Open-Source Aerial Imagery platform
Revision as of 16:18, 8 March 2011 by Omapwiki
- We plan on developing a open-source and easily hackable aerial imagery platform, targeting autonomous aerial vehicles. Recently there has been a huge explosion in growth in the Open Source UAV community, as well as a similar growth in the commercial viability of these platforms for geographic/ecological surveys, crisis management, emergency response, agriculture monitoring, filming and photography, and many more. Our platform will be relatively low-cost (especially compared to commercial solutions) and highlight the features of the OMAP and PandaBoard to excel in this application. (Onboard video camera, hardware-accelerated jpeg compression/decompression, high-speed ethernet, wireless connectivity, low-power, and more!)
- We will support controlling external digital cameras using a variety of protocols over the USB 2.0 hosts, the onboard camera support, and both at once. For storage we'll support efficient and secure wireless transfer to a ground station (based on the implementation we used to win our competition last year) while simultaneously creating a local backup on a USB harddrive or on the SD card. Finally we will use GIS (Geospatial Information Systems) software running that will allow users to select which areas of the ground they want imagery of, and have the platform trim out undesired (or overlapping) portions of images to maximize wireless throughput.
- Edit: I think this will mesh well with the rest of the Unmanned/Autonomous Aerial Vehicle projects already listed.
- Time frame
- 4-5 months, we plan on having this platform mission ready for the 2011 competition (see below).
- Background & work by project submitter/s
- Lead software developer and Imagery system designer for the NCSU Aerial Robotics Team. Last year (2010) NCSU Aerial Robotics got top marks and First Place at the international AUVSI Unmanned Aerial Systems competition. The competition is to develop the best mission-capable autonomous aerial imagery platform for quickly and efficiently finding targets and points of interest. We believe the pandaboard is perfectly suited to this application, as it is smaller, lighter, and more efficient than previous platforms (from us and other teams). Our team has experience with ARM development (Linux and FreeRTOS as well as plain EABI) in other aspects of our competition, including designing and building target boards, developing and debugging software, and live in-mission testing/evaluation.
- Wiki/URL Links
- Main group site: http://art1.mae.ncsu.edu/
- Wiki page about the imagery system (and the pandaboard proposal): http://art1.mae.ncsu.edu/twiki/bin/view/Main/ImagerySystems.
- Todo list: http://art1.mae.ncsu.edu/twiki/bin/view/Main/ToDo
- Contact information
- Black Market Winner -- 2/28/11 - 3/4/11 event