Frequency-Domain Range Data Registration for 3-D Space Modeling in Robotic Applications By Phillip Curtis.

Slides:



Advertisements
Similar presentations
Real-Time Template Tracking
Advertisements

Image Registration  Mapping of Evolution. Registration Goals Assume the correspondences are known Find such f() and g() such that the images are best.
Arc-length computation and arc-length parameterization
The fundamental matrix F
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.
Kinematic Synthesis of Robotic Manipulators from Task Descriptions June 2003 By: Tarek Sobh, Daniel Toundykov.
Evaluating Color Descriptors for Object and Scene Recognition Koen E.A. van de Sande, Student Member, IEEE, Theo Gevers, Member, IEEE, and Cees G.M. Snoek,
Fast Algorithms For Hierarchical Range Histogram Constructions
November 12, 2013Computer Vision Lecture 12: Texture 1Signature Another popular method of representing shape is called the signature. In order to compute.
1.  Texturing is a core process for modeling surface details in computer graphics applications › Texture mapping › Surface texture synthesis › Procedural.
Integrating a Short Range Laser Probe with a 6-DOF Vertical Robot Arm and a Rotary Table Theodor Borangiu Anamaria Dogar
An Integrated Robotic Laser Range Sensing System for Automatic Mapping of Wide Workspaces P. Curtis P. Payeur University of Ottawa.
Uncertainty Representation. Gaussian Distribution variance Standard deviation.
Instructor: Mircea Nicolescu Lecture 13 CS 485 / 685 Computer Vision.
Semi-automatic Range to Range Registration: A Feature-based Method Chao Chen & Ioannis Stamos Computer Science Department Graduate Center, Hunter College.
Registration of two scanned range images using k-d tree accelerated ICP algorithm By Xiaodong Yan Dec
CENG 789 – Digital Geometry Processing 06- Rigid-Body Alignment Asst. Prof. Yusuf Sahillioğlu Computer Eng. Dept,, Turkey.
Iterative closest point algorithms
Reflective Symmetry Detection in 3 Dimensions
Direct Methods for Visual Scene Reconstruction Paper by Richard Szeliski & Sing Bing Kang Presented by Kristin Branson November 7, 2002.
Fractal Image Compression
Image Enhancement in the Frequency Domain Part I Image Enhancement in the Frequency Domain Part I Dr. Samir H. Abdul-Jauwad Electrical Engineering Department.
Preprocessing ROI Image Geometry
Fitting a Model to Data Reading: 15.1,
Multiple View Geometry Marc Pollefeys University of North Carolina at Chapel Hill Modified by Philippos Mordohai.
3D full object reconstruction from kinect Yoni Choukroun Elie Semmel Advisor: Yonathan Afflalo.
Digital Imaging and Remote Sensing Laboratory Automatic Tie-Point and Wire-frame Generation From Oblique Aerial Imagery Seth Weith-Glushko Advisor: Carl.
CSci 6971: Image Registration Lecture 5: Feature-Base Regisration January 27, 2004 Prof. Chuck Stewart, RPI Dr. Luis Ibanez, Kitware Prof. Chuck Stewart,
כמה מהתעשייה? מבנה הקורס השתנה Computer vision.
1 REAL-TIME IMAGE PROCESSING APPROACH TO MEASURE TRAFFIC QUEUE PARAMETERS. M. Fathy and M.Y. Siyal Conference 1995: Image Processing And Its Applications.
Topic 7 - Fourier Transforms DIGITAL IMAGE PROCESSING Course 3624 Department of Physics and Astronomy Professor Bob Warwick.
Introduction to Image Processing Grass Sky Tree ? ? Review.
Constraints-based Motion Planning for an Automatic, Flexible Laser Scanning Robotized Platform Th. Borangiu, A. Dogar, A. Dumitrache University Politehnica.
Geodesic Minimal Paths Vida Movahedi Elder Lab, January 2010.
Parallelism and Robotics: The Perfect Marriage By R.Theron,F.J.Blanco,B.Curto,V.Moreno and F.J.Garcia University of Salamanca,Spain Rejitha Anand CMPS.
Automatic Registration of Color Images to 3D Geometry Computer Graphics International 2009 Yunzhen Li and Kok-Lim Low School of Computing National University.
Shape Matching for Model Alignment 3D Scan Matching and Registration, Part I ICCV 2005 Short Course Michael Kazhdan Johns Hopkins University.
MESA LAB Multi-view image stitching Guimei Zhang MESA LAB MESA (Mechatronics, Embedded Systems and Automation) LAB School of Engineering, University of.
A Method for Registration of 3D Surfaces ICP Algorithm
Realtime 3D model construction with Microsoft Kinect and an NVIDIA Kepler laptop GPU Paul Caheny MSc in HPC 2011/2012 Project Preparation Presentation.
Global Parametrization of Range Image Sets Nico Pietroni, Marco Tarini, Olga Sorkine, Denis Zorin.
A Frequency-Domain Approach to Registration Estimation in 3-D Space Phillip Curtis Pierre Payeur Vision, Imaging, Video and Autonomous Systems Research.
Plenoptic Modeling: An Image-Based Rendering System Leonard McMillan & Gary Bishop SIGGRAPH 1995 presented by Dave Edwards 10/12/2000.
1 Complex Images k’k’ k”k” k0k0 -k0-k0 branch cut   k 0 pole C1C1 C0C0 from the Sommerfeld identity, the complex exponentials must be a function.
A Passive Approach to Sensor Network Localization Rahul Biswas and Sebastian Thrun International Conference on Intelligent Robots and Systems 2004 Presented.
Computer Vision Lecture #10 Hossam Abdelmunim 1 & Aly A. Farag 2 1 Computer & Systems Engineering Department, Ain Shams University, Cairo, Egypt 2 Electerical.
Bahadir K. Gunturk1 Phase Correlation Bahadir K. Gunturk2 Phase Correlation Take cross correlation Take inverse Fourier transform  Location of the impulse.
A Flexible New Technique for Camera Calibration Zhengyou Zhang Sung Huh CSPS 643 Individual Presentation 1 February 25,
Using simplified meshes for crude registration of two partially overlapping range images Mercedes R.G.Márquez Wu Shin-Ting State University of Matogrosso.
CSE 185 Introduction to Computer Vision Feature Matching.
MultiModality Registration Using Hilbert-Schmidt Estimators By: Srinivas Peddi Computer Integrated Surgery II April 6 th, 2001.
CENG 789 – Digital Geometry Processing 07- Rigid-Body Alignment Asst. Prof. Yusuf Sahillioğlu Computer Eng. Dept,, Turkey.
BLOCK BASED MOTION ESTIMATION. Road Map Block Based Motion Estimation Algorithms. Procedure Of 3-Step Search Algorithm. 4-Step Search Algorithm. N-Step.
Registration and Alignment Speaker: Liuyu
Camera calibration from multiple view of a 2D object, using a global non linear minimization method Computer Engineering YOO GWI HYEON.
CSCI 631 – Foundations of Computer Vision March 15, 2016 Ashwini Imran Image Stitching.
If you have a transaction processing system, John Meisenbacher
Lesson Plan: Drafting and Design J6-2. What is 3D solid modeling? How do 3D solid modeling programs work?
11/25/03 3D Model Acquisition by Tracking 2D Wireframes Presenter: Jing Han Shiau M. Brown, T. Drummond and R. Cipolla Department of Engineering University.
The content of lecture This lecture will cover: Fourier Transform
CENG 789 – Digital Geometry Processing 08- Rigid-Body Alignment
Automatically Collect Ground Control Points from Online Aerial Maps
Approximate Models for Fast and Accurate Epipolar Geometry Estimation
Multi-modality image registration using mutual information based on gradient vector flow Yujun Guo May 1,2006.
Image Segmentation Techniques
Image Based Modeling and Rendering (PI: Malik)
CSE 185 Introduction to Computer Vision
--- Range Image Registration
Chapter 4 . Trajectory planning and Inverse kinematics
Presentation transcript:

Frequency-Domain Range Data Registration for 3-D Space Modeling in Robotic Applications By Phillip Curtis

What is Registration? Registration is the act of making an image of an object from one point of view, match an image of the same object from a different point of view.

What is Needed? A registration technique for robotics applications must be: –Quick, with a low computational burden –Flexible (precision adjusted to task) –Accurate –Scalable

Registration Prior Art Classical approaches –Three Point Problem Requires direct knowledge of point correspondents and 3-D spatial locations: P2=Q*P1, solve for Q –Iterative solutions Classic iterative closest point (ICP) algorithm by Besl and McKay [1]. –Most research in the field of range image registration is centred on modifications on the ICP approach [1] P.J. Besl, N.D. McKay, “A Method for Registration of 3-D Shapes”, IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 14, pp , Feb

ICP Besl and McKay’s ICP Algorithm –1 st : match points between images using the criterion of closest point –2 nd : determine the optimal registration for that match by first estimating the rotation, then the translation –3 rd : rotate the 1 st image by the estimation –4 th : Repeat the 1 st, 2 nd, and 3 rd steps until the error delta between iterations is small enough

ICP Advantages –Allows for arbitrary data sampling structures –Simple and precise Disadvantages –Tends toward local minima, unless a precise initial estimate is used –Slow due to its matching algorithm (N- squared)

Frequency Domain Registration Well known in 2-D registration Extended to 3-D by Lucchese et al. [2] –Takes advantage of the fact that the Fourier transform decouples the estimation of the rotational parameters from that of the translational parameters –Uses correlation and geometric projection techniques to extract rotational and translational parameters [2] L. Lucchese, G. Doretto, G.M. Cortelazzo, “A Frequency Domain Technique for Range Data Registration”, IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 24, No. 11, pp , Nov

Frequency Domain Registration Advantages –No initial estimate –No matching of features required –Avoid local minima solutions that are inherent in ICP Disadvantages –Luccese et al. require many transformations of the data (FFT and correlation histograms) to achieve results

Frequency Domain Benefits The frequency domain allows for a low computational burden, due to the availability of the Fast Fourier Transform The frequency domain techniques have the ability to scale well to an increase in dataset size Precision can be adjusted by changing the resolution.

Frequency Domain Registration Fourier Transform allows for the effective segregation of the rotation parameters from the translation parameters. Fourier Transform Magnitude Phase

Determining The Axis of Rotation All objects which rotate have an axis of rotation. By eliminating the effects of the translation parameters when two frequency domain magnitude images are subtracted, the axis of rotation can be found by determining the zero line through the frequency origin, since this line is the only constant between the two images.

Determining the Angle of Rotation The angle of rotation can be determined via a minimum sum of the difference of squares search of possible rotation values about the axis of rotation. Due to the Hermitian symmetry property of the Fourier transform, there are two possible rotation angles, separated by 180°.

Solution Selection A phase correlation between the first image, and the second image derotated by both solutions is used to determine the proper solution. The proper solution will yield a more impulse- like space function.

Estimation of Translation The location of the impulse of the phase correlation corresponds to the estimate of the translation parameters

What Needed to be Done Lucchese et al. provide a nice rigorous start to frequency domain registration, but to be practical for robotics applications the following must be improved –A more efficient method for the estimation of the axis of rotation –A more efficient and flexible method for the estimation of the angle of rotation –Handle images with multiple properties (colour, texture, etc), in addition to range

Descretisation of Data Apply continuously positioned data to a discrete grid for the FFT –Need to add ‘volume’ to each point, so that the energy is captured, prevent impulsive effects, and aliasing –Cubic voxels to prevent scaling factors –Minimize the data space

Determining the Axis of Rotation Minimize calculation time, while maintaining accuracy comparable to Lucchese et al. Solution was to develop the normalised percentage difference equation (eq. 18) to find the difference between F-D images Use a moving window search technique to find the axis

Determine the Angle of Rotation Lucchese et al. use a correlation histogram technique using the projections of rotated then re-transformed data to estimate the angle of rotation –Huge computational penalty This thesis uses a coarse to fine minimum of least squares iterative approach

Solution Selection Observations of Correct solution vs. complementary solution –Correct solution is more impulsive, and that impulse is higher than the average energy Uses peak energy / average energy measure along the projections of each dimension

Translation Estimate The solution with the highest ratio “wins” The location of the maximal peak in the winning solution is the estimate of the translation parameters

Experimental Setup Used both simulated data and real data acquired –Simulated data from Dr. P. Payeur’s simulator –Real Data from robotic range sensing system developed primarily for this thesis

Robotic Range Sensing System Combines CRS 6 degree of freedom serial robotic arm with a track containing an additional degree of freedom, plus a laser range line scanner, and a standard PC

Robotic Range Sensing System Used standard RS-232 serial communications protocol and Windows 2000 Win32 API to develop control and acquisition software. Inverse kinematics was modified from that developed by W.M. Keck virtual factory lab.

Robotic Range Sensing System Cut down on number of tools required to acquire data Saves data in many formats used by members in VIVA lab Repeatable and quick (64 lines from 3 different perspectives takes about half an hour) Expandable code base

Test Data Used both simulated data sets and real data sets –The simulated house frame was selected to evaluate the performance of the algorithm using objects with a high degree of symmetry –The real house frame data was selected to see how the algorithm performed under “real” data vs. simulated data.

Test Data –Simulated computer setup was used to evaluate performance with occlusions –Skewed planar surfaces were used to evaluate overlap performance

Some Results Histogram of rotation error of simulated house frame (top) and real house frame (bottom) data sets.

Some Results Selected registration point clouds of registered data sets (top simulated house frame, bottom is real house frame)

Some Results Histogram of rotation error (top) and selected registration point cloud (bottom) for simulated computer data sets

Some Results Execution times of the frequency domain registration algorithm developed in this thesis, compared to that of ICP

Results The implementation as described in this thesis is accurate, and flexible Have improved computational efficiency, compared to Lucchese et al. without observable loss of accuracy More scalable than ICP (execution time is faster and does not grow as rapidly as ICP) Have added in the capability for enhanced data sets containing colour in addition for range data (untested)

Conclusion Proposed, implemented, and tested an automatic registration estimation algorithm that: –does not require human intervention –does not require an initial estimate –is independent of the geometry of the object –provides the framework for the multi-dimensional, multi-property data registration –is scalable with regards to data set size, and desired precision –is more efficient than that of Lucchese et al. and of Besl and McKay.

Conclusion The following innovations were contributed to the area of frequency domain registration research –More computationally efficient difference equation for calculating the difference between frequency domain magnitude images –Moving window to determine axis of rotation –Coarse to fine approach to determine the angle of rotation –Development of a robotic range sensing system

Future Work Improve solution selection mechanism Investigate other transform domains Test with enhanced data sets containing multiple data attributes Develop the mathematical frame work for higher dimensional registration using the axis of rotation model

Resulting Research Publications P. Curtis, P. Payeur, “An Integrated Robotic Laser Range Sensing System for Automatic Mapping of Wide Workspaces”, Proceedings of the IEEE Canadian Conference on Electrical And Computer Engineering, Vol. 2, pp , May P. Curtis, C.S. Yang, P. Payeur, "An Integrated Robotic Multi-Modal Sensing System", Proceedings of the IEEE International Instrumentation and Measurement Technology Conference, pp , Ottawa, ON, C.S. Yang, P. Curtis, P. Payeur, "Calibration of an Integrated Robotic Multi-Modal Range Scanner", submitted to IEEE Transactions on Instrumentation and Measurement, P. Curtis, P. Payeur, “A Frequency Domain Approach to Registration in 3-D Space”, submitted to IEEE International Instrumentation and Measurement Technology Conference, Oct P. Curtis, P. Payeur, “Registering 3-D Data with Multiple Properties using a Frequency Domain Approach“, in preparation for IEEE Transactions on Pattern Analysis and Machine Intelligence.

Questions