Computational Intelligence Techniques in Visual Pattern Recognition

Abstract

Efficient feature selection and classification algorithms are necessary for the effective recognition of visual patterns. The initial part of this dissertation presents fast feature selection and classification algorithms for multiple feature data, with application to visual pattern recognition. A fuzzy-rough approach is utilized to develop a novel classifier which can classify vague and indiscernible data with good accuracy. The proposed algorithm translates each quantitative value of a feature into fuzzy sets of linguistic terms using membership functions. The fuzzy membership functions are formed using the feature cluster centers identified by the subtractive clustering technique. The lower and upper approximations of the fuzzy equivalence classes are obtained and the discriminative features in the dataset are identified. The classification is done through a voting process. Two algorithms are proposed for the feature selection, an unsupervised algorithm using fuzzy-rough approach and a supervised method using genetic algorithm. The algorithms are tested in different visual pattern classification tasks: hand posture recognition, face recognition, and general object recognition. In order to prove the generality of the classifier for other multiple feature patterns, the algorithm is also applied to cancer and tumor datasets. The proposed algorithms identified the relevant features and provided good classification accuracy, at a less computational cost, with good margin of classification. On comparison, the proposed algorithms provided equivalent or better classification accuracy than that provided by a Support Vector Machines classifier, at a lesser computational time.

The later part of the thesis presents the results of the utilization of computational model of visual cortex for addressing problems in hand posture recognition. The image features have invariance with respect to hand posture appearance and its size, and the recognition algorithm provides person independent performance. The features are extracted in such a way that it provides maximum inter class discrimination. The real-time implementation of the algorithm is done for the interaction between the human and a virtual character Handy.

A system for the recognition of hand postures against complex natural backgrounds is presented in the last part of the dissertation. A Bayesian model of visual attention is utilized to generate a saliency map, and to detect and identify the hand region. Feature based visual attention is implemented using a combination of high level (shape, texture) and low level (color) image features. The shape and texture features are extracted from a skin color map, using the computational model of the visual cortex. The skin color map, which represents the similarity of each pixel to the human skin color in HSI color space, enhanced the edges and shapes within the skin colored regions. The hand postures are classified using the shape and texture features, with a support vector machines classifier. The algorithm is tested using a newly developed complex background hand posture dataset namely NUS hand posture dataset-II. The experimental results show that the algorithm has a person independent performance, and is reliable against variations in hand sizes. The proposed algorithm provided good recognition accuracy despite clutter and other distracting objects in the background, including the skin colored objects.