Slide 1Slide 2Slide 3Slide 4Slide 5Slide 6Slide 7VisionComputer vision class, fast-forwardImages and image formationCameras, lenses, and sensorsRadiometrySlide 13Low-level visionImage filteringNon-linear filtering, and applicationsModels of textureShape from shadingBayesian framework for visionSlide 20Slide 21Edge detection and optical flowSlide 23Learning and visionSlide 25Support vector machines and boostingSlide 27ApplicationsComputer vision applications as ocean-going vesselsGame: DecathleteOptical-flow-based Decathlete figure motion analysisDecathlete 100m hurdlesDecathlete javelin throwSlide 34videoNintendo Game Boy CameraSlide 37The course, in broad categorieshttp://www.ai.mit.edu/courses/6.801/Vision•What does it mean, to see? “to know what is where by looking”.•How to discover from images what is present in the world, where things are, what actions are taking place.from Marr, 1982Computer vision class, fast-forwardImages and image formationCameras, lenses, and sensorsFrom Computer Vision, Forsyth and Ponce, Prentice-Hall, 2002.•Pinhole cameras•Lenses•Projection models•Geometric camera parametersRadiometryWolfgang Luchthttp://geography.bu.edu/brdf/brdfexpl.htmlFrom Foundations of Vision, by Brian Wandell, Sinauer Assoc., 1995ColorLow-level visionImage filtering•Review of linear systems, convolution•Bandpass filter-based image representations•Probabilistic models for imagesImageOriented, multi-scale representationviewer television displayIEEE Computer Graphics and Applications, 18, no. 3, 1998templateimage Normalized correlationNon-linear filtering, and applicationsModels of textureA Parametric Texture Model based on Joint Statistics of Complex Wavelet Coefficients J. Portilla and E. Simoncelli, International Journal of Computer Vision 40(1): 49-71, October 2000.© Kluwer Academic Publishers. Parametric modelNon-parametric modelA. Efros and W. T Freeman, Image quilting for texture synthesis and transfer, SIGGRAPH 2001Shape from shadingShape from shading, Horn and Brooks, MIT Press, 1986Bayesian framework for visionGahan Wilson’s Still Weird, Forge, 1994Bayesian framework for visionhttp://www.cs.dartmouth.edu/whites/old_man.htmlCoincidental appearance of face profile in rock?Bayesian framework for visionhttp://bensguide.gpo.gov/3-5/symbols/print/mountrushmore.htmlCoincidental appearance of faces in rock?Edge detection and optical flowM. Heath, S. Sarkar, T. Sanocki, and K.W. Bowyer, "A Robust Visual Method for Assessing the Relative Performance of Edge-Detection Algorithms" IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 19, No. 12, December 1997, pp. 1338-1359.Edge detection and optical flowin "Statistical Theories of the Brain", edited by R. Rao, B. Olshausen, and M. Lewicki, MIT Press, 2001ImagesTrue optical flowEstimated optical flowLearning and visionStatistical classifiers–MIT Media Lab face localization results.–Applications: database search, human machine interaction, video conferencing.Support vector machines and boostingwww.support-vector.net/nello.htmlLarge-margin classifierSupport vector machines and boostingwww.support-vector.net/nello.html“The kernel trick”Applications•Computer vision for computer gamesthisapplicationComputer vision applicationsas ocean-going vesselsGame: DecathleteOptical-flow-based Decathlete figure motion analysisDecathlete 100m hurdlesDecathlete javelin throwDecathlete javelin throwvideoNintendo Game Boy Camera•Several million sold (most of any digital camera). Imaging chip is Mitsubishi Electric’s “Artificial Retina” CMOS detector.videoThe course, in broad categories•Images and image formation•Geometry•Low-level vision•High-level vision•Implementations and