Deep learning-based segmentation and classification of leaf images for detection of tomato plant disease

Abstract

Plants contribute significantly to the global food supply. Various Plant diseases can result in production losses, which can be avoided by maintaining vigilance. However, manually monitoring plant diseases by agriculture experts and botanists is time-consuming, challenging and error-prone. To reduce the risk of disease severity, machine vision technology (i.e., artificial intelligence) can play a significant role. In the alternative method, the severity of the disease can be diminished through computer technologies and the cooperation of humans. These methods can also eliminate the disadvantages of manual observation. In this work, we proposed a solution to detect tomato plant disease using a deep leaning-based system utilizing the plant leaves image data. We utilized an architecture for deep learning based on a recently developed convolutional neural network that is trained over 18,161 segmented and non-segmented tomato leaf images—using a supervised learning approach to detect and recognize various tomato diseases using the Inception Net model in the research work. For the detection and segmentation of disease-affected regions, two state-of-the-art semantic segmentation models, i.e., U-Net and Modified U-Net, are utilized in this work. The plant leaf pixels are binary and classified by the model as Region of Interest (ROI) and background. There is also an examination of the presentation of binary arrangement (healthy and diseased leaves), six-level classification (healthy and other ailing leaf groups), and ten-level classification (healthy and other types of ailing leaves) models. The Modified U-net segmentation model outperforms the simple U-net segmentation model by 98.66 percent, 98.5 IoU score, and 98.73 percent on the dice. InceptionNet1 achieves 99.95% accuracy for binary classification problems and 99.12% for classifying six segmented class images; InceptionNet outperformed the Modified U-net model to achieve higher accuracy. The experimental results of our proposed method for classifying plant diseases demonstrate that it outperforms the methods currently available in the literature.

An advanced deep learning models-based plant disease detection: A review of recent research

“…Deep CNN ( Shoaib et al., 2022a ; Shoaib et al., 2022b )is a type of feedforward AI model that is consisting of several hidden layers of convolutional and pooling layers, the CNN model are the best of the DL model for achieving higher detection accuracy using imaging data The CNN model consist of two blocks, the features learning and classification blocks. The features learning block extract various kind of features using the convolutional layer where the features learning is performed at the fully connected layers.…”

“…U-Net is a popular CNN architecture for image segmentation tasks. The architecture is named U-Net because it is U-shaped, with encoder and decoder sections connected by a bottleneck ( Shoaib et al., 2022a ). The encoder section of the network consists of a series of convolutional and clustering layers that extract entities from the input image.…”

“… An example of performance comparison between three models using the ROC curve ( Shoaib et al., 2022a ). …”

An effective deep learning approach for the classification of Bacteriosis in peach leave

“…The authors presented an innovative method for detecting rice and leaf disease based on deep convolutional neural networks (CNNs) Lu et al, 2017 andShoaib et al, 2022. Various models were trained to detect ten types of rice diseases. They experimented with a dataset consisting of 500 images of healthy and infected rice leaves.…”

Cropformer: A new generalized deep learning classification approach for multi-scenario crop classification

“…In contrast to classical machine learning, deep learning no longer requires manually designed features, but can learn complex semantic features from high-dimensional data. Currently, deep learning has been widely used in agriculture due to its effectiveness, including crop classification (Kussul et al, 2017;Minh et al, 2018), pest and disease detection (Akbar et al, 2022;Shoaib et al, 2022a;Shoaib et al, 2022b), yield estimation (Nevavuori et al, 2019;Khaki et al, 2020), etc. The complex network structure of deep learning requires a large amount of labeled data for support, which creates difficulties for the agricultural fields where deep learning is applied.…”