Precision Agriculture Deep Neural Network Driven Multi-hop Plant Image Noisy Data Transmission and Plant Disease Detection
Résumé
Under precision agriculture (PA), plant disease detection
(PDD) is imperative regarding farm crops’ life quality and
crop yield. However, the data captures for PDD is influenced by
the noisy data captured by farm sensors due to wireless noisy
transmission channels. Hence, this work considers the onsite
or offsite (remote) farm PDD through onsite farm monitoring
PA sensor networks (PAN). Here, effects on captured sensor
image (plant leaf image) data transmitted through the PAN to
an PDD application are studied. Where both traditional decodeand-
forward (DF) data routing and channel-effect considering
machine learning data autoencoder routing are used for image
data transmission. In addition, a PDD deep learning algorithm
is developed to predict whether or not a farm plant is diseased,
based on the noisy image data captured by the PAN through data
routing. From the PAN-PDD simulation, the proposed ML PANPDD
algorithm showed fair performance over the DF PAN-PDD.
Mots clés
Deep learning (DL)
multi-hop wireless sensor network (WSN)
plant disease detection (PDD)
convolutional neural network (CNN)
autoencoder. Offsite Application Access Onsite Farm Containing Precision Agriculture Network Internet Camera Other Farm Gateway/ Sensor Type Central System Image Data Signal Routing Path
Domaines
Apprentissage [cs.LG]| Origine | Fichiers produits par l'(les) auteur(s) |
|---|