Using self-propelled sprayers for the targeted application of herbicides

The presented paper deals with the targeted application of pre-emergent herbicides. The field experiment was established in 2022 and 2023; two plots sized 26 ha and 30 ha were selected and the accuracy of the application was tested. There were local outbreaks of weeds on selected plots of land, which had to be eliminated before sowing the main crop - soybeans. Both plots were monitored using an unmanned aerial vehicle (UAV) � Mavic 3M with the multispectral camera. After the end of the monitoring, spectral analysis of images taken by UAV was made using the Pix4D software and vegetation indexes were calculated. The weed outbreaks were identified based on multispectral maps using the MagicTooll algorithm. The final application map was prepared in the QGIS program - two zones with and without herbicide application were defined. The application was carried out by the JD 4150i machine (9 sections per 30 m), the regulatory map was transferred to the machine terminal (4600) via the MyJD link. Subsequently, the entire application process was monitored and analysed using the MyJD programme. The accuracy of the application was analysed by comparing the prescription map and the actual dose applied. The accuracy of application was found to be significantly greater than 90%.