A Model for African Fabrics Analysis and Recognition

Abstract

In this paper, a model to analyse and distinguish traditional African fabric patterns is presented. The model can be a valuable tool in image retrieval systems. Selected African fabric patterns were analysed using image processing and wavelet analysis techniques to extract relevant features for the recognition purpose. The recognition model consisted of multiple multi-layered artificial neural networks that used statistical and structural properties to recognize African fabrics' patterns. The model was simulated in MATLAB environment and its performance with respect to the accuracy of the recognised fabrics patterns was evaluated using the following metrics sensitivity, specificity and efficiency.