A Novel Approach Of Pest Recognition By Analyzing Ensemble Modeling

Abstract

One of the significant problems in agriculture is the pest in the crop, which causes a considerable
loss to the farmer. Nowadays, farmers are using the same pesticide for all the pests. We recommend the
specific pesticide needed for the given pest. Pest recognition is the process of identifying the pest from the
provided photo and suggesting a suitable insecticide. For this, we use machine learning algorithms. Different
image processing algorithms are using to process the images, and multiple variations of these datasets are
trained on different Classifiers to determine which one is the most efficient in predicting the pests.
Ensembles of different sets of classifiers trained on different variations of the images are making to increase
the accuracy of the models. This work aims to recognize the best model dataset combination to achieve the
highest precision and performance.