|CCL HomeSoftware Community Operations||
REU Project: MapScryMapScry works in the larger context of TeamTrak and is one of a constellation of tools designed to manage the data produced by that system. Teamtrak is basically a distributed sensor network. It permits many users to share location information over an ad-hoc network and gather various sensory information. Analyzing its outputs allows us to delve into the potential applications of wireless cooperative mobile computing systems.
My Project was to produce a system for navigating the geolocated images produced by TeamTrak. In particular, it provides means of visualizing the many photographs gathered from various uses of theTeamTrak system in relation to a simple map, and tools for searching those images both visually and through sql syntax. Tools like mapscry will gain increasing importance as more and more images come accompanied with data about their location, time taken, facing, etc. Making a system that can effectively navigate such images gives an entirely new level of accessibility to photo collections without requiring lengthy manual inspection and renaming of the images.
MapScry currently provides a SQL savvy user with an enormously powerful tool for navigating the thousands of images produced by teamtrak. However, more work needs to be done in creating easy to use visual tools for users without SQL experties, or without the time to construct complex queries. Additionally, data mining and image recognition techniques might permit even more powerful tools for navigating and searching the images. Finally, scalability issues need to be addressed if we expect this system to operate on extremely large sets of data. The first priority of the system should be increasing the usability so that a non-technical user can effectively access the power of the underlying system. Once this has been achieved scalability and data mining improvements could be explored.