Study on the Prediction Model of Litchi Downy Blight Damage Based on IoT and Hyperspectral Data Fusion

The Internet of Things (IoT) and hyperspectral technology have been widely applied in the field of crop disease monitoring. However, the effective integration of these two data modalities remains an exigent challenge. This study concentrates on the litchi downy blight disease and proposes a model th...

Full description

Saved in:
Bibliographic Details
Published in:IEEE internet of things journal 2024-08, Vol.11 (16), p.27184-27200
Main Authors: Lu, Jianqiang, Wu, Zhiyun, Lan, Yubin, Deng, Xiaoling, Huang, Jiewei
Format: Article
Language:English
Subjects:
Adaptive algorithms
Algorithms
Artificial intelligence
Blight
Crop diseases
data fusion
Data integration
Data models
disease forecast processing
Diseases
Feature extraction
Hyperspectral imaging
Internet of Things
litchi downy blight
Memory devices
multimodal sensors
Multisensor fusion
Plant diseases
Prediction models
Predictions
Predictive models
Citations: Items that this one cites
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The Internet of Things (IoT) and hyperspectral technology have been widely applied in the field of crop disease monitoring. However, the effective integration of these two data modalities remains an exigent challenge. This study concentrates on the litchi downy blight disease and proposes a model that combines IoT and hyperspectral data for precocious prediction utilizing artificial intelligence algorithms.In this model, IoT data is collected by IoT sensor devices. We proposed 15 sensitive feature factors closely related to litchi downy blight and utilized a long short-term memory (LSTM) network to extract serialized features from IoT data. Hyperspectral data is collected by a ground object spectrometer, and the Savitzky-Golay (S-G) algorithm is applied for data preprocessing. The successive projections algorithm (SPA) is utilized for the extraction of spectral features for subsequent modeling and prediction. Finally, a Bayesian probability model predicated on adaptive kernel density estimation is incorporated into the framework, and an adaptive weight algorithm is devised to construct a multimodal data fusion-based predictive model for litchi downy blight affliction. The results show that the proposed model achieves a prediction accuracy of 89.58%, significantly higher than using only environmental data or hyperspectral data. This method effectively integrates IoT, hyperspectral data and artificial intelligence technologies, providing new insights for the application of IoT technology and the development of modern agriculture.
ISSN:2327-4662
2327-4662
DOI:10.1109/JIOT.2024.3397625