IMAGE BASED VISUAL SERVOING

Slides:



Advertisements
Similar presentations
Visual Servo Control Tutorial Part 1: Basic Approaches Chayatat Ratanasawanya December 2, 2009 Ref: Article by Francois Chaumette & Seth Hutchinson.
Advertisements

QR Code Recognition Based On Image Processing
Change Detection C. Stauffer and W.E.L. Grimson, “Learning patterns of activity using real time tracking,” IEEE Trans. On PAMI, 22(8): , Aug 2000.
Ray tracing. New Concepts The recursive ray tracing algorithm Generating eye rays Non Real-time rendering.
Street Crossing Tracking from a moving platform Need to look left and right to find a safe time to cross Need to look ahead to drive to other side of road.
1. A given pattern p is sought in an image. The pattern may appear at any location in the image. The image may be subject to some tone changes. 2 pattern.
Automatic in vivo Microscopy Video Mining for Leukocytes * Chengcui Zhang, Wei-Bang Chen, Lin Yang, Xin Chen, John K. Johnstone.
Virtual Dart: An Augmented Reality Game on Mobile Device Supervisor: Professor Michael R. Lyu Prepared by: Lai Chung Sum Siu Ho Tung.
Machine vision is not a subset of: Computer Science Image Processing Pattern Recognition Artificial Intelligence (whatever this is!) However, tools and.
Probabilistic video stabilization using Kalman filtering and mosaicking.
Stereoscopic Light Stripe Scanning: Interference Rejection, Error Minimization and Calibration By: Geoffrey Taylor Lindsay Kleeman Presented by: Ali Agha.
7/24/031 Ben Blazey Industrial Vision Systems for the extruder.
Virtual Dart – An Augmented Reality Game on Mobile Device Supervised by Prof. Michael R. Lyu LYU0604Lai Chung Sum ( )Siu Ho Tung ( )
A Novel 2D To 3D Image Technique Based On Object- Oriented Conversion.
Xinqiao LiuRate constrained conditional replenishment1 Rate-Constrained Conditional Replenishment with Adaptive Change Detection Xinqiao Liu December 8,
An FPGA implementation of real-time QRS detection H.K.Chatterjee Dept. of ECE Camellia School of Engineering & Technology Kolkata India R.Gupta, J.N.Bera,
FEATURE EXTRACTION FOR JAVA CHARACTER RECOGNITION Rudy Adipranata, Liliana, Meiliana Indrawijaya, Gregorius Satia Budhi Informatics Department, Petra Christian.
Shane Tuohy.  In 2008, rear end collisions accounted for almost 25% of all injuries sustained in road traffic accidents on Irish roads [RSA Road Collision.
Shape Recognition and Pose Estimation for Mobile Augmented Reality Author : N. Hagbi, J. El-Sana, O. Bergig, and M. Billinghurst Date : Speaker.
CSSE463: Image Recognition Day 30 This week This week Today: motion vectors and tracking Today: motion vectors and tracking Friday: Project workday. First.
Landing a UAV on a Runway Using Image Registration Andrew Miller, Don Harper, Mubarak Shah University of Central Florida ICRA 2008.
3D SLAM for Omni-directional Camera
Scientific Writing Abstract Writing. Why ? Most important part of the paper Number of Readers ! Make people read your work. Sell your work. Make your.
Submitted by: Giorgio Tabarani, Christian Galinski Supervised by: Amir Geva CIS and ISL Laboratory, Technion.
BARCODE IDENTIFICATION BY USING WAVELET BASED ENERGY Soundararajan Ezekiel, Gary Greenwood, David Pazzaglia Computer Science Department Indiana University.
Acquiring 3D models of objects via a robotic stereo head David Virasinghe Department of Computer Science University of Adelaide Supervisors: Mike Brooks.
Stable Multi-Target Tracking in Real-Time Surveillance Video
CSC508 Convolution Operators. CSC508 Convolution Arguably the most fundamental operation of computer vision It’s a neighborhood operator –Similar to the.
Sejong Univ. Edge Detection Introduction Simple Edge Detectors First Order Derivative based Edge Detectors Compass Gradient based Edge Detectors Second.
Univ logo Research and Teaching using a Hydraulically-Actuated Nuclear Decommissioning Robot Craig West Supervisors: C. J. Taylor, S. Monk, A. Montazeri.
A Tutorial on using SIFT Presented by Jimmy Huff (Slightly modified by Josiah Yoder for Winter )
Team Members Ming-Chun Chang Lungisa Matshoba Steven Preston Supervisors Dr James Gain Dr Patrick Marais.
Visual Odometry David Nister, CVPR 2004
Implementation of Real Time Image Processing System with FPGA and DSP Presented by M V Ganeswara Rao Co- author Dr. P Rajesh Kumar Co- author Dr. A Mallikarjuna.
Zhaoxia Fu, Yan Han Measurement Volume 45, Issue 4, May 2012, Pages 650–655 Reporter: Jing-Siang, Chen.
Instantaneous Geo-location of Multiple Targets from Monocular Airborne Video.
Ehsan Nateghinia Hadi Moradi (University of Tehran, Tehran, Iran) Video-Based Multiple Vehicle Tracking at Intersections.
Vision-based Android Application for GPS Assistance in Tunnels
CMSC5711 Image processing and computer vision
Signal and Image Processing Lab
Presented by Jason Moore
Hiba Tariq School of Engineering
Adaptive Block Coding Order for Intra Prediction in HEVC
Paper – Stephen Se, David Lowe, Jim Little
Injong Rhee ICMCS’98 Presented by Wenyu Ren
Velocity Estimation from noisy Measurements
Multidisciplinary Engineering Senior Design Project P06441 See Through Fog Imaging Preliminary Design Review 05/19/06 Project Sponsor: Dr. Rao Team Members:
A language assistant system for smart glasses
3D Stereoscopic Image Analysis Ahmed Kamel, Aashish Agarwal
See Through Fog Imaging Project: P06441
Ali Ercan & Ulrich Barnhoefer
Image filtering Hybrid Images, Oliva et al.,
CMSC5711 Image processing and computer vision
Example: Applying EC to the TSP Problem
Cos 429: Face Detection (Part 2) Viola-Jones and AdaBoost Guest Instructor: Andras Ferencz (Your Regular Instructor: Fei-Fei Li) Thanks to Fei-Fei.
Vision Tracking System
Project P06441: See Through Fog Imaging
Image and Video Processing
Estimation of relative pose
An Infant Facial Expression Recognition System Based on Moment Feature Extraction C. Y. Fang, H. W. Lin, S. W. Chen Department of Computer Science and.
CSSE463: Image Recognition Day 30
Image filtering
Image filtering
Reduction of blocking artifacts in DCT-coded images
CSSE463: Image Recognition Day 30
CSSE463: Image Recognition Day 30
Fig 2: System in action with athlete
Scalable light field coding using weighted binary images
Nome Sobrenome. Time time time time time time..
Sign Language Recognition With Unsupervised Feature Learning
Presentation transcript:

IMAGE BASED VISUAL SERVOING Supervisor : Dr. Ming LIU Associate Supervisor: Dr. F. CRUSCA Seng Guan TAN INTRODUCTION Vision as a means of non-contact measurement has been quite popular due in fact that it is very close to the human visual sense. The application of vision as a feedback in a control loop has became a trend because it is a non-contact measurement method that remains useful in a non-controlled environment. Improvement in computing capability allowed a real-time control to be implemented. THE PUMA 560 Shown below is a typical control loop for an image based visual servoing system : Fig. 1 The PUMA 560 Fig. 2 Screen short of the QNX controller program Simulation done using the classic method of applying a normal PD controller taking in to account compensation for gravity and friction. (V,w)T = K (J+ (fd - f) ) V = (v,w)T J+ is a control jacobi based on the pseudo inverse of image jacobi (Ji). In an example the J+ is (JiJr)T From (published paper), The suggested controller becomes: EQUIPMENTS PUMA Controller OS - QNX Will be the OS for the control of the PUMA 560. - QNX Platform allow for real time processing / control. - The ACRA program in Fig. 2 is the product of previous Student and the source code can be easily modified with a different “controller”. CAMERA A high S/N CCD camera is to be mounted on the end-effector, the image data is processed by a independent PC that run in synch with the PUMA 560 system. Camera Model is TM - 6E The Video Capture Card used is DT3155. Image Feature Extraction There are several challenge in feature extraction. 1. Correlation of feature points between frames. 2. Processing time. We to try to finish the image control calculation before the next sampling interval Windowing method proposed. Estimation of window position: - Prediction of window position by implementation of image jacobi: f = f(old) + Ts x Ji (v,w)T (v,w)T from the controller. - A general fixed estimation based on previous feature coordinate. Assuming a slow feature velocity. Image Jacobi Ji is actually based on the current feature points . f Ji Note that for Ji (f1) we can actually stack the matrix as below: Note: In addition to basing on current feature, Ji can be based on (fd) : Ji( fd ) However this method can only work well if the initial error is very small. It is used to avoid reaching a local minima in the J matrix a b c Fig. 3 (a),(b) : location shown on top left corner (c): with cross section of line showing current feature point. Fig 3 Final Control System Images above have gone through background reduction, thresholding, ,median filter and the feature coordinated are shown on the top left corner. (This is a bit too much processing, however, due to interference the image data captured is too noisy to be used otherwise) Processing time (without display) was around ~ 0.4 s in a Pentium3 888Hz : (768 x 576 pixel) By implementing the windowing method: time is greatly reduced ~ 0.032s (camera acquisition 30fps rate ) The new timing include an additional edge detection process. Still the process should be improved, since the control algorithm will be taking its own processing time. (0.1s total) sampling time is currently being considered. Electrical and Computer Systems Engineering Postgraduate Student Research Forum 2001