University of Dayton, USA
Title: Visual perception for autonomy application to object detection and behavior analysis in complex environmental conditions
Biography: K Asari Vijayan
Automated visual monitoring involves data acquisition, analysis, and interpretation for understanding objects and object behaviors. Automated visual data analysis systems are mostly used for military, law enforcement and commercial applications. Sensors of different types and characteristics in various platforms are used for the acquisition of data. In recent years, there has been a spurt in the development of palm-sized cameras for the consumer market that are equipped with fish eye lenses on front and back sides to deliver 360 degree spherical views. Intelligent analysis of these data is an important task in applications such as automatic human detection, identification, activity recognition, behavior analysis, anomaly detection, alarming, etc. Object motion analysis and interpretation are integral components for activity monitoring and situational awareness. We present the development of a robust automated system which can detect and identify people using imagery captured using visible, infrared and omnidirectional cameras in a mobile platform and track their actions and activities by a spatiotemporal feature tracking mechanism. The automated visual analysis procedure includes preprocessing of data for distortion correction, novel methods for feature extraction, and machine learning based approaches for object classification.