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.

Keywords: U-Net CNN; deep learning; inception-net; object detection and recognition; plant disease detection.

Figure 1

(A) Original Baseline U-Net Architecture, (B) Modified improved U-net Deep Neural Network Architecture Livne et al., 2019.

Figure 2

Sisualization of tomato leaf images using the Score-CAM tool, demonstrating affected regions where CNN classifier makes the majority of its decisions.

Figure 3

Proposed Tomato Plants leaf diseases Classification Model.

Figure 4

Some random samples of tomatoes leaf images from the benchmar Plant Village Dataset.

Figure 5

Original Tomota Leaf Images, Ground Truth Mask and Segmented Leaf using Modified U-Net CNN Model.

Figure 6

Working feature curves for (A) binary segmented leaf classification, (B) sixth class segmented leaf classification, and (C) tenth class classification of the segmented leaf.

Figure 7

Image classification using compound scaling CNN-based models of healthy and diseased tomato leaves for segmented leaf images (A) for 2 class classification, (B) for 6 class classification, and (C) for 10 class classification.

Figure 8

Accurate classification and visualization of ROI using the CAM-Score tool: The red intensity indicates the severity of the lesion.

Figure 9

Visual Results of Proposed Modified U-Net CNN model.

Figure 10

Proposed Model Comparison with state-of-the-artwork.