Robots have played an increasingly important role in the process of agricultural operations. This technology depends on the robots ability to localize in its environment. This is generally accomplished by relying on global position signals such as GPS. However, some agricultural environments like orchards and vineyards can obstruct GPS signals and therefore require a different navigation aid. This work proposes a navigation framework that incorporates perception measurements to objects in the environment that follow an expected structure to maintain accurate localization in GPS-denied environments. The work uses measurements to tree trunks or vineyard posts as navigation aids to maintain accurate localization estimates.
展开▼
