Unmanned aerial system and satellite: Which one is a better platform for monitoring of the peanut crops?

• Published in: Agronomy journal Vol. 115; no. 3; pp. 1146 - 1160
• Main Authors: Costa Souza, Jarlyson Brunno, Luns Hatum de Almeida, Samira, Lopes de Brito Filho, Armando, Morlin Carneiro, Franciele, Santos, Adão Felipe dos, da Silva, Rouverson Pereira
• Format: Journal Article
• Language: English
• Published: 01.05.2023
• ISSN: 0002-1962; 1435-0645
• DOI: 10.1002/agj2.21310

Remote sensing tools are helpful in monitoring and managing crop production. However, each remote sensing technology responds to crop variability differently. In this way, the objective of this work was to compare sensors on airborne and orbital platforms and to observe which one has the best quality to determine the behavior of the peanut (Arachis hypogaea L.) crop variability. The experimental design followed the premises of the statistical quality control (SQC), with samples collected over time. The experimental area was composed of 30 sampling points spaced every 50 m. The multispectral images were acquired with an unmanned aerial system (UAS) consisting of a DJI Matrice quad‐copter and a Micasense RedEdge multispectral camera and with the PlanetScope multispectral imaging satellites. It was verified that in all periods evaluated for spectral bands and vegetation indices (VI), satellite images presented better process quality. The enhanced vegetation index (EVI) and normalized difference vegetation index (NDVI) generated from satellite images were able to detect the peanut maturation variation better. The behavior of the bands and the VIs generated from the Planet images show quality for peanut crop monitoring. While UAS showed sensitivity to detect the saturation of the bands, making it difficult to visualize the temporal variability. Core Ideas Monitoring of the peanut crops through remote sensing tools. Statistical quality control has satisfactory results in monitoring the quality of agricultural processes. Satellite images showed lower data variability in the behavior of bands and vegetation indices.