Unformatted text preview:

16.098 Digital and Computational Photography 6.882 Advanced Computational PhotographyBill FreemanFrédo DurandMIT - EECSToday's plan• Introduction of Computational Photography• Course facts• Camera advice• Syllabus• HistoryWhat is computational photography• Convergence of image processing, computer vision, computer graphics an dphotography• Digital photography:– Simply replaces traditional sensors and recording by digital technology– Involves only simple image processing• Computational photography– More elaborate image manipulation, more computation– New types of media (panorama, 3D, etc.)– Camera design that take computation into accountQuick demosComputational Photography @ MIT• Tone mapping• Defocus Matting• Motion magnification• SuperresolutionTone mapping• One of your assignments!BeforeAfterDefocus Matting• With Morgan McGuire, Wojciech Matusik, Hanspeter Pfister, John “Spike” Hughes• Data-rich: use 3 streams with different focus2Super-resolution• Get a sharp high resolution image from low resolution• Important principle: Learn from examplesOriginal70x70Our technique [Freeman et al]Bicubic spline AltamiraOur technique [Freeman et al]Motion magnification Today's plan• Introduction of Computational Photography• Course facts• Camera advice• Syllabus• HistoryAdministrivia• Staff– Bill Freeman [email protected]– Frédo Durand [email protected]–CeLiu [email protected][email protected]• Office hours (email for other time)– Bill Freeman: Tuesday 2:30-4pm, 32-D476 – Frédo Durand: Friday 2:30-4pm, 32-D426– Ce Liu: Wednesday 2:30-4pm, 32-D460• Prereq: 18.06 & 6.003– or equivalent level• Web page: http://groups.csail.mit.edu/graphics/classes/CompPhoto06/– Lecture notes will be postedGrading policy• 6.098 – Assignments 75% – Final project 25%• 6.882 – Assignment 70%• With additional questions compared to 6.098– Final project 22%– Paper review 8%• Read and write a review (Siggraph form) for a paper from the literature+ participation3Assignment• Every two week• written questions + programming• Camera?– Not required, but can help. Can be borrowed from us. • Matlab– First office hour (Tuesday), Ce will give an intro– Or see Xiaoxu Ma's slides:http://courses.csail.mit.edu/6.869/handouts/6869%20Matlab%20Tutorial.ppt• Turn in code and results• Final project– Proposal due with PSet 5– Individual or teams of 2Textbook• No textbook required• Lots of resources on the net• Siggraph course notes– http://www.merl.com/people/raskar/photo/• Will post references with lecturesQuestions? Introductions• Who are you?• What do you know about photography?• Why you want to take this class?What do you think you will learn? What the class is not about• Little about art, photographers• Little about EE (sensors, A/D, etc)• Not much about optics – but some cool stuff such as wavefront coding• Not how to use Photoshop– But how its coolest tools work• Not much about 3D imaging• Not too much fundamentals of signal processing• Not much computational imaging, no tomography, no radar, no microscopy– See Berthold Horn’s class!• Not much computer vision, computer graphics– We avoided overlap with 6.837 and 6.801/6.8664What the class is about• Software aspects of computational photography– but a bit of hardware as well, lens technology, new camera designs• Basic tools– Linear & non-linear image processing, color• Emphasis on applications– High-dynamic range photography, photomontage, panoramas, foreground extraction, inpainting, morphing• Emphasis on recent research resultsSkills you will acquire• Implementation of basic tools– Color demosaicing– Multiscale blending– Matting– Bilateral filter– Gradient reconstruction– Panorama stitching– Markov Random Field– Optical flow• General approaches to computational photography• Important problems ion computational photography• By the end of the class, you should be able to make contributions to the state of the art– Your final project could lead to a publicationNon-photo motivation• It's about any kind of data !– Speech, motion, geometry, etc. – Example: •Music• Motion graphs• Poisson mesh editing• BTF shopQuestions?Today's plan• Introduction of Computational Photography• Course facts• Camera advice• Syllabus• HistorySyllabus• Image formation • Color and color perception • Demosaicing• Image processing and wavelets • Applications of wavelets; pyramid texture synthesis • Matting • High Dynamic Range Imaging • Bilateral filtering and HDR display • Gradient image manipulation • Taking great pictures • Markov Random Fields • Non-parametric image synthesis, inpainting, analogies • Tampering detection and higher-order statistics • Panoramic imaging • Image and video registration • Spatial warping operations • Motion analysis • Temporal sequence re-rendering • Active flash methods • Lens technology • Depth and defocus • Non-photorealistic rendering • Future cameras • Plenoptic function and light fields • Student project presentations5Syllabus• Image formation • Color and color perception • DemosaicingSyllabus• Image processing and wavelets • Applications of wavelets; pyramid texture synthesis Syllabus• High Dynamic Range Imaging • Bilateral filtering and HDR display • Matting Syllabus• Gradient image manipulation Syllabus• Taking great pictures Art Wolfe Ansel AdamsSyllabus• Markov Random Fields• Non-parametric image synthesis, inpainting, analogies6Syllabus• Tampering detection and higher-order statistics Syllabus• Panoramic imaging • Image and video registration • Spatial warping operations Syllabus• Motion analysis Syllabus• Temporal sequence re rendering Syllabus• Active flash methods • Lens technology • Depth and defocus No-flashFlashour resultSyllabus• Non-photorealistic rendering7Syllabus• Future cameras • Plenoptic function and light fields Questions?Today's plan• Introduction of Computational Photography• Course facts• Camera advice• Syllabus• HistoryEquipment• Do get an SLR, compacts are way too limited• Don't worry about brand• Don't worry about the body, get the cheapest one• Worry about lenses– Zooms are convenient but quality can be a problem• avoid the basic zoom, but the one above is usually great•


View Full Document

MIT 6 098 - Lecture Notes

Download Lecture Notes
Our administrator received your request to download this document. We will send you the file to your email shortly.
Loading Unlocking...
Login

Join to view Lecture Notes and access 3M+ class-specific study document.

or
We will never post anything without your permission.
Don't have an account?
Sign Up

Join to view Lecture Notes 2 2 and access 3M+ class-specific study document.

or

By creating an account you agree to our Privacy Policy and Terms Of Use

Already a member?