Intensive and precise agriculture is one of the fields where autonomous driving found its first real applications. Navigating an agricultural tractor in an open field is relatively easily done with Global Navigation Satellite Systems. Navigating orchards and vineyards is however more challenging as the vehicle needs to position itself not only with respect to a global map but also to the crops. This paper designs a positioning system for (semi-) autonomous navigation of agricultural tractors in vineyards that uses standard ultra-sonic automotive sensors as the main sensors. The goal of the system is to estimate the distance from the left vineyard row and the incidence angle. The paper shows that a standard Extended Kalman Filter (EKF) cannot estimate the position of the vehicle accurately enough because of holes in the vegetation or branches. We solve these issues by using an augmented model for the EKF and by introducing a data-selection stage that discards measurements that do not fall on the tracked row. An extensive experimental campaign exemplifies the main features of the estimation algorithm and provides a validation of the method. In particular, we show that the Root Mean Square Error (RMSE) of the distance is 16 cm, while the angular RMSE is 2.6 degrees.
Cost-effective Tractor Localization for Semi-Autonomous Vineyard Operations
Furioli S.;Onesto L.;Corno M.;Cesana P.;Savaresi S.
2020-01-01
Abstract
Intensive and precise agriculture is one of the fields where autonomous driving found its first real applications. Navigating an agricultural tractor in an open field is relatively easily done with Global Navigation Satellite Systems. Navigating orchards and vineyards is however more challenging as the vehicle needs to position itself not only with respect to a global map but also to the crops. This paper designs a positioning system for (semi-) autonomous navigation of agricultural tractors in vineyards that uses standard ultra-sonic automotive sensors as the main sensors. The goal of the system is to estimate the distance from the left vineyard row and the incidence angle. The paper shows that a standard Extended Kalman Filter (EKF) cannot estimate the position of the vehicle accurately enough because of holes in the vegetation or branches. We solve these issues by using an augmented model for the EKF and by introducing a data-selection stage that discards measurements that do not fall on the tracked row. An extensive experimental campaign exemplifies the main features of the estimation algorithm and provides a validation of the method. In particular, we show that the Root Mean Square Error (RMSE) of the distance is 16 cm, while the angular RMSE is 2.6 degrees.File | Dimensione | Formato | |
---|---|---|---|
[furioli] Cost effective Tractor Localization for Semi Autonomous Vineyard Operations.pdf
Accesso riservato
:
Publisher’s version
Dimensione
1.66 MB
Formato
Adobe PDF
|
1.66 MB | Adobe PDF | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.