Computer-Aided Plant Diseases Prediction Systems using Image Processing: Challenges and Methodologies

Abstract

This paper outlines the important, motivating, and inspiring problems in agriculture, the problem of plant diseases prediction systems using hyperspectral images. An computer-aided diagnosis (CAD) system for plant diseases prediction is designed to improve the survival of plant. The numerous researches have been investigated in recent times. A typical plant diseases prediction system is composed of five main processing steps: plant image acquisition, to improve the quality of the plant image using pre-processing, selection of more informative bands from preprocessed images by band selection process, then various features to discriminate different types of plant diseases is extracted using the feature extraction technique, and finally the extracted features of the plant disease trained for classification. This paper summarizes the recent findings proposed to design the plant diseases prediction system by the researchers. Also, the paper lists out various challenges that are faced by researchers in designing the plant diseases prediction system and outline the advantage and disadvantage of the existing system.