Intelligent Agricultural Machinery and Field Robots

Abstract

This chapter discusses classification of intelligent machines and presents examples of autonomous vehicles and field robots. It also discusses perception sensors and their selection for agricultural applications primarily in vehicle navigation and vehicle safeguarding. Monocular vision provides the best spatial resolution, which is very helpful in applying feature-based algorithms for object identification. Stereo vision and Lidar are often used in parallel with monocular vision through point cloud rendering to provide better object identification capability. The FroboMind architecture level consists of four modules, which are perception, decision making, action, and safety modules, all of which are encompassing the layered framework. For small field robots, a common limitation is the low work rate associated with them. Many technologies, as required for developing intelligent agricultural machinery and discussed in the previous sections, are still in early development stages.