Urban Lawn Monitoring in Smart City Environments

Abstract

Control over water usage for irrigation purposes is a key factor in order to achieve the sustainability in agriculture. The irrigation of urban lawns represents a high percentage of the urban water usage. The use of information and communication technology (ICT) offers the possibility of monitoring the grass state in order to adjust the irrigation regime. In this paper, we propose an Arduinobased system with a camera set on a drone. The drone flies along the garden taking pictures of the grass. Those pictures are processed with a rule-based algorithm that classifies them according to the grass quality. Pictures can be tagged in three categories: high coverage, low coverage, or very low coverage. After designing our algorithm, twelve pictures are used to verify its correct operation. The results show a 100% hit rate. To analyze the suitability of using drones to perform this task, we carried out a comparative study for gardens with different sizes, where the drone and a similar system mounted on a small autonomous vehicle have been used. The results show that, for gardens bigger than 1000m2, the use of drone is needed due to the time consumed by the vehicle to cover the entire surface. Finally, we show the results of sending the image information after processing it in different manners.