Litcius/Paper detail

Weed Plant Detection from Agricultural Field Images using YOLOv3 Algorithm

Yukta Dandekar, Kshitija Shinde, Jai Gangan, Sabil Firdausi, Smita Bharne

20222022 6th International Conference On Computing, Communication, Control And Automation (ICCUBEA20 citationsDOI

Abstract

Agriculture is the only way for humans to survive in this world. Weed plant detection and classification are critical technical and economic issues in agriculture. Weed creates problems on the field as they extract good nutrients required by other crops. To encounter this, manual weed detection has recently been carried out with specialized people. Later, as technology advanced, people began using herbicides to kill weeds. People are trying to detect weeds without human intervention, but they were unable to reach the public due to a lack of precision. This paper focuses on weed plant detection techniques that can be used to supplement physical detection methods. We have attempted to use the CNN technique called YOLOV3 which is an object detection technique that will help us to accurately identify weed crops. We aim to counteract and provide a better solution for the existing problems to detect weeds in crops.

Topics & Concepts

WeedAgricultureComputer scienceWeed controlPrecision agricultureField (mathematics)Weed scienceObject detectionAgricultural engineeringArtificial intelligenceAgronomyMathematicsPattern recognition (psychology)EngineeringEcologyBiologyPure mathematicsSmart Agriculture and AIIoT-based Smart Home Systems