Plant Disease Detection Using Sequential Convolutional Neural Network

Abstract

The main warnings in the area of food preservation and care are crop diseases. It has been recognized speedily, but it is not as easy as in any area of the world because no required framework exists. Both the healthy and diseased plant leaves were gathered and collected under the condition and circumstances. For this purpose, a public set of information was used. It was 20,639 images of plants that were infected and healthy. In order to recognize three different crops and 12 diseases, a sequential convolutional neural network from Keras was trained and applied. The perfection and exactness was 98.18% onset of information of the above trained mentioned model using CNN . It has also indicated the probability and possibility of this strategy and procedure. The over-fitting occurs and neutralizes by putting the dropout value to 0.25.