Abstract:Deep learning (DL) plays an important role in modern agriculture, especially in plant pathology using leaf images where convolutional neural networks (CNN) are attracting a lot of attention. While numerous reviews have explored the applications of DL within this research domain, there remains a notable absence of an empirical study to offer insightful comparisons due to the employment of varied datasets in the evaluation. Furthermore, a majority of these approaches tend to address the problem as a singular prediction task, overlooking the multifaceted nature of predicting various aspects of plant species and disease types. Lastly, there is an evident need for a more profound consideration of the semantic relationships that underlie plant species and disease types. In this article, we start our study by surveying current DL approaches for plant identification and disease classification. We categorise the approaches into multi-model, multi-label, multi-output, and multi-task, in which different backbone CNNs can be employed. Furthermore, based on the survey of existing approaches in plant pathology and the study of available approaches in machine learning, we propose a new model named Generalised Stacking Multi-output CNN (GSMo-CNN). To investigate the effectiveness of different backbone CNNs and learning approaches, we conduct an intensive experiment on three benchmark datasets Plant Village, Plant Leaves, and PlantDoc. The experimental results demonstrate that InceptionV3 can be a good choice for a backbone CNN as its performance is better than AlexNet, VGG16, ResNet101, EfficientNet, MobileNet, and a custom CNN developed by us. Interestingly, there is empirical evidence to support the hypothesis that using a single model for both tasks can be comparable or better than using two models, one for each task. Finally, we show that the proposed GSMo-CNN achieves state-of-the-art performance on three benchmark datasets.

Deep learning for plant identification using vein morphological patterns

Deep Learning for Plant Species Classification Using Leaf Vein Morphometric

Research on the Fine-Grained Plant Image Classification

Plant identification based on very deep convolutional neural networks

Plant Species Identification by Bi-channel Deep Convolutional Networks

SWP-LeafNET: A novel multistage approach for plant leaf identification based on deep CNN

Plant Identification Using New Architecture Convolutional Neural Networks Combine with Replacing the Red of Color Channel Image by Vein Morphology Leaf

Leaf Identification Using a Deep Convolutional Neural Network

Plant Leaf Recognition Based On Small Datasets Using Deep Learning Algorithm

Identification of Soybean Leaf Diseases Via Deep Learning

Deep Convolutional Neural Network for Plant Seedlings Classification

Deep Learning for Plant Identification and Disease Classification from Leaf Images: Multi-prediction Approaches

Automated and accurate segmentation of leaf venation networks via deep learning

Hybrid Deep Learning for Plant Leaves Classification

Plant-part Segmentation Using Deep Learning and Multi-View Vision

Classification of Plants Using Convolutional Neural Network

Transfer Learning with Deep Convolutional Neural Network for Automated Plant Identification

Optimized convolutional neural network model for plant species identification from leaf images using computer vision

A DEEP LEARNING METHODOLOGY FOR PLANT SPECIES RECOGNITION USING MORPHOLOGY OF LEAVES

Novel deep learning assisted plant leaf classification system using optimized threshold-based CNN

Feature Extraction of Plant Leaf Using Deep Learning