An Autonomous Pesticide Sprayer Robot with a Color-based Vision System

Abstract

The usage of robots has been due to cost reduction and increasing operation accuracy. Furthermore, employing robots in farming can decrease human tasks in hard and dangerous duties such as plowing, spraying pesticides, etc. Spraying chemicals is a common operation in agriculture crop protection. This operation is essential, but it can create some problems such as human and environmental damages by overdosing using pesticides. Recently, researchers focused on precision agriculture to make this smart. Sensors are employed to detect leaves of plants on the ground and spray them as much as required. Thus, pesticide dose will be under control. The current paper aims to introduce a wheeled robot that is developed to detect plants by color sensor and spray them. This robot can move between planting rows and detect weeds based on the leave color. A microcontroller-based board was used as the main controller, which sends spray commands to the sprayer nozzle. Outdoor and indoor tests were carried out to study the accuracy of this system. Results of experiments showed that this robot could work with acceptable accuracy in identifying weeds in the field. Thus, this robot can be commercialized for applying in the field to spray pesticides.