PLANT DISEASE DETECTION USING TRANSFER LEARNING

PLANT DISEASE DETECTION USING TRANSFER LEARNING

N. PRAKASH

Assistant Professor, Dept. of Information Technology, Sreenidhi Institute of Science and Technology

prakashnakirekanti@sreenidhi.edu.in

DR. K. KRANTHI KUMAR

Associate Professor, Dept. of Information Technology, Sreenidhi Institute of Science and Technology

kranthikumark@sreenidhi.edu.in

ALEKHYA CHENNOJU

B.Tech Student, Dept. of Information Technology, Sreenidhi Institute of Science and Technology

alluchennoju1205@gmail.com

MOUNIKA MADAS

B.Tech Student, Dept. of Information Technology, Sreenidhi Institute of Science and Technology

mouniikamadas2706@gmail.com

DODLA HARINI

B.Tech Student, Dept. of Information Technology, Sreenidhi Institute of Science and Technology

harinir694@gmail.com

Abstract - Plant diseases are essential in agricultural output, and early diagnosis of plant diseases is seen as critical. Current computational intelligence and computer vision approaches show promise in improving illness diagnosis. Convolutional Neural Network (CNN) models can identify plant illnesses in agricultural fields and in photographs of plantation leaves. Xception is an extension of InceptionV3 with fewer but superior feature extraction settings. This recommends using an ensemble of Densenet and Xception to increase plant disease detection performance by concatenating the derived features. When evaluating the whole Plant Village dataset, the Densenet, Xception, and ensemble models obtained 98.9% accuracy.  The suggested model's performance evaluation shows that it successfully categorizes healthy, Scab, Rust, and Multiple Diseases, outperforming pre-trained models.

Keywords: CNN, Transfer Learning, Xception, Densenet, Plant Disease.