Diagnosis of Early Blight Disease in Tomato Plant based on Visible/Near-Infrared Spectroscopy and Principal Components Analysis- Artificial Neural Network Prior to Visual Disease Symptoms
Journal Title: Journal of Agricultural Machinery - Year 2022, Vol 12, Issue 1
Abstract
Early diagnosis of plant diseases before the occurrence of symptoms can reduce the loss of the yield and increase the quality of agricultural crops. It also reduces the consumption of pesticides, environmental risks, and the cost of production. For this reason, the objectives of the present study were non-destructive diagnosis of early blight of tomato plant and discrimination of the most important agents of early blight (A. solani and A. alternate) in the primary stages of incidence of the disease before appearing visual symptoms using Vis-NIR spectroscopy (400-900 nm). The spectral data were acquired from the leaves of the plants infected with A. solani and A. alternate, 48 hours, 72 hours, 96 hours, and 120 hours after inoculation. To develop the recognition model based on the spectral data, principal components analysis (PCA) coupled with artificial neural network (ANN) was used. The results showed that the PCA-ANN model could diagnose the infected plants and pathogen species with accuracy of 93-100% for test set samples. In 96 hours after inoculation, in addition to the simpler model (8 PCs and 3 neurons in hidden layer), accuracy of 100% was obtained. At all times after inoculation, there was no error in diagnosis of the plants infected with A. solani that is more pathogenic and aggressive than other species, from healthy plants. Early blight in tomato plant and the type of pathogen before visual symptoms, without any plant sample preparation, could be diagnosed non-destructively (with accuracy of 93-100%) using Vis-NIR (400-900 nm) spectroscopy coupled with PCA-ANN. It was concluded that this technology could be used for rapid, low-cost, and early diagnosis of this disease in tomato plant instead of time-consuming, expensive, and destructive laboratory methods.
Authors and Affiliations
F. Azadshahraki,K. Sharifi,B. Jamshidi,R. Karimzadeh,H. Naderi,
Keywords
Related Articles
- EP ID EP718019
- DOI https://doi.org/10.22067/jam.2021.67436.1001
- Views 54
- Downloads 0
Viscoelastic modeling of apples under quasi-static loading using finite element method to investigate the causes of bruising
Introduction: Apple is one of the most important horticultural crops of Iran. Its production in the country stands in the second place after citrus. Iran holds the fourth place in the world production of apples and gains...
Evaluation of Different Methods of Sugarcane Hilling Up in Two Soil Textures and Three Forward Speeds
Introduction Sugarcane cultivation in Khuzestan province is in the form of planting in-furrow. Due to the fact that in a machine harvesting, the reaper is not able to fully harvest the straw in the furrow, in the plantin...
Modeling and Optimization of the Grinding Cereals to Produce Poultry Feed in the Hammer Mill using Response Surface Method (RSM)
IntroductionThe most costly part of poultry breeding is feeding. Due to the noticeable developments in animal husbandry and agricultural sectors, it is necessary to use the mechanized methods to reduce the casualties, in...
Simulation of Heat and Mass Transfer in a Refractance Window Dryer for Aloe vera gel
IntroductionDrying is one of the oldest methods of food preservation. To increase the efficiency of heat and mass transfer while maintaining product quality, the study of the drying process is crucial scientifically and...
The Effect of Three Common Grain Drills on Dryland Wheat Yield
A three-year field experiment (2004-2007) was conducted on a silty clay loam soil at Tajarak Research Station of Hamedan to determine proper grain drill for wheat in Hamedan dryland areas. In this study, three grain dril...