An implement able to perform physical weed control in maize was developed at the University of Pisa within the RHEA Project. The implement is an automatized operative machine able to perform mechanical non-selective treatments in the inter-row and site-specific flaming in the row space. Flaming system is basically composed by a LPG feeding system, couples of burners and an instantaneous ignition system. When the machine is coupled with the GMU developed within the RHEA project the system becomes autonomous. The application of flaming is based on a camera vision weed detection system installed on the GMU and developed by the University Complutense of Madrid. When the vision system detects weeds the signal is send to the HLDMS installed on the GMU and then to the PLC mounted on the implement that manages flaming. Three levels of weed cover (zero, low and high weed cover) were preestablished in order to perform no-treatment or treatments with low or high LPG pressure. According to the percentage of weed cover, the choice between two levels of LPG pressure is made in real-time in order to obtain the proper dose. The machine was realized to treat independently four rows of maize at each passage of the GMU. The weed detection system (WDS) is composed by a camera mounted at the top of the GMU pointed to the crop rows. Images are taken as the GMU moves along the rows. Every image is processed in real-time in order to detect, locate and quantify weed cover by means of an evaluation of the green percentage. This percentage is used by the HLDMS to decide whether to start flaming treatment or not and, if yes, adjust LPG pressure. These processes are done in real-time. The synchronization between the detection of the weeds and the flaming treatment is critical for the success of the whole project.
Calibration and synchronization between WDS and flaming system within the RHEA project
FRASCONI, CHRISTIAN;MARTELLONI, LUISA;FONTANELLI, MARCO;RAFFAELLI, MICHELE;PERUZZI, ANDREA
2014-01-01
Abstract
An implement able to perform physical weed control in maize was developed at the University of Pisa within the RHEA Project. The implement is an automatized operative machine able to perform mechanical non-selective treatments in the inter-row and site-specific flaming in the row space. Flaming system is basically composed by a LPG feeding system, couples of burners and an instantaneous ignition system. When the machine is coupled with the GMU developed within the RHEA project the system becomes autonomous. The application of flaming is based on a camera vision weed detection system installed on the GMU and developed by the University Complutense of Madrid. When the vision system detects weeds the signal is send to the HLDMS installed on the GMU and then to the PLC mounted on the implement that manages flaming. Three levels of weed cover (zero, low and high weed cover) were preestablished in order to perform no-treatment or treatments with low or high LPG pressure. According to the percentage of weed cover, the choice between two levels of LPG pressure is made in real-time in order to obtain the proper dose. The machine was realized to treat independently four rows of maize at each passage of the GMU. The weed detection system (WDS) is composed by a camera mounted at the top of the GMU pointed to the crop rows. Images are taken as the GMU moves along the rows. Every image is processed in real-time in order to detect, locate and quantify weed cover by means of an evaluation of the green percentage. This percentage is used by the HLDMS to decide whether to start flaming treatment or not and, if yes, adjust LPG pressure. These processes are done in real-time. The synchronization between the detection of the weeds and the flaming treatment is critical for the success of the whole project.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
