A Survey on Weed Detection System Using Deep Learning

Abstract

In agriculture, weed is the major component in the field that affects crop production and crop quality. Therefore, it is essential to detect and classify weed in the field at its early growth stage. To avoid weed growth in the field farmers follow conventional techniques such as cultural, biological, and mechanical methods. Later on, as the technology has been improved the farmers started using chemical substances such as herbicides and pesticides to avoid weed growth and pests in the field. Farmers spray herbicides uniformly throughout the field which will also be sprinkled on the crops and the chemicals in the herbicides causes an effect on crop growth, crop quality, and crop production. The chemical substances that are present in herbicides are causing harm to crops so it’s necessary to spray herbicides specifically only on weeds and it results in achieving site-specific weed management. This paper focuses on the deep learning techniques which are used for a weed detection system that achieve site-specific weed management..