To add to what James presented, some specific applications could be 1) search and rescue operations (e.g., identifying a hiker in a forest or on a mountain); 2) crop analysis (e.g., estimating water requirements from imagery or determining pesticide requirements); 3) monitoring pollution levels in lakes or rivers using aerial imagery.
Joseph - There's been a lot of work on post-disaster damage mapping using images taken with UAVs. The processing entails both extracting of direct image parameters (color, texture), but, more importantly, 3D information. Therefore these images are acquired with overlap to process them photogrammetrically. The resulting point clouds are then used to detect things like (geometric) planarity, they can be integrated with data taken from other sources (such as terrestrial laser scanners), etc., Of course also videos can be flows on UAVs and the data processed in similar ways. Have a look at some of the papers our group has published, such as
Article: UAV-based urban structural damage assessment using object-based image analysis and semantic reasoning
Fernandez Galarreta, Norman Kerle, Markus Gerke
Article: Identification of damage in buildings based on gaps in 3D point clouds from very high resolution oblique airborne images
Anand Vetrivel, Markus Gerke, Norman Kerle, George Vosselman
Conference Paper: Segmentation of UAV-Based Images Incorporating 3D Point Cloud Information
You might consider studying the elements and practices of airborne, real-time, urban pedestrian swarm activity predictive analysis. It is an essential component in determining safe downing site criteria for weaponized sUAS BVLOS interdiction planning, and automated first response infrastructure evolution for 'Smart Cities' architecture.