Output details
13 - Electrical and Electronic Engineering, Metallurgy and Materials
Newcastle University
Region-of-Interest Extraction in Low Depth of Field Images Using Ensemble Clustering and Difference of Gaussian Approaches
Extracting salient objects from images is a difficult problem without efficient solution. This research presents an automatic solution for extracting regions of interest from complex visual images, regardless of the complexity of foreground and background. The solution achieves a high level of performance approaching perfect matching and is computationally three times faster than existing state of the art approaches. This work was used as foreground knowledge for a KTP project with ADL Smartcare KTP008259 £154,918 (contact peter.gore@adlsmartcare.com), where a single poor quality image taken by customers can be used to infer dimensions for stairlifts for the elderly.