computing 3D Vision Topic 2 of Part II Calibration CSc I6716 Spring2013 Zhigang Zhu, City College of New York
Parametric Surfaces and their Area Part I
3D Sensing 3D Shape from X Perspective Geometry Camera Model Camera Calibration General Stereo Triangulation 3D Reconstruction.
Basic Theory (for curve 01). 1.1 Points and Vectors  Real life methods for constructing curves and surfaces often start with points and vectors, which.
3D Computer Vision and Video Computing 3D Vision Topic 3 of Part II Calibration CSc I6716 Spring 2008 Zhigang Zhu, City College of New York
1 Teaching Innovation - Entrepreneurial - Global The Centre for Technology enabled Teaching & Learning, N Y S S, India DTEL DTEL (Department for Technology.
Motion Segmentation CAGD&CG Seminar Wanqiang Shen
Hough Transform CS 691 E Spring Outline Hough transform Homography Reading: FP Chapter 15.1 (text) Some slides from Lazebnik.
COORDINATE SYSTEMS & TRANSFORMATION
Fundamentals of Data Analysis Lecture 11 Methods of parametric estimation.
3D HUMAN BODY POSE ESTIMATION BY SUPERQUADRICS 26/02/2012 1/21 Ilya Afanasyev, Massimo Lunardelli, Nicolo' Biasi, Luca Baglivo, Mattia Tavernini, Francesco.
An SVM Learning Approach to Robotic Grasping Raphael Pelossof December Advanced Machine Learning.
3D Single Image Scene Reconstruction For Video Surveillance Systems
University of Ioannina
3D Graphics Rendering PPT By Ricardo Veguilla.
Principal Component Analysis (PCA)
Fitting Curve Models to Edges
Outline Peter N. Belhumeur, Joao P. Hespanha, and David J. Kriegman, “Eigenfaces vs. Fisherfaces: Recognition Using Class Specific Linear Projection,”
Dongwook Kim, Beomjun Kim, Taeyoung Chung, and Kyongsu Yi
Automatic cylinder detection using Hough Transform.
Object Detection with Superquadrics
Introduction to Artificial Intelligence Lecture 22: Computer Vision II
Presentation transcript:

Object Detection with Superquadrics Presenter: Ilya Afanasyev Facoltà di Ingegneria Trento, /12/2010 1/20

Superquadrics (SQ) A superquadric surface can be defined by the 3D vector : - are parameters of shape squareness; - define the superquadric sizes. 14/12/20102/20 Superquadrics are a family of parametric shapes, which are used as primitives for shape representation and play the role of prototypical parts and can be further deformed and glued together into realistic looking models.

Vector r(η,ω) sweeps out a closed surface in space when η,ω change in the given intervals: η ω x y z r(η,ω) Creation of SQ surface η,ω – independent parameters (latitude and longitude angles) of vector r(η,ω) expressed in spherical coordinates. a1a1 a2a2 a3a3 14/12/20103/20

Parametric equation of SQ The parametric equation of a superquadric surface: - are parameters of shape squareness; - define the superquadric sizes. 14/12/20104/20 Superquadrics can model a large set of standard building blocks, like spheres, cylinders, parallelepipeds and shapes in between. Modeling capabilities of SQ can be enhanced by tapering, bending and making cavities.

x’ x y z 2 nd way for creation of SQ surface x,y – independent parameters (Cartesian coordinates of SQ), which are used to obtain z coordinate. y’ z’ Use the parametric equation in Cartesian coordinates, considering f(x,y,z) = 1 -a 1 ≤ x ≤ a 1 -a 2 ≤ y ≤ a 2 z=NaN y=a 2 x=a 1 14/12/20105/20

Warning: complex numbers in SQ equation 1. If ε 1 or ε 2 < 1 and cos or sin of angles ω or η < 0, then vector r(η,ω) has complex values. To escape them, it should be used absolute values of the vector components, taking into account the sign of sin or cos for further SQ surface reconstruction. 2. Analogically if x or y 1, the function f(x,y,z) willl have the complex values of z. To overcome it, use the f(x,y,z) in power of exponent ε 1. f(x,y,z) ε1 = 1 6/2014/12/2010

ε 1 = 0.1 ε 1 = 1 ε 1 = 2 ε 2 = 0.1 ε 2 = 1 ε 2 = 2 Samples of SQ, depending on ε 1 or ε 2 14/12/20107/20 [1]

Reconstruction of complex object [2] Applications with SQ 14/12/20108 Reconstruction of complex object [4] Reconstruction of complex object [3]

Reconstruction of multiple objects [5] Applications with SQ 14/12/20109 Reconstruction of multiple objects [5]

What are we trying to achieve? To build the robust algorithm of recovery of pallets surface by Superquadrics with the real 3D data from a rangefinder (possibly, without segmentation). 14/12/ Develop a robust algorithm of recovery of simulated SQ objects (excluding noise). 2.Test it with 3D real data. 3.Build a superquadric model of the europallet. 4.Take 3D data of pallet from rangefinder. 5.Fit the superquadric model of the europallet to real 3D data by the algorithm.

Idea of work with SQ simulation 1. Create SQ with known parameters a 1, a 2, a 3, ε 1, ε Get representation of SQ in points (like “Cloud of points”). 3. Rotate and translate SQ according to “viewpoint of interest”. 4. Get the silhouette of SQ (visible part of SQ from viewpoint). 5. Add the noise to SQ silhouette to obtain more realistic model. 6. Detect the orientation and location of SQ, using the known model of SQ and detection algorithm (based on solving a system of nonlinear equation). Azimuth, Elevation, Distance ??? /2014/12/2010

Rotation and translation of SQ Elevation Azimuth x y z zWzW xWxW yWyW T – transformation matrix. n – amounts of points in SQ surface. SQ – coordinates of points of SQ surface. P w – coordinates of points of rotated SQ surface. x W, y W, z W – world system of coordinates (with center in viewpoint). Warning: Standard MATLAB function: viewmtx is useless for rotation and translation purposes. 12/2014/12/2010

Rotation and translation of SQ Elevation Azimuth x y z zWzW xWxW yWyW TSQPwPw pxpx az,el,px,py,pz,x,y,z – are given; x W,y W,z W – should be found. 1314/12/2010

Getting the silhouette of SQ 1) Calculate x’,y’,z’ (R’,Az’,El’) coordinates of every point of line laid between center of world coordinate system and point of interest from rotated SQ. ywyw xwxw p2p2 p’ p” Az’ El’ zwzw p’ – lays outside SQ; p’’ – lays inside SQ. (x’; y’) p1p1 14/2014/12/2010

Getting the silhouette of SQ 2) To recover the real x,y,z coordinates of every point of rotated SQ (rotate and translate it back): T – transformation matrix; SQ – non-rotated point of SQ silhouette; P’ w – rotated point of SQ silhouette. Elevation Azimuth x y z zWzW xWxW yWyW Silhouette (visible part of SQ from viewpoint) P(x;y;z) 1514/12/2010

Getting the silhouette of SQ - the point is on the surface of SQ. - the point is outside SQ. - the point is inside SQ. 3) Calculation of inside-outside function f(x,y,z) ε1. 4) If all points of a line from the “SQ point of interest” to the world coordinate center lay outside – this point of SQ is visible, otherwise – invisible. x y z zWzW xWxW yWyW Silhouette p1p1 p2p2 p 1 – is the point of SQ silhouette; p 2 – is not. 16/2014/12/2010

Adding the noise to SQ silhouette 2) Also the algorithm should process the SQ surface points, which cab be slightly deviated from real positions of SQ surface. We can create the deviation of mean values of z-coordinate for points of SQ surface according to law of normal distribution for small standard deviation (for example, σ=0.5). To build the robust algorithm of recovery of SQ surface it is necessary to take into account the noise, which can exist in the real data (from a rangefinder). 1) The algorithm should be robust to outlets. To simulate the outlets, we can add to SQ silhouette some points obtained by uniform distribution. silhouette Noised silhouette outliers 17/2014/12/2010

Algorithm of Recovery of SQ surface from silhouette To detect az,el from silhouette of rotated SQ, the given points should rotated and translated back to SQ coordinate system. T – transformation matrix. SQ – non-rotated points of SQ silhouette. P’ w – rotated points of SQ silhouette. Elevation Azimuth x y z zWzW xWxW yWyW Silhouette P(x’;y’;z’) T SQ P’ w ε 1, ε 2,a 1 -a 3, p x,p y,p z,x W,y W,z W – are given; Az,El,ω,η – should be found. 1814/12/2010

Algorithm of Recovery of SQ surface from silhouette To solve this system of nonlinear equations, when silhouette consists n ≥10 points, use MATLAB fsolve function (from Optimization toolbox) with Levenverg-Marquardt method. Future improvement of the algorithm deal with adding the new variable: p x, using the parameter dist for initial estimates x 0. x y z zWzW xWxW yWyW Silhouette T SQ P’ w p y =0; p z =0. ε 1, ε 2,a 1 -a 3,p y,p z,x W,y W,z W – are given; Az,El,ω,η, p x – should be found. pxpx dist Variables 1914/12/2010

Links 1. Solina F. and Bajcsy R. Recovery of parametric models from range images: The case for superquadrics with global deformations. // IEEE Transactions on Pattern Analysis and Machine Intelligence PAMI-12(2): , Chella A. and Pirrone R. A Neural Architecture for Segmentation and Modeling of Range Data. // 10 pages. 3. Leonardis A., Jaklic A., and Solina F. Superquadrics for Segmenting and Modeling Range Data. // IEEE Transactions On Pattern Analysis And Machine Intelligence, vol. 19, no. 11, Bhabhrawala T., Krovi V., Mendel F. and Govindaraju V. Extended Superquadrics. // Technical Report. New York. 93 pages. 5. Jaklic Ales, Leonardis Ales, Solina Franc. Segmentation and Recovery of Superquadrics. // Computational imaging and vision 20, Kluwer, Dordrecht, /12/201020/20 Grazie per attenzione!!