Decision Tree-based Machine Learning Algorithms to Classify Rice Plant Diseases

Abstract

Rice is one of the most important foods on earth for human beings. India and China are two countries in the world mostly depend on rice. The output of this crop depends on the many parameters such as soil, water supply, pesticides used, time duration, and infected diseases. Rice Plant Disease (RPD) is one of the important factors that decrease the quantity and quality of rice. Identifying the type of rice plant disease and taking corrective action against the disease in time is always challenging for the farmers. Although the rice plant is affected by many diseases, Bacterial Leaf Blight (BLB), Brown Spot (BS), and Leaf Smut (LS) are major diseases. Identification of this disease is really challenging because the infected leaf has to be processed by the human eye. So in this paper, we focused on machine learning techniques to identify and classify the RPD. We have collected infected rice plant data from the UCI Machine Learning repository. The data set consists of 120 images of infected rice plants in which 40 images are BLB, 40 are BS, and 40 are LS. Experiments are conducted using Decision tree-based machine learning algorithms such as RandomForest, REPTree, and J48. In order to extract the numerical features from the infected images, we have used ColourLayoutFilter supported by WEKA. Experimental analysis is done using 65% data for training and 35% data for testing. The experiments unfold that the Random Forest algorithm is exceptional in predicting RPD.

Diagnosis of Leaf Surface Disease Using Two Datasets of Tomato and Rice Obtained from Image Processing Techniques

“…In this section, the proposed method is compared with other methods. The accuracy of the proposed method using data sets of rice has been compared with CNN methods [13], Decision Tree-based Machine Learning Algorithms [14] and improved random forest [15], Detection and Classification of Rice Plant Diseases [26], Deep Neural Network Algorithm [27], An Approach Using Textural Features [28], and tomato data set has been compared with Computer Vision Based Detection and Classification [16], Automatic Tomato Plant Leaf Disease and pests [17], Deep Neural Network-Based Tomato Plant Diseases [18],Tomato plant disease classification methods in digital images [29], Automatic Tomato Plant Leaf Disease [30].…”

“…It should be noted that the data sets of the methods compared for both rice and tomato diseases are the same as the proposed method, and therefore the experiments are the same in terms of data sets. 93.3% [14] 76.19% [15] 97.80% [26] 93.33% [27] 96% [28] 94.16% Proposed Method 99.12%…”

“…Rice disease was specifically addressed. Pictures of disease symptoms on leaves and stems are taken from rice fields [14]. A total of 120 images of sick rice plants have been collected from real farm conditions, according to the decision tree classification.…”