Automated Weed Removal System using Machine Learning and Robotics: A Systematic Review

Abstract

As the demand for food grains continues to rise and resources continue to be limited, sustainable crop production has emerged as a major concern in modern agriculture. According to estimates, weeds in India lower crop output by 31.5%. Manual weed removal process is labor-intensive & time-consuming. It may include the extensive use of herbicides, which could be detrimental to the quality of the soil and water. The proposed work presents a comprehensive study on automated weed removal techniques by employing Machine Learning algorithms. The various types of weeds and their morphological features are discussed. The benchmark weed datasets are reported in the proposed study. The state-of-the-art approaches employed in automatic weed removal systems are presented in a well-defined manner. The issues and challenges in the field of automated weed removal are discussed.