Image Processing. The Module. Lab Sessions and Courseworks. Prerequisites. Reference Book. Text Book Image Processing

Transcription

1 Processing Pengwei Hao Topic 1: Introduction ECS605U / ECS776P School of EECS Queen Mary University of London The Module Lectures: Mondays, 9-11am, ArtsOne 1.28 Pengwei Hao (p.hao@qmul.ac.uk) Lab sessions: Wednesday, 11am-1pm, ITL 2F, no lab in week 1. Mr Zhaoyang Xu, acw267, zhaoyang.xu@qmul.ac.uk Miss Xindi Zhang, eex291, xindi.zhang@qmul.ac.uk Miss Bingqing Guo, acw355, b.guo@qmul.ac.uk Lab Sessions and Courseworks Labs are compulsory Week 2: Lab1, complete GUI with template code Week 3-9: Lab2-8 programming for image processing Week 10: coursework completion The deadline to complete the coursework is the end of Week 10. Week 11&12: coursework assessment (assessed during the last 2 lab sessions.) Prerequisites Java programming experience Mathematical background: calculus, complex arithmetic, statistics, linear algebra Basic understanding of signal processing: function, signal and spectrum Text Book Processing Digital Processing Global Edition, 4th Edition by Rafael C. Gonzalez, and Richard E. Woods 1024 pages, Pearson, 2017 ISBN13: ISBN10: Reference Book Digital Processing using MATLAB by Rafael C. Gonzalez, Richard E. Woods, Steven L. Eddins 827 pages, Gatesmark Publishing 2nd edition (2009) ISBN-10: , ISBN-13:

2 Reference Journals IEEE Trans. Processing IEEE Trans. Signal Processing IEEE Trans. Medical Imaging Computer Vision and Understanding (CVIU) Graphical Modeling and Processing Computer Vision, Graphics and Processing (CVGIP) Reference Proceedings British Machine Vision Conference (BMVC) International Conference on Processing (ICIP) IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR) IEEE International Conference on Computer Vision (ICCV) Websites the module The course: Lecture notes : Labs: Coursework: Past exam papers: available on QM+ Text book: Assessment Coursework 20% Assessed during the last 2 lab sessions Examination (2h30m) 80% Processing Processing : Process images by means of digital computers : photograph/picture scanned/produced Digital image : An electronic photograph made up of a set of picture elements, "pixels" Each pixel is at a spatial position and has a tonal value (intensity, gray level) represented digitally in binary code (zeros and ones). First The oldest surviving camera photograph, by Nicéphore Niépce, 1826 or 1827, requires an exposure of at least 8 hours. 2

3 First of Person First photograph including a person, by Daguerre, 1838 First Colour James Clerk Maxwell presents the first durable colour photograph in 1861, uses three separate BW photographs taken with red, green, and violet-blue filters. First Colour Print A 1877 photographic colour print on paper by Louis Ducos du Hauron, superimposing cyan, red and yellow components. First Moving Pictures Muybridge used high-speed photography to make the first animated image sequences photographed in realtime ( ) First Easy-to-Use Camera 1888 The Kodak n 1 box camera, the first easy-to-use camera First Digital Used in newspaper industry Sent between London and New York Bartlane cable picture transmission system was introduced in a week => 3 hours Reproduced on a telegraph printer with typefaces to simulate a halftone pattern 3

4 Photographic Reproduction Better tonal quality (5 gray levels) Higher resolution Photographic Reproduction Better tonal quality (15 gray levels) Higher resolution First Scanned The first picture scanned, stored, and recreated in digital pixels by the SEAC scanner in SEAC: Standards Eastern Automatic Computer. Computers Involved in Creation From early 1960s US space missions First Pixel Array 1967 First MOS 10 by 10 active pixel array shown by Noble 4

5 First Colour Filter Array 1975 Bryce Bayer of Kodak develops the Bayer filter mosaic pattern for CCD colour image sensors. Medical Imaging From late 1960s and early 1970s Computerized tomography (CT) Using X-rays Bone scan Gamma-Ray Imaging Positron emission tomography (PET) Cygnus Loop Gamma radiation from a reactor valve Chest X-ray Aortic angiogram Head CT Circuit boards Cygnus Loop X-Ray Imaging Ultraviolet Band Imaging Normal corn Smut corn Cygnus Loop Light Microscopy s Anticancer agent, 250X Cholesterol, 40X Microprocessor, 60X Nickel oxide thin film, 600X Audio CD surface, 1750X Organic superconductor, 450X 5

6 Landsat Satellite 7 Bands: 1. Visible blue, 2. visible green, 3. visible red, 4. near infrared, 5. middle infrared, 6. thermal infrared, 7. middle infrared Meteorological Satellite Visible + Infrared Multispectral - Hyperspectral image - 80 bands (highly correlated) - 512x346 - Road, water, rice paddy Manufactural s Circuit board controller Packaged pills Bottles Bubbles in product Cereal Intraocular implant s in Visual Spectrum Spaceborne Radar images Fingerprint Paper money Automated license number reading 6

7 Radio Band Imaging (MRI) Electromagnetic Spectrum Magnetic Resonance Imaging Electromagnetic Spectrum Ultrasound Imaging Crab Pulsar Scanning Electron Microscope Hair with dandruff Holography 1947 Dennis Gabor invents holography 7

8 Commercial Digital Camera Kodak DC40 and the Apple QuickTake 100 become the first digital cameras marketed for consumers. 3D Camera 2009 FujiFilm launches world's first digital 3D camera with 3D printing capabilities Light-Field Camera Computer Generated s 2011 Lytro releases the first pocket-sized consumer light-field camera, capable of refocusing images after being taken. Computer Generated s Processing System Cloud Network displays Computer Mass storage Hardcopy Specialized image processing hardware processing software sensors Problem domain 8

9 Outputs of these processes generally are image attributes Digital Processing What You Will Learn Outputs of these processes generally are images CHA PTER 6 CHA PTER 7 CHA PTER 8 CHA PTER 9 Wavelets and Color image Compression and Morphological other multiresolution image processing watermarking processing transforms processing CHA PTER 5 CHA PTERS 10 & 11 Segmentation restoration CHA PTERS 3 & 4 CHA PTER 12 Knowledge base Feature filtering and extraction enhancement Level (brightness) operations Algebraic and logical operations Geometric transformations Filtering (both spatial and frequency-based) Sampling Restoration/Reconstruction Compression Problem domain CHA PTER 2 acquisition CHA PTER 13 pattern classification Applications Multimedia (just look at the web) Editing and Manipulation (Photoshop) Medical Imaging (CT, MRI) Compression (PNG, JPEG) Document Processing (OCR) Libraries (restoration/cleanup, storage, retrieval) Many More Relation to Other Fields Processing: Transform an image into another image, often to make the image better to give information Computer Vision: Find the model of an image Computer Graphics: Create an image of a model Processing Computer Vision Computer Graphics Model 9