Upgrade to Pro — share decks privately, control downloads, hide ads and more …

Gigapixel Imaging

Gigapixel Imaging

Presentation on the concept of Gigapixel Imaging, part of my college final year project

Mayank Khanna

March 10, 2011
Tweet

More Decks by Mayank Khanna

Other Decks in Technology

Transcript

  1.  In Digital Imaging, a Pixel is a single point

    in a raster image or the smallest addressable screen element in a display device.  It is the smallest unit of picture that can be represented or controlled.  Each pixel is a sample of an original image wherein more samples typically provide more accurate representations of the original. Gigapixel Imaging SCE, Dept. of ISE 2
  2.  It is an application of ultra-high resolution techniques to

    the field of large-format photography, that started in late 2000 under the impetus of Graham Flint.  It led to design and construction of a special camera and associated procedures.  They achieved a 4-Gigapixel resolution after scanning of 9’’x18’’ negative and enlarge it to 96’’x 192’’ four panel assembly. Gigapixel Imaging SCE, Dept. of ISE 5
  3.  Gigapan is a collaborative project between Carnegie Mellon University

    & NASA Ames Intelligent System Division’s Robotics group with support from Google.  It is the newest development of the Global Connection Project, which aims to bring distant communities & people together through images that have so much detail that they are themselves the object of exploration, discovery & wonder.  It’s goal is to facilitate the taking of large composite pictures, presenting data as a single image & providing efficient web storage, browsing & zooming of such images. Gigapixel Imaging SCE, Dept. of ISE 7
  4.  Gigapan consists of three technological developments:  A Robotic

    Camera mount for capturing very high-resolution images using a standard digital camera.  A Custom software for constructing those high-resolution Gigapixel Panoramas.  A Website for exploring, sharing & commenting on Gigapixel Panoramas & the details one can discover within them. Gigapixel Imaging SCE, Dept. of ISE 8
  5.  Commercially available Gigapan Robotic mounts turn most Digital cameras

    into extremely high-resolution Panoramic Imagers.  The main components are:  A Camera Mount bracket  A Tripod Mount plate  Two Stepper/ Servo motors  Motor Driver electronics  Two Gear boxes  Micro-Controller board and Software Gigapixel Imaging SCE, Dept. of ISE 9
  6.  The Software is one that stitches individual images to

    create a single image.  It takes overlapping digital images shot from a common view point and aligns them and blends them to form a High-Resolution Mosaic.  It uses a Stitch Algorithm to combine the numerous images based on their overlapping regions.  GigaPan uses the GigaPan Stitch software for this purpose. Gigapixel Imaging SCE, Dept. of ISE 10
  7.  Another Stitching Software available- Autopano  Created by the

    company- Kolor  It has been rated the no. 1 Stitching Software presently available in the market.  Difference between the GigaPan Stitch and the Kolor Autopano is of the Quality of the final image that is rendered and how well the images are blended together. Gigapixel Imaging SCE, Dept. of ISE 11
  8.  Gigapan.org allows users to upload, share and explore Brilliant

    Gigapixel Panoramas from around the Globe.  Allows users to comment and take snapshots of GigaPan Images. Gigapixel Imaging SCE, Dept. of ISE 12
  9.  Uploaded Gigapans reside on servers at Carnegie Mellon University,

    Pittsburgh, Pennsylvania, USA.  To achieve the smoothest Panning and Zooming, the core trick is that at any moment, one views a modest number of Pixels.  The server rather than sending the entire image in full detail to your computer over the network sends only the pixels that are needed for one’s current view.  This depends on what subrectangle one’s looking at (The Pan) and how magnified it is (The Zoom). Gigapixel Imaging SCE, Dept. of ISE 14
  10.  The Server stores an Image Pyramid, known as The

    Image Quadtree, in which multiple versions of the image have been precomputed and decide into “Tiles”.  As one pans and zooms around the image, the server sends new tiles as they come into view, at appropriate zoom level over the network.  Typical panning & zooming requires only a few tiles per second to be transmitted, so even on Low-Bandwidth networks, one can have a reasonably smooth experience. Gigapixel Imaging SCE, Dept. of ISE 15
  11.  Gigapixel Photographs are ideal for Tourism, Real-Estate, Architecture, Medical

    Imaging, Archiving and Documenting special events.  High Resolution Images create the impression of “Being There” by immersing the viewer within the scene.  The extreme level of detail captured in a Gigapixel Image allows artifacts to be accurately preserved for Future Generations.  It will help us meet our neighbours across the globe and learn about our planet itself. Gigapixel Imaging SCE, Dept. of ISE 17
  12.  The mean size of Gigapans, has increased from 140

    in the year 2007 to 940 megapixels in 2010.  The number of Gigapans from 10,000 in 2009 to around 40,000 in 2010.  As of September 2010, Gigapan.org has about 31,000 users, of whom 4880 have uploaded atleast one public Gigapan.  Majority of users register in order to add comments to a Gigapan or to create snapshots to others' Gigapans. Gigapixel Imaging SCE, Dept. of ISE 21
  13.  Gigapan is more than just a tool to produce

    highly detailed panoramic images.  It can be an innovative way to teach people to see, to hold a community conversation and to gather scientific data.  Gigapan offers exciting possibilities for a wide range of truly novel avenues for formal and informal education, for community engagement and for scientific discovery. Gigapixel Imaging SCE, Dept. of ISE 24
  14.  Websites:  http://en.wikipedia.org  http://www.gigapan.org  http://www.paris-26-gigapixels.com  http://www.cs.cmu.edu/~globalconn/gigapan.html

     http://www.gigapan.blogspot.com  E-Papers:  Gigapan Uses In Education - Jerry Schoen Stevenson  How Many Pixels- Paul Heckbert Miriam  How many ways do I use thee - Brian Yamamoto Gigapixel Imaging SCE, Dept. of ISE 25