Uso de aeronave remotamente pilotada na análise qualitativa no plantio de cafeeiro

Abstract

Precision farming technologies contribute to multi-step monitoring of agricultural management. Sensors aboard remotely piloted aircraft (RPA) are systematically exploited to collect crop information and support decision making in the field. In coffee farms aerial remote sensing can contribute to observations about the quality of operations. The implementation phase in perennial crops is seen as one of the most important due to errors in this stage may compromise the next operations. Investigations related to the quality of planting in coffee crops may help to improve the crop implantation techniques. Therefore, the objective of this research was to evaluate the quality of manual planting and semi-mechanized planting techniques in high slope lands by statistical process control (SPC) of data remotely obtained by RPA. The experiments were conducted in a farm located at the municipality of Bom Sucesso – MG, in three coffee areas six months after their implantation. I these areas were performed a 30 meters height flights with an RPA and point tracking whith a pair of Global Navigation Satellite System (GNSS). In the first experiment, the tests were performed in an area implanted with a manual planting system and another area implanted by a semi-mechanized planting system. In order to evaluate

quality in alignment and distribution were used the SCP and density maps. The semi- mechanized system planted 7% less plants than the expected, while the manual system planted

4.7% above projected. From the SCP analyses were found in the semi-mechanized system points out of bounds, so with improvements this system can present better results. It was found that the manual system presents intrinsic variations in the process, preventing improvements in quality. In the second experiment, the quality of semi-mechanized planting system in regions with a slope between 0-40% was evaluated. In an area of 0.8 hectares, 18 points were traced and georeferenced for geometric corrections of the digital terrain model (DTM). In order to study the slope influence on the quality of planting it was proceeded analysis of variance and SCP. The averages tests showed low influence of slope on the alignment between plants in the line. From the SCP analyses were observed random variations in the spacing between planting lines in higher slopes. When evaluated spacing errors between planting lines it was observed that considerable errors occurred in regions with slope above 25%. The study allowed the knowledge about the effect of transplanting techniques on the quality of the coffee plant implantation and the influence of the slope on quality of semi-mechanized plantingsystem. The manual system despite presenting acceptable quality in the SCP analises presents intrinsic variations precluding improvements. The most relevant errors were found when the alignment between planting lines was evaluated. There is no significant interference in the alignment between plants in the planting line when the semi-mechanized system is performed in declivities between 0-40%.