Development of Agribot System for Weed Detection and Automatic Spraying of Herbicide

Agriculture plays an important role in Indian economy. Technology helps in increasing the production of food. However the production of food can be affected by number of factor such as climatic change, diseases, soil fertility etc. Weeds have the potential to result in 33 percent of crop loss. Economic loss of about USD 11 billion was estimated to weeds alone in 10 major crops of India. Weeds increases the cost of agriculture and hinders the progress of work and damages crops by affecting their growth. Weed control is an important aspect of crop management, as failure to adequately control weeds it leads to reduce yields and product quality. Automatic detection of Weed proves benefits in monitoring large fields of crops. The proposed work consists of modules namely Weed detection module to detect the weed and automate the agribot to spray the herbicides, Soil moisture detection module to automate the water supply based on soil moisture, Robot movement module to control the agribot movements based on the presence of weeds. Tools and technologies used to implement the proposed work are arduino IDE, python IDLE, Embedded C. YoloV5 is a framework where CNN algorithm is used in it. The outcome of the proposed project is to help the farmers through an agribot which detects the weeds in the field by image processing and sprays the herbicide to prevent the damages for crops. Water supply is automated based on soil moisture. It will be very useful to increase the production of food and also reduces the man power.