Hybrid Deep Learning Based Plant Leaf Disease Identification and Crop Recommendation System to Enhance Agriculture Productivity

Abstract

In the Gross Domestic Product (GDP), agriculture production plays very crucial role and is a growth engine of a nation. Various countries are using advance technology like Artificial Intelligence (AI), Machine Learning (ML) and Internet of Things (IoT) within the agriculture sector which is referred as precision agriculture. Precision agriculture is a concept which involves all above technologies for increasing overall productivity of agriculture and to reduce time. Precise agriculture consists many domains which can be accomplished using AI and IoT. Plant leaf disease identification and crop recommendation are the two important domains of precise agriculture which uses AI and IoT technologies. Deep Learning (DL) on the other hand has the capacity for processing image data in proper way by extracting features. This research work ML and DL respectively for crop recommendation and plant leaf disease identification. Two separate ML models were developed by utilizing data obtained from kaggle. In order to utilize strength of multiple ML and DI techniques, a hybrid approach has been proposed in which two ML techniques: Support Vector Machine (SVM) and Artificial Neural Network (ANN) are combined together for crop recommendation. Results revels that hybrid of ANN and SVM achieved highest$\text{accuracy}=97.50 \%,\ \text{sensitivity}=97.61 \%$and$\text{specificity}=99.88 \%$as compared to individual technique. A hybrid of Convolutional Neural Network (CNN), VGG16 and EfficientNetV2S are combined for plant leaf disease identification resulting in$\text{accuracy}=97.33 \%,\ sensitivity =96.87 \%$and$\text{specificity} = 98.84 \%$.

Plant phenotyping relevance

葉画像から植物病害を識別する深層学習手法の開発と性能評価が研究の主要な貢献であり、植物の病害状態を直接推定しているため含める。作物推薦部分は対象外だが、病害識別手法だけで基準を満たす。

Code and data availability

公開本文の所在を確認できませんでした。非公開または購読が必要な可能性があります。

No evidence-backed public reproduction asset is currently recorded.