Detection And Classification Of Sunflower Leaf Disease Using Convolutional Neural Network (Cnn)

Abstract

Sunflower Is A Seed Crop For Essential Oils That Can Be Employed In Agricultural And Oil Production Systems. Sunflower Crops Are Harmed By A Variety Of Diseases, Insects, And Nematodes, Resulting In A Wide Range Of Productivity Losses. Disease Detection Can Be Done With The Naked Eye; However, This Procedure Is Impractical On Large Farms. To Address This Problem, Several Scholars Used A Variety Of Methods And Techniques To Conduct Research. Previous Research Has Used Particle Swarm Optimization, Digital Image Processing, K-Means ++ Clustering, And Other Methods To Solve This Problem. Ethiopian Researchers Conduct Numerousplant Disease Detection And Classification Investigations. Even With Ethiopia's Large Production, However, The Detection And Classification Of Sunflower Diseases Receive Little Attention.The Main Objective Is To Develop A System For Detecting And Classifying Sunflower Leaf Disease Using Deep Learning Technologies. To Achieve The Main Objective Convolution Neural Network Methods Are Used In This Research. CNN Is Used In Image Classification Because Of Its Accuracy, While RNN Is Commonly Utilized With Sequences Such As Text, Sound, Video, And More. The Sunfnet Model Was Created In This Study Utilizing These Strategies. The Constructed Model Has Ten Layers: Nine Convolution Layers And One Output Layer. This Model Is Developed Using A 7850(Seven Thousand Eight Hundred Fifty) Training Data Set Augmented From Original Images. Back Propagation And Gradient Descent Optimization Algorithms Are Used To Develop This Model. The Confusion Matrix Is Used To Assess Model Performance. Sunfnet Model Is Tested In Different Epochs And Learning Rates To Generate An Efficient Model. Our Result Of The Experiment Is 99.88% Training Accuracy With 99% Validation Accuracy. This System's Accuracy Is About ~99.8.