Visualization of large data set on Small screens

Abstract

Representing large, valuable information in a comprehensible manner has been a key challenge of 21st century. Human cognition relies mainly on their visual system. However, visual analysis includes both, visual complexities, as well as computational challenges. Visual complexities include the limitations of the display devices; whereas computational challenges involve the complex algorithm to extract and process the information from large data sets. This thesis aims to devise strategies in advancing the techniques used for visual analysis in the sports industry. Different algorithms are used to extract the location of the player from a video stream of football game. The obtained information is processed in a way, to be presented on a smaller screen like cell phones. The integration between the two devices is managed by a cloud, which works in real time. Furthermore, conclusions are made based on the average precision and frame throughput. Moreover, most processes are automated to minimize the human input and be adopted by multiple application domains.