Disease Prediction in Plants: using new Era Technologies

A country’s growth rate highly depends on the Agricultural sector. If the agricultural sector fails to meet the rising demand for food products, then it will negatively affect the growth rate of the economy. To meet the rising demand, the healthiness of plants is a must that can be secured by detecting disease in plants. In the past times, plant pathologists or agriculture professionals used to detect plant leaf disease by naked eye inspection. But this naked-eye inspection consumes a lot of time and is also expensive for most small farmers. Disease in plants can be detected experimentally by software developed on Machine Learning (ML) and Deep Learning (DL). The author team used various ML algorithms such as Logistic Regression, K Nearest Neighbors, Random Forest, Naïve Bayes, and Support Vector Machine for detecting the disease in the leaves of plants. The dataset includes images of tomatoes, potatoes, apples, Corn, Grapes, etc. The Random Forest algorithm has the highest accuracy of 95 percent for the dataset used. The accuracy of other algorithms is as follows: LR-90 percent, KNN-92 percent, RF-95 percent, NB-86 percent, and SVM-77 percent. So it can be concluded that the RF algorithm is best suited for this particular type of dataset.