HYBRID DEEP LEARNING WITH ALEXNET FEATURE EXTRACTION AND UNET CLASSIFICATION FOR EARLY DETECTION IN LEAF DISEASES
Abstract
This study addresses the imperative need for early detection of leaf diseases in tobacco, pepper, and tomato plants, as these diseases significantly impact crop yield and quality. Existing methods often fall short in accurately identifying diseases across diverse plant species. The research aims to bridge this gap by proposing a hybrid deep learning approach, combining the robust feature extraction capabilities of AlexNet with the precise segmentation and classification prowess of UNet. The proposed hybrid model leverages AlexNet proficiency in extracting hierarchical features from plant leaf images and subsequently utilizes UNet for accurate disease classification. This synergistic combination enables the model to overcome the challenges posed by the varied morphologies of tobacco, pepper, and tomato leaves. Experimental results demonstrate the effectiveness of the proposed methodology, showcasing superior performance in terms of accuracy, sensitivity, and specificity compared to existing techniques. The hybrid deep learning approach exhibits promising potential for early and accurate detection of leaf diseases, contributing to sustainable crop management practices.

Authors
Uddyalok Chakraborty1, D. Thilagavathy2, Suresh Kumar Sharma3, Awadh Kishor Singh4
UPL University of Sustainable Technology, India1,4, Adhiyamaan College of Engineering, India2, Sri Karan Narendra Agriculture University, India3

Keywords
Leaf Disease Detection, Hybrid Deep Learning, AlexNet, UNet, Agriculture
Yearly Full Views
JanuaryFebruaryMarchAprilMayJuneJulyAugustSeptemberOctoberNovemberDecember
0297600000000
Published By :
ICTACT
Published In :
ICTACT Journal on Soft Computing
( Volume: 14 , Issue: 3 , Pages: 3255 - 3262 )
Date of Publication :
January 2024
Page Views :
131
Full Text Views :
42

Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.