Cotton harvester through the application of machine learning and image processing techniques

Abstract

Abstract Contributing 35% of global cotton production, India stands at the top. Harvesting is the most delicate process which requires clinically skilled labourers, but Statistics shows 10% decline in agricultural workers in every 3 years. The project deals with a cotton harvester which uses advanced digital analytics, robotics and image processing to complete the tasks of Identification, harvesting and storage. The machine would use multiple robotic arms constructed in the harvest zone to achieve multiple plucking of crops. powered electrically and controlled via remote the machine omits the need of a driver making it women friendly. The machine will be precise and diligent causing the least margin of errors. Studies show that 33 kg/ha of cotton remains on the plant un harvested after the harvesting is done and pickers are infected due to the contact and inhalation of defoliating agents. The harvest loss will be eradicated which increases the yield by 20–25% thereby increasing the farmers profit up to 20%.