Lilong Shi and Brian Funt School of Computing Science, Simon Fraser University, Canada.

Slides:



Advertisements
Similar presentations
Electromagnetic Waves and Light
Advertisements

Intrinsic Images by Entropy Minimization (Midway Presentation, by Yingda Chen) Graham D. Finlayson, Mark S. Drew and Cheng Lu, ECCV, Prague, 2004.
Computer Vision Radiometry. Bahadir K. Gunturk2 Radiometry Radiometry is the part of image formation concerned with the relation among the amounts of.
Scene illumination and surface albedo recovery via L1-norm total variation minimization Hong-Ming Chen Advised by: John Wright.
Illumination Estimation via Thin Plate Spline Weihua Xiong ( OmniVision Technology,USA ) Lilong Shi, Brian Funt ( Simon Fraser University, Canada) ( Simon.
Robust statistical method for background extraction in image segmentation Doug Keen March 29, 2001.
COLOUR YEAR 11 - UNIT ONE PHYSICS
Simon Fraser University Computational Vision Lab Lilong Shi, Brian Funt and Tim Lee.
Hyperspectral Imaging Seminar Dmitry Yudovsky and Laurent Pilon Retrieving Skin Properties From In Vivo Spectral Reflectance Measurements.
Acquiring the Reflectance Field of a Human Face Paul Debevec, Tim Hawkins, Chris Tchou, Haarm-Pieter Duiker, Westley Sarokin, Mark Sagar Haarm-Pieter Duiker,
Color spaces CIE - RGB space. HSV - space. CIE - XYZ space.
Light, Surface and Feature in Color Images Lilong Shi Postdoc at Caltech Computational Vision Lab, Simon Fraser University.
Physics-based Illuminant Color Estimation as an Image Semantics Clue Christian Riess Elli Angelopoulou Pattern Recognition Lab (Computer Science 5) University.
A novel concept for measuring seawater inherent optical properties in and out of the water Alina Gainusa Bogdan and Emmanuel Boss School of Marine Sciences,
1 Practical Scene Illuminant Estimation via Flash/No-Flash Pairs Cheng Lu and Mark S. Drew Simon Fraser University {clu,
BMVC 2009 Specularity and Shadow Interpolation via Robust Polynomial Texture Maps Mark S. Drew 1, Nasim Hajari 1, Yacov Hel-Or 2 & Tom Malzbender 3 1 School.
ECCV 2002 Removing Shadows From Images G. D. Finlayson 1, S.D. Hordley 1 & M.S. Drew 2 1 School of Information Systems, University of East Anglia, UK 2.
ICCV 2003 Colour Workshop 1 Recovery of Chromaticity Image Free from Shadows via Illumination Invariance Mark S. Drew 1, Graham D. Finlayson 2, & Steven.
1 Invariant Image Improvement by sRGB Colour Space Sharpening 1 Graham D. Finlayson, 2 Mark S. Drew, and 2 Cheng Lu 1 School of Information Systems, University.
1 A Markov Random Field Framework for Finding Shadows in a Single Colour Image Cheng Lu and Mark S. Drew School of Computing Science, Simon Fraser University,
Modelling, calibration and rendition of colour logarithmic CMOS image sensors Dileepan Joseph and Steve Collins Department of Engineering Science University.
School of Computer Science Simon Fraser University November 2009 Sharpening from Shadows: Sensor Transforms for Removing Shadows using a Single Image Mark.
Quaternion Colour Texture
Segmentation Divide the image into segments. Each segment:
Image Segmentation. Introduction The purpose of image segmentation is to partition an image into meaningful regions with respect to a particular application.
Illumination Estimation via Non- Negative Matrix Factorization By Lilong Shi, Brian Funt, Weihua Xiong, ( Simon Fraser University, Canada) Sung-Su Kim,
Mark S. Drew and Amin Yazdani Salekdeh School of Computing Science,
Global Illumination May 7, Global Effects translucent surface shadow multiple reflection.
Introduction of the intrinsic image. Intrinsic Images The method of Finlayson & Hordley ( 2001 ) Two assumptions 1. the camera ’ s sensors are sufficiently.
Color Fidelity in Multimedia H. J. Trussell Dept. of Electrical and Computer Engineering North Carolina State University Raleigh, NC
LESSON 4 METO 621. The extinction law Consider a small element of an absorbing medium, ds, within the total medium s.
Describing Visual Air Quality Is A Complex Issue Depending On: characteristics of observer optical characteristics of target illumination of scene optical.
Lilong Shi, Brian Funt, and Ghassan Hamarneh School of Computing Science, Simon Fraser University.
Statistical Color Models (SCM) Kyungnam Kim. Contents Introduction Trivariate Gaussian model Chromaticity models –Fixed planar chromaticity models –Zhu.
Computer Graphics Inf4/MSc Computer Graphics Lecture Notes #16 Image-Based Lighting.
Dye Sublimation Color Management
Matthias Wimmer, Bernd Radig, Michael Beetz Chair for Image Understanding Computer Science Technische Universität München Adaptive.
Optical flow (motion vector) computation Course: Computer Graphics and Image Processing Semester:Fall 2002 Presenter:Nilesh Ghubade
The IOCCG Atmospheric Correction Working Group Status Report The Eighth IOCCG Committee Meeting Department of Animal Biology and Genetics University.
Optical Models Jian Huang, CS 594, Spring 2002 This set of slides are modified from slides used by Prof. Torsten Moeller, at Simon Fraser University, BC,
Computer Science 631 Lecture 7: Colorspace, local operations
Ch. 5 - Basic Definitions Specific intensity/mean intensity Flux
MV-4920 Physical Modeling Remote Sensing Basics Mapping VR/Simulation Scientific Visualization/GIS Smart Weapons Physical Nomenclature Atmospherics Illumination.
Mestrado em Ciência de Computadores Mestrado Integrado em Engenharia de Redes e Sistemas Informáticos VC 15/16 – TP4 Colour and Noise Miguel Tavares Coimbra.
Michal Tepper Under the supervision of Prof. Israel Gannot.
February 2004 Chuck DiMarzio, Northeastern University a-1 ECEU692 Subsurface Imaging Course Notes Part 2: Imaging with Light (3): Strong Scattering.
CHAPTER 5 Atmospheric Influence and Radiometric Correction PRE-PROCESSING A. Dermanis.
Using computer vision for analysis of plant growth condition: what to consider? Hans Jørgen Andersen Computer Vision and Media Technology laboratory Aalborg.
Integumentary System (Skin). What Can You Observe About Skin? Look at the skin on your arms and hands. What does your skin do for your body?
Computer Graphics: Illumination
Date of download: 7/5/2016 Copyright © 2016 SPIE. All rights reserved. (a) Simulated measurement system scanning a surface element at position P(r→). (b)
Heechul Han and Kwanghoon Sohn
3D Rendering 2016, Fall.
IMAGE PROCESSING COLOR IMAGE PROCESSING
Miguel Tavares Coimbra
Computer Graphics Chapter 9 Rendering.
Intrinsic images and shape refinement
The Skin Functions of the skin: Protects the body from injury,
Utilizing Spectrophotometry in Life Science
HyperSpectral Skin Imaging Tianchen Shi, Prof. Charles A. DiMarzio
Illumination Model How to compute color to represent a scene
Colour Theories.
High Dynamic Range Images
Integumentary System Functions:
Chapter 14: Bones, Muscles, and Skin Section 4: The Skin
Volume 104, Issue 1, Pages (January 2013)
CS5500 Computer Graphics May 29, 2006
Vysakh Vasudevan*, N. Sujatha
Speaker: YI-JIA HUANG Date: 2011/12/08 Authors: C. N
Specularity, the Zeta-image, and Information-Theoretic Illuminant
Presentation transcript:

Lilong Shi and Brian Funt School of Computing Science, Simon Fraser University, Canada

Normalize skin tones of human faces Eliminate the effects of illumination Preserve skin colour Allow variations of melanin concentration Normalize skin tones of human faces Eliminate the effects of illumination Preserve skin colour Allow variations of melanin concentration

Two-layered Skin Model [1] Epidermis Layer: Melanin Absorbance Dermis Layer: Hemoglobin Absorbance A layer has properties of an optical filter Two-layered Skin Model [1] Epidermis Layer: Melanin Absorbance Dermis Layer: Hemoglobin Absorbance A layer has properties of an optical filter

Reflection spectrum of skin [1]: where,  ’s are pigment densities of melanin & hemoglobin  ’s are absorbance of melanin and hemoglobin, l ’s are mean path lengths of photons,  is other scattering loss and absorbance. Reflection spectrum of skin [1]: where,  ’s are pigment densities of melanin & hemoglobin  ’s are absorbance of melanin and hemoglobin, l ’s are mean path lengths of photons,  is other scattering loss and absorbance. 4

Wien’s blackbody radiation models where, I is power of radiation, c 1 and c 2 are constants, T is blackbody temperature, Wien’s blackbody radiation models where, I is power of radiation, c 1 and c 2 are constants, T is blackbody temperature, 5

Proposed to combine Skin & Illum. Models Assume narrowband sensors used: In log space: Then, let Π represent camera RGB: Proposed to combine Skin & Illum. Models Assume narrowband sensors used: In log space: Then, let Π represent camera RGB: 6

where,  m &  h are melanin & hemoglobin bases,  is a blackbody radiator basis, b = log(I),  = 1/T, c is a constant vector.  m &  h span all possible skin colours where,  m &  h are melanin & hemoglobin bases,  is a blackbody radiator basis, b = log(I),  = 1/T, c is a constant vector.  m &  h span all possible skin colours 7 σmσm 1 ω c log G log B log R σhσh

Our simplified Skin-Illumination Model For varying illumination colour temperature; For varying skin melanin concentration;  m and  span the chromaticity space of arbitrary skin under different illuminations. Given a skin pixel, melanin concentration can be recovered, so is true skin colour. Our simplified Skin-Illumination Model For varying illumination colour temperature; For varying skin melanin concentration;  m and  span the chromaticity space of arbitrary skin under different illuminations. Given a skin pixel, melanin concentration can be recovered, so is true skin colour. 8

Results based on UOPB[2] database (#94) (a) a series of 16 face images under different camera calibration and illumination conditions. (faces segmented from the background) (b) the same images with corrected skin tones based on our model. Results based on UOPB[2] database (#94) (a) a series of 16 face images under different camera calibration and illumination conditions. (faces segmented from the background) (b) the same images with corrected skin tones based on our model. 9

Results based on UOPB[2] database (#111) (a) a series of 16 face images under different camera calibration and illumination conditions. (faces segmented from the background) (b) the same images with corrected skin tones based on our model. Results based on UOPB[2] database (#111) (a) a series of 16 face images under different camera calibration and illumination conditions. (faces segmented from the background) (b) the same images with corrected skin tones based on our model. 10

Based on physical models Estimate skin melanin concentration Skin colour varies along melanin axis Shift colour along illum. axis Simple and computationally inexpensive References: [1] Shimada, M., Y. Yamada, M. Itoh and T. Yatagai Melanin and blood concentration in a human skin model studied by multiple regression analysis: assessment by Monte Carlo simulation. Phys. Med. Biol. 46(9): [2] Marszalec, E., B. Martinkauppi, M. Soriano, M. Pietikäinen A physics-based face database for color research. Journal of Electronic Imaging 9(1): Based on physical models Estimate skin melanin concentration Skin colour varies along melanin axis Shift colour along illum. axis Simple and computationally inexpensive References: [1] Shimada, M., Y. Yamada, M. Itoh and T. Yatagai Melanin and blood concentration in a human skin model studied by multiple regression analysis: assessment by Monte Carlo simulation. Phys. Med. Biol. 46(9): [2] Marszalec, E., B. Martinkauppi, M. Soriano, M. Pietikäinen A physics-based face database for color research. Journal of Electronic Imaging 9(1):