Published on March 2021 | MACHINE LEARNING AND COMPUTER VISION

AGRICULTURAL ROBOT USING MACHINE LEARNING AND COMPUTER VISION

Abstract - Agriculture is the primary source of survival for about 60 percent of India’s population. Over 18 percent of India’s GDP is contributed by agriculture. But yearly there was about 7-8 percent of farmers quit agriculture. The reason behind this is a low profit, weather, manpower, and many more. No technology has evolved to help farmers. To overcome some of the problems, a robot was designed that is capable of seeding, planting, spraying weedicides & insecticides, and cultivation. The planting operation has a moisture sensor in it to check whether the soil is suitable for planting. The robot is built with an efficient sprayer that can spray as tall as the plant is and another nozzle for spraying insecticides. The weeds and insects are monitored with the help of computer vision and Machine Learning algorithms. The robot also has a robotic arm for picking the products from plants. The robotic arm is controlled with machine learning and computer vision. This robot is equipped with a Magnetometer and Global Positioning System (GPS) module that is capable of receiving GPS coordinates at higher accuracy so that it can maintain its path according to the path provided by the farmer. The robot will be smart in that it monitors the growth of weed and controls it with the help of a cutter at its bottom. Also, the robot is built with IoT capable of sharing the process going on with the farmer. Here the robot has a battery as a power source. It can be both electrically and solar charged. The robot that is designed are suitable for limited vegetable farming. The future work is to build a robot that can work in all crop fields. In the future, the robot will get updates on the mechanisms in the robot based on the performance.