Color object recognition methods that are based on image retrieval algorithms can handle changes of illumination via image normalization, e.g. simple color-channel-normalization1 ...
We propose a new fast algorithm for solving one of the standard formulations of image restoration and reconstruction which consists of an unconstrained optimization problem where t...
This contribution proposes a compositionality architecture for visual object categorization, i.e., learning and recognizing multiple visual object classes in unsegmented, cluttered...
Overcomplete representations are attracting interest in image processing theory, particularly due to their potential to generate sparse representations of data based on their morp...
Linear Discriminant Analysis (LDA) technique is an important and well-developed area of image recognition and to date many linear discrimination methods have been put forward. Desp...