Developing a gigapixel web-service for mobile devices
| Thesis type |
|
|---|---|
| Student | Markus Soworka |
| Status | Finished |
| Submitted in | 2012 |
| Presentation on | 11. Nov 2012 00:00 |
| Presentation room | Seminarraum I5 |
| Add presentation to calendar |
vCal
iCal
|
| Supervisor(s) | |
| Advisor(s) |
A gigapixel image is a digital image bitmap composed of one billion (10^9) pixels , more than 1000 times the information captured by a 1 megapixel digital camera. Gigapixel technology is being deployed in a wide range of applications ranging from remote sensing to the field of arts. Current technology for creating such very high-resolution images involves making mosaics of a large number of high-resolution digital photographs. The Bachelor Thesis is to explore in detail current available applications and their underlying image-mosaic technologies in order to develop a concept for making the consumption of these images available on an ANDROID mobile device.
Conventional Panoramic digital photography relies on creating panoramic images by stitching relatively few multiple images together. In the recent years, however, advances in software, hardware, and digital camera capabilities now allow possible resolutions into the range of several gigapixels per single image. The resulting gigapixel image is one that has a tremendous range of detail contained, allowing unsurpassed viewing and exploration. This ultimately allows for large prints that show no softening or degradation from size, while allowing intricate detail to be revealed at close inspection via a variety of media.
The equipment required to shoot spherical panoramic imagery has typically been the use of a simple panoramic nodal head, which is manually turned to dozens of preset shooting positions, covering every possible angle within a sphere of the camera. Gigapixel photography, however, requires many hundreds or thousands of shots to be taken and typically demands the use of an automated motion control head which has the precision to execute the large amount of overlapping images. The resulting shots are then stitched together in dedicated image stitching software, often taking several hours of uninterrupted processing for a single image to be assembled.
While the underlying image processing algorithms for large images are already employed in a wide range of web based application (i.e. Smart Zooming of satellite imagery known by Google Maps or others) the creation of large images in non-expert domains is actually fostered by low-cost equipment (Gigapan) and easy-to-use software packages (MS HD View). The digital camera system to be used in this bachelor thesis is available at the chair.
Based on an overview of technologies this thesis will concentrate on HD View and Gigapan to evaluate the underlying technologies for image-stichting and upload of processed images and explore the possibilities for making the uploaded images available on mobile devices. Suitable metadata standards from different domains shall be explored in order to evaluate aspects of of portability and interoperability of these technologies across devices and application domains.
Some further reading:
http://research.microsoft.com/en-us/um/redmond/groups/ivm/HDView/
Prerequisites
Interest in digital photography and development skill on mobile devices is expected.
Keynote Talk at GDN 2017 