Role of Hyperbox Classifiers for Color Recognition

Abstract

In this research work, we explore the application of hyperbox classifier for the color recognition problem. Color is an important cue in computer vision and its recognition plays an important role in many vision tasks. Color boundary separation has earlier been performed by use of traditional classifiers like SVM, Naive Bayes etc. However, the intrinsics of color are very delicate and their separation using the traditional classifiers have not achieved acceptable accuracy. In this research work we show the applicability of hyperbox classifiers for color recognition task and perform comparative analysis with traditional classifiers to validate that hyperbox classifiers achieve better color boundary separation results.