A Canny Edge Detection Algorithm for Automatic Identification of Plant Diseases

Abstract

Plant diseases are important to investigate because they cause loss of plant life and product. Various forms of losses can occur in the field, in storage, or at any point between sowing and harvesting. Direct monitory loss and material loss are caused by the diseases. Plant diseases continue to cause agony to untold millions of people around the world, resulting in an estimated annual yield loss of 14% and a global economic loss of $220 billion dollars. Plants were afflicted by several diseases 250 million years ago, according to fossil evidence. The Plant sickness has been linked to a number of significant events in the Earth's history. In this paper, we have analyzed various types of diseases which affects the plants and applied some machine learning classification approaches to plant dataset. Among these classification steps, we identified the percentage of accuracy in each algorithm so that proper approach can be followed to identify the plant diseases at the earlier stage.