Implications In any collocated camera+projector setup, there is a special illumination pattern that appears undistorted to the camera, for arbitrary scene.

Slides:



Advertisements
Similar presentations
Image Registration  Mapping of Evolution. Registration Goals Assume the correspondences are known Find such f() and g() such that the images are best.
Advertisements

Planar Orientation from Blur Gradients in a Single Image Scott McCloskey Honeywell Labs Golden Valley, MN, USA Michael Langer McGill University Montreal,
COS 429 PS2: Reconstructing a Simpler World
Chapter 5 Pictorial Sketching.
Seeing 3D from 2D Images. How to make a 2D image appear as 3D! ► Output and input is typically 2D Images ► Yet we want to show a 3D world! ► How can we.
Multimedia Specification Design and Production 2012 / Semester 1 / week 6 Lecturer: Dr. Nikos Gazepidis
S TRUCTURED L Prasanna RangarajanDr. Marc P Christensen Vikrant R BhaktaDr. Panos Papamichalis.
Three Dimensional Viewing
Muhammad Hasan Danish Khan University of Vaasa, Finland.
Shared Graphics Skills Cameras and Clipping Planes
Freehand Sketching.
Engineering Graphics Stephen W. Crown Ph.D.
Sharon Hornstein, PhD Optical Engineering Conf. February 26 th,
Camera Models A camera is a mapping between the 3D world and a 2D image The principal camera of interest is central projection.
Image Formation1 Projection Geometry Radiometry (Image Brightness) - to be discussed later in SFS.
Uncalibrated Epipolar - Calibration
Passive Object Tracking from Stereo Vision Michael H. Rosenthal May 1, 2000.
Lecture 5 Monday, 29 June ENGINEERING GRAPHICS 1E7 Lecture 5: Isometric Projections.
Today: Calibration What are the camera parameters?
© 2009, TSI Incorporated Stereoscopic Particle Image Velocimetry.
Visualization and Graphics Introduction
雅虎邮箱地址 : PW:zjuopt. Chapter 2 System Evaluation.
The Digital Image.
ENGI 4559 Signal Processing for Software Engineers Dr. Richard Khoury Fall 2009.
A critical review of the Slanted Edge method for MTF measurement of color cameras and suggested enhancements Prasanna Rangarajan Indranil Sinharoy Dr.
C OMPUTER V Prasanna Rangarajan04/09/10 Dr. Panos Papamichalis.
Basics of Rendering Pipeline Based Rendering –Objects in the scene are rendered in a sequence of steps that form the Rendering Pipeline. Ray-Tracing –A.
CAP4730: Computational Structures in Computer Graphics 3D Concepts.
776 Computer Vision Jan-Michael Frahm Fall Camera.
Rujchai Ung-arunyawee Department of Computer Engineering Khon Kaen University.
Geometric Models & Camera Calibration
Lecture 2b Readings: Kandell Schwartz et al Ch 27 Wolfe et al Chs 3 and 4.
Course 9 Texture. Definition: Texture is repeating patterns of local variations in image intensity, which is too fine to be distinguished. Texture evokes.
How natural scenes might shape neural machinery for computing shape from texture? Qiaochu Li (Blaine) Advisor: Tai Sing Lee.
V part 1 Obtained from a Guildford County workshop-Summer, 2014.
Discrete Fourier Transform in 2D – Chapter 14. Discrete Fourier Transform – 1D Forward Inverse M is the length (number of discrete samples)
CSE 185 Introduction to Computer Vision Stereo. Taken at the same time or sequential in time stereo vision structure from motion optical flow Multiple.
Microgrid Modulation Study High Resolution Sinusoidal ImageHigh Resolution Image Image FFT Simulated Microgrid Image (Fully Polarized) Fully Polarized.
Optical Holography Martin Janda, Ivo Hanák Introduction Wave Optics Principles Optical holograms Optical Holography Martin Janda, Ivo Hanák Introduction.
1. 1. What do you notice in this image?. 2. What do you think this is?
Feature Matching. Feature Space Outlier Rejection.
ELEC 106 FUNDAMENTALS OF ELECTRICAL ENGINEERING Engineering Drawing.
12/24/2015 A.Aruna/Assistant professor/IT/SNSCE 1.
Explanatory notes added to a drawing.
Unit 5 Shap Description (Orthographic Projection) نظرية الأسقاط.
Navigating in 3D MAX CTEC V part 1. Viewing Objects and/or Scenes Depending upon the software program, the image on the monitor could be a Perspective.
Fixed-Center Pan-Tilt Projector and Its Calibration Methods Ikuhisa Mitsugami Norimichi Ukita Masatsugu Kidode Graduate School of Information Science Nara.
1Ellen L. Walker 3D Vision Why? The world is 3D Not all useful information is readily available in 2D Why so hard? “Inverse problem”: one image = many.
A Photograph of two papers
Answer Key Graph # 1 - The graph is misleading because the lower part of the vertical axis is missing, the differences in prices are exaggerated. Graph.
1 PTT 105/3: Engineering Graphics. TAXONOMY OF PLANAR GEOMETRIC PROJECTIONS PTT 105/3: Engineering Graphics 2 Planar Geometric Projections Parallel Multiview.
Image features and properties. Image content representation The simplest representation of an image pattern is to list image pixels, one after the other.
ENGINEERING DRAWING VISUALIZATION. Axonometric & Oblique Projection.
The Frequency Domain Digital Image Processing – Chapter 8.
Date of download: 6/6/2016 Copyright © 2016 SPIE. All rights reserved. System overview for direct optical fringe writing in PRP. The direct fringe writing.
GE1021 Engineering Drawing and Graphics
Date of download: 6/27/2016 Copyright © 2016 SPIE. All rights reserved. Typical experimental setup. The FPA camera is focused onto the plane of the projection.
Quality of Images.
AAFS 2004 Dallas Zeno Geradts
Date of download: 10/20/2017 Copyright © ASME. All rights reserved.
PERSPECTIVE PROJECTION…...
From: The effect of perceived surface orientation on perceived surface albedo in binocularly viewed scenes Journal of Vision. 2003;3(8):2. doi: /3.8.2.
Overview Pin-hole model From 3D to 2D Camera projection
Three Dimensional Viewing
3D Graphics.
Picture Analysis Terms
Course 6 Stereo.
Picture Analysis Terms
Guilford County SciVis V part 1
SAVI: Synthetic apertures for long-range, subdiffraction-limited visible imaging using Fourier ptychography by Jason Holloway, Yicheng Wu, Manoj K. Sharma,
Presentation transcript:

Implications In any collocated camera+projector setup, there is a special illumination pattern that appears undistorted to the camera, for arbitrary scene geometry. The pattern may or may not be periodic !! Existing structured light scanners can now be used to super- resolve in addition to recovering depth information ( with suitable modifications ) Preliminary Experimental Results on surpassing the diffraction limit for crossed optical axes Prasanna Rangarajan Indranil Sinharoy Vikrant R Bhakta

Picture of Experimental Setup Top View Side View Special thanks to Indranil for the pictures Camera Projector Angle between optical axes of camera and projector is ≈ 30 degrees The projector center-of-perspective is designed to be in the XZ plane of the camera coordinate system, which is centered at the camera center-of-perspective  horizontally collocated setup Projector Optical Axis Camera Optical AxisBaseline

Picture of Experimental Setup Top View Side View Camera Projector The target is parallel to the camera image plane during calibration In all other cases, the target is parallel to the projector. Consequently, the camera perceives the calibration target as being tilted, and having depth variation Special thanks to Indranil for the pictures

Details of Camera, Projector Setup

More details…

EXPERIMENT - 1 USAF target positioned in front of calibration target such that it is parallel to the projector image plane Illuminate scene with pre-warped vertical pattern of spatial frequency 90 cycles/mm in camera image plane Pre-warped illumination pattern 1400 x 1050 ( appears aliased due to MATLAB display issues ) Zoomed-in view of pre-warped illumination pattern 201x 201 patch Notice the perspective warp in illumination pattern

super-resolving a planar USAF target that is oriented at an angle of 30 degrees w.r.t camera Notice the perspective distortion in the image of the USAF target camera optical cutoff ≈ 180 lp/mm in the image plane size of image is 601 x 601

super-resolving a planar USAF target that is oriented at an angle of 30 degrees w.r.t camera Notice the periodic pattern super-imposed on the USAF target. Although the target appears to experience perspective distortion, the pattern appears un-distorted to the camera ( evident in the Fourier domain ) carrier frequency ≈ 90 lp/mm in the camera image plane

super-resolving a planar USAF target that is oriented at an angle of 30 degrees w.r.t camera Notice that modulation is able to resolve vertical edges in the oriented USAF target NOTE: The carrier pattern might appear slightly aliased due to MATLAB’s display issues The fringe patterns on the target appear oriented due to the tilt in the target

super-resolving a planar USAF target that is oriented at an angle of 30 degrees w.r.t camera Notice the peak in the fourier spectrum of the cosine/sine modulated image of the scene, at ≈ 90 lp/mm in the camera image plane

super-resolving a planar USAF target that is oriented at an angle of 30 degrees w.r.t camera Notice the peak in the fourier spectrum of the complex sinusoidal modulated image of the scene, at ≈ 90 lp/mm in the camera image plane

super-resolving a planar USAF target that is oriented at an angle of 30 degrees w.r.t camera Notice that modulation in the vertical direction is able to extend the bandwidth of the imaging system in the vertical direction

super-resolving a planar USAF target that is oriented at an angle of 30 degrees w.r.t camera size of raw image is 601 x 601

super-resolving a planar USAF target that is oriented at an angle of 30 degrees w.r.t camera size of super resolved image is 841 x 841 Notice enhanced resolution in vertical elements of group -1 ( zoom-in for better view ) There is a faint artifact that is regular and appears to be perspectively distorted

EXPERIMENT - 2 Barcode target positioned in front of calibration target such that it is parallel to the projector image plane Illuminate scene with pre-warped vertical pattern of spatial frequency 90 cycles/mm in camera image plane Pre-warped illumination pattern 1400 x 1050 ( appears aliased due to MATLAB display issues ) Zoomed-in view of pre-warped illumination pattern 201x 201 patch Notice the perspective warp in illumination pattern

super-resolving a planar barcode target that is oriented at an angle of 30 degrees w.r.t camera Notice the perspective distortion in the image of the barcode target camera optical cutoff ≈ 180 lp/mm in the image plane size of image is 601 x 601 the fine spatial in the barcodes is missing ( zoom-in for better view )

super-resolving a planar USAF target that is oriented at an angle of 30 degrees w.r.t camera Notice the periodic pattern super-imposed on the USAF target. Although the target appears to experience perspective distortion, the pattern appears un-distorted to the camera ( evident in the Fourier domain ) carrier frequency ≈ 90 lp/mm in the camera image plane

super-resolving a planar USAF target that is oriented at an angle of 30 degrees w.r.t camera Notice that modulation is able to resolve vertical edges in the 1D & 2D barcodes NOTE: The carrier pattern might appear slightly aliased due to MATLAB’s display issues The fringe patterns in the barcodes appear oriented due to the tilt in the target

super-resolving a planar barcode target that is oriented at an angle of 30 degrees w.r.t camera Notice the peak in the fourier spectrum of the cosine/sine modulated image of the scene, at ≈ 90 lp/mm in the camera image plane

super-resolving a planar barcode target that is oriented at an angle of 30 degrees w.r.t camera Notice the peak in the fourier spectrum of the complex sinusoidal modulated image of the scene, at ≈ 90 lp/mm in the camera image plane

super-resolving a planar barcode target that is oriented at an angle of 30 degrees w.r.t camera Notice that modulation in the vertical direction is able to extend the bandwidth of the imaging system in the vertical direction

super-resolving a planar barcode target that is oriented at an angle of 30 degrees w.r.t camera size of raw image is 601 x 601

super-resolving a planar barcode target that is oriented at an angle of 30 degrees w.r.t camera size of super resolved image is 841 x 841 Notice enhanced resolution in the vertical barcode & the 2D bar code The vertical barcode is fully resolved There is a faint artifact that is regular and appears to be perspectively distorted