Classify Color Quality of a Newspaper Image Using Fuzzy Logic and Neural Networks

Abstract

In the Newspaper industry, the use of multicolor printing has been increased rapidly during the last decade. With these rapidly increased requirements it is not possible to identify color quality of a newspaper image using naked eye. Many color vision systems require a first step of classifying pixels in a given image into a discrete set of color classes. In this paper we describe an approach to pixel color segmentation to classify color quality of a newspaper image. Fuzzy sets are defined on the H, S and V components of the HSV color space and provide a fuzzy logic model that aims to follow the color detection. And then use threshold value and saturation value of a color as neural networks inputs to train neural network to classify color quality. This will make an efficient color quality assessment system to Sri Lankan news paper industry to reduce wastage and cost.