Optimized convolutional neural network for identification of maize leaf diseases with adaptive ageist spider monkey optimization model

In recent years, the number of maize disease species has increased, which obviously increases the level of damages in leaves. The reason for maize leaf disease is due to variations in agriculture systems, the variants of pathogen, and it also occurs due to the scarcity of plant conservation measures...

Full description

Saved in:
Bibliographic Details
Published in:International journal of information technology (Singapore. Online) Vol. 15; no. 2; pp. 877 - 891
Main Authors: Arjunagi, Shravankumar, Patil, Nagaraj B.
Format: Journal Article
Language:English
Published: Singapore Springer Nature Singapore 01.02.2023
Springer Nature B.V
Subjects:
Accuracy
Agricultural production
Algorithms
Artificial Intelligence
Artificial neural networks
Classification
Computer Imaging
Computer Science
Corn
Crop diseases
Crops
Deep learning
Feature extraction
Identification
Image Processing and Computer Vision
Leaves
Machine Learning
Monkeys
Optimization
Optimization models
Original Research
Pathogens
Pattern Recognition and Graphics
Plant diseases
Software Engineering
Spectrum analysis
Vision
AOSMO model
CNN
Performance measures
Maize leaf detection
4D-LBP features
ISSN:2511-2104, 2511-2112
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:In recent years, the number of maize disease species has increased, which obviously increases the level of damages in leaves. The reason for maize leaf disease is due to variations in agriculture systems, the variants of pathogen, and it also occurs due to the scarcity of plant conservation measures. The disease in maize leaves can be exhibited by varied symptoms; however, it might be complex for farmers to identify the disease in naked eye. Therefore, this paper intends to present a new automatic system for identifying and diagnosing maize leaf diseases. The proposed model includes two major phases: Proposed Feature Extraction and Classification. The first phase is feature extraction, where the proposed 4D-Local Binary Pattern (4D-LBP) based texture features will be extracted. More particularly, Dimension 1 insists pixel intensity, dimension 2 insists angle, dimension 3 insists local frequency from intensity patch and dimension 4 insists global frequency as well. Once the features get extracted, they are subjected for classification process, where the optimized Convolutional Neural Network (CNN) is used, where the count of convolutional layers is optimally tuned. For this optimal selection, a new Adaptive Opposition based Spider Monkey optimization (AOSMO), which is the enhanced version of SMO algorithm. At last, the performance of proposed work is evaluated over other traditional models with respect to accuracy.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:2511-2104
2511-2112
DOI:10.1007/s41870-021-00657-3