Multi-channel capsule network ensemble for plant disease detection

Abstract

This study presents a new deep learning approach based on capsule networks and ensemble learning for the detection of plant diseases. The developed method is called as multi-channel capsule network ensemble. The main innovation behind the proposed method is the use of multi-channel capsule networks, individually trained on images applied different preprocessing techniques and then combined together. In this way, the final ensemble can better detect plant diseases by making use of different attributes of the data. Our experiments carried out using a well-known data set and various state-of-the-art classification methods demonstrate that our classification approach can provide competitive advantages in terms of classification accuracy. Article Highlights An ensemble of capsule networks has been developed for the automated detection of plant diseases with high accuracies. Better accuracy has been achieved with ensemble learning compared to a single model With the proposed method, better results have been obtained compared to state-of-the-art classification methods in the literature

Transformative Role of Artificial Intelligence in Advancing Sustainable Tomato (Solanum lycopersicum) Disease Management for Global Food Security: A Comprehensive Review

“…This approach combines ensemble learning with a five-channel capsule network, which draws information from five distinct data sources. The ensembled capsule network model outperformed all other models that have been previously reported in the literature, demonstrating its superior performance [31].…”

“…Dataset Type of Dataset [7,19,21,22,27,31, PlantVillage Public [101,102] AIChallenger Public [17,18] Own dataset Public [103] Tomato Leaf Disease Detection Public [104] Dataset of Tomato Leaves Public [105] PlantVillage + Plant Disease Severity Public [106] PlantVillage + New Plant Diseases Public [107] PlantVillage + AIChallenger Public [33] PlantVillage + AIChallenger + PlantDoc Public [32,108] PlantVillage + PlantDoc Public [109][110][111][112][113][114][115][116][117] PlantVillage + own dataset Public + Private [8,23,24,[118][119][120][121][122][123][124][125][126][127] Own dataset Private…”

A Capsule Attention Network for Plant Disease Classification

“…Peker [19] proposed the ensemble capsule network with five channels for plant disease detection. Gabor, PCA, filters were combined with the capsule network to achieve the best performance in detecting the texture, shape, or leaf-affected area more accurately.…”

“…Gabor, PCA, filters were combined with the capsule network to achieve the best performance in detecting the texture, shape, or leaf-affected area more accurately. The limitation of this network is high training complexity and overfitting problems arise at the preprocessing level [19].…”

Multi-Scale Convolution-Capsule Network for Crop Insect Pest Recognition

“…Wang et al [20] proposed a multiscale convolutional CapsNet for hyperspectral image classification, which is composed of a multi-scale convolutional layer, a single-scale convolutional layer, a PrimaryCaps layer, a DigitCaps layer, and a fully connected layer. Peker [21] proposed a multi-channel CapsNet ensemble for plant disease detection and individually trained the network on the image set. Thenmozhi et al [22] proposed a deep CNN model to classify insects, where transfer learning was applied to fine-tune the pre-trained models.…”