A novel autonomous weedicide spraying machine for row based crops for agricultural applications using embedded linux

Abstract An innovative approach to control the weeds by spraying the weedicide using ARM processor based single board credit card sized computer called Raspberry Pi with a selective spraying to reduce the wastage of chemicals was developed. The system was developed for precision spraying of weedicide using quantified information of the weeds from the image processing in Octave. This unit consists of a camera to capture the input image, Octave installed in a Raspberry Pi board and an autonomous robotic sprayer. The captured image was converted into gray scale then area thresholding was done to separate the crops and the weeds. The position of weed obtained from Octave was transferred to the microcontroller to control the sprayer movement by serial communication. Thus, the developed autonomous weeding system saves an amount of 48% of weedicide and provides an efficiency of 89%.