Análise do desenvolvimento do milho por meio do processamento de imagens obtidas a partir de aeronave remotamente pilotada.

Abstract

Agricultural monitoring is important to anticipate information about productivity or production and detect problems that may affect it. One way to follow the development of a crop is to monitor its growth and the changes that occur in plants throughout their phenological cycle. This information is commonly obtained from manual measurements that have limitations such as scale, workplace and execution delay. Here we present the data collection in the field using Remotely Piloted Aircraft (RPA), popularly known as Drone. Through a low-cost RGB remote sensor embedded in the equipment, photographs can be collected that, after being processed, generate digital files. These allow the analysis and generation of information in a similar way to manual surveys not only locally, but in a considerable geographic area. Aerophotogrammetric surveys were carried out at different dates of the maize phenological cycle using an RPA Mavic 2 Pro, equipped with the L1D-20c camera. The Agisoft Metashape program was used to process the photos, through which point clouds, Digital Terrain Model (DTM) and orthomosaics were generated. By calculating the difference between the point clouds and the MDT using the Compute cloud/mesh distance technique of the CloudCompare program, the estimated canopy heights were obtained. Using spectral transformations calculated in the Agisoft Metashape program, the soil/plant separation was obtained, which is necessary to analyze the crop without the influence of soil or straw in early stages. The point clouds were analyzed using descriptive statistics and the average values of vegetation indices provided information about the state of the crop. The techniques provided the quantitative and qualitative evaluation of the point clouds referring to 5 stages of maize development, allowing the estimation of plant growth and the monitoring of the dynamics of the phenological evolution of the canopy. The possibility of using low-cost cameras brings new opportunities for agricultural research and monitoring.