Detection and Providing Suggestion for Removal of Weeds Using Machine Learning Techniques

Abstract

Weed generally refers to any plant growing where it is not wanted. These unwanted plants are undesirable as they are aggressive in growth, consuming light, water, nutrients, and space that desirable crops utilize. It would affect production efficiency both in quality and quantity. For decades, weeds have been removed using chemicals that are not very effective. With the advancement in technology, this paper proposes an image processing-based framework and machine learning model for weed detection using Convolutional Neural Network (CNN) and the Xception model for classification. With increased layers, the efficiency and accuracy of the system are improved compared to existing methods.