Analyzing The Prediction Accuracy Of Corn Leaf Diseases Using A Pre-Trained Network Model

During corn’s exploration and manufacturing phases, farmers have a complicated issue in accurately diagnosing corn crop infections. To solve this issue, this work provides a method for specific position three prevalent diseases of corn leaves: grey spot, leaf blight, and rusty depending on grouping and an upgraded deep learning method. Deep learning advancements paved the path for enhancing prediction accuracy as its significance is adopted over various research fields. Disease prediction and classification is time-consuming due to the lesser foreground and background intensity information. To identify three illnesses, first cluster reference images using the clustering technique, then input them into the enhanced deep network. The influences of various ‘k’ values on corn diagnostic techniques are investigated in this research. The trial findings show that the approach has the most significant identification impact on observations with the analytical prediction of corn disease, rust, and grey spot disease. Here, VGG-16 and ResNet18 likewise produce the best diagnostic findings, with an average diagnostic accuracy, respectively. The technique presented in this study has a diagnostic performance of 95% for the three corn diseases. It has a more substantial diagnostic impact than another four techniques and can be used to safeguard crops in the agricultural area. The simulation is done in the MATLAB 2020a environment, and various performance metrics like accuracy, precision, F1-score, recall and some other statistical measures are evaluated and compared with existing approaches.