Classification of Multiple Maize Leaf Diseases Using a Blended Convolutional Neural Network

Abstract

India is mostly an agricultural nation and produces an abundance of agricultural goods. Corn is one of the crops that is grown in India and utilized as a staple meal for the people there. This corn plant has to be safeguarded against illnesses to ensure that the harvested maize will be of the highest possible quality. Farmers need to identify diseases in maize plants at an early stage so that they can provide medication in a timely and accurate manner. Previous studies have attempted to overcome this issue by using various machine-learning approaches. The prior study did not provide the best possible findings since just a little quantity of data was utilized, and the data that was used was not very diverse. As a consequence, we suggest a method that is capable of processing large amounts of diverse data with the hope that the final system will be more accurate than the studies that came before it. In this study, feature extraction is accomplished by the use of transfer learning methods, and classification is performed using a convolutional neural network. We performed an analysis of the Convolution and pooling operations in conjunction with the Flatten layer. The findings of the experiments indicate that the accuracy generated by the suggested model is superior to that provided by existing Deep Learning techniques. The accuracy of 97% that was achieved demonstrates that the new method is more accurate than the previous one.