Adaptive neuro-fuzzy inference system for recognition of cotton leaf diseases

The objective of this work was to develop and to evaluate adaptive neuro-fuzzy inference system as methodology to identify the leaf diseases on cotton. This paper presents automatic system for classification of three cotton leaf diseases namely Bacterial Blight, Myrothecium and Alternaria. Graph cut method is used for segmentation of images to extract color layout descriptors as features to train the adaptive fuzzy inference system. The testing samples are collected from Central Institute for Cotton Research, Nagpur and from the fields in Buldana and Wardha district.