A Software Framework for Image Retrieval and Visual Understanding in Dynamic and Sensor Rich Environments
Abstract
Performing search and retrieval operations with massive amounts of visual and environmental sensor information is problematic in time sensitive and mission critical situations such as emergency management and disaster response. Distinct sensor readings can be fused together to create a compact multimodal representation of a location. Efficient search and retrieval can be an answer to the problem of scale. Content Based Image Retrieval systems inherently rely on the search and retrieve operations to support timely and accurate responses. However, there is currently no adequate software framework system for multimodal CBIR to support situational awareness in dynamic and sensor rich environments. In this thesis, an extensible framework for CBIR is proposed to support an understanding of a sensor rich environment through the automated search and retrieval of relevant images and the context of their capture. This constitutes assisted CBIR as embodied in the proposed framework for multi-sensor assisted CBIR system.