A Survey on Leaf Disease Prediction Algorithms using Digital Image Processing

摘要

Objective : To investigate the plant leaf disease prediction algorithms that utilizes the digital image processing techniques in agricultural environments. Findings : In digital image processing, the segmentation process of healthy and diseased tissue was mainly focused in order to detect and diagnose the plant leaf diseases accurately.Semi-automatic segmentation technique was mostly utilized among various segmentation methods, which was developed based on the grayscale histogram. However, the issue of accuracy in segmentation process was still not improved. In this paper, the leaf disease prediction algorithms are investigated briefly according to the digital image processing techniques and evaluated the performance effectiveness of different algorithms. Results : In this paper, various segmentation algorithms are studied which are used to predict the leaf diseases through digital image processing techniques in terms of their merits and demerits to prove segmentation based on grayscale histogram is better than other segmentation techniques to predict the leaf diseases. Application/Improvements : The finding of this study shows that the segmentation technique based on grayscale histogram is better than the other digital image processing techniques.