116. A human-borne sensor device to foster participatory sensing in precision orchard management

A novel wearable sensor device is proposed to ease the participation of farmers in data acquisition and labelling. This wearable device complies with artificial neural network approaches and can be used to develop supervised learning models based on the farmer’s cultural practices. This paper focuses on system design and development using open-source software and hardware. During field testing, the system GNSS-based location estimates showed an average error of 0.7±0.45 m and three fisheye cameras enable the system to image its surroundings at a maximum resolution of 2,592 by 1,939 pixels. Combined with the ability of the system to take user input, the experimental results demonstrate that it is a suitable tool for simultaneous data collection and labelling.