Image Edge Orientation Estimation via Fuzzy Logic

Abstract

This paper proposes a novel image edge orientation estimation using fuzzy logic approach. The idea behind orientation estimation is to design edge directed image interpolation for image super resolution enhancement. The calculation of image edge pixel orientation is based on different types of gradients measured across different angles from particular edge pixel. Orientation of edge pixel is classified in to nine angles to ensure the quality of orientation estimation. The advantage of proposed fuzzy based method is that, orientation calculation is done in human language with the help of simple fuzzy rules. Use of simple membership function and rules helps to design an efficient orientation estimation frame work. The experimental results show that fuzzy based orientation estimation outperforms even if the image is affected by noise.