Textbook in PDF format

This book provides a complete and comprehensive reference for agricultural mobile robots, covering all aspects of the design process, from sensing and perceiving to planning and acting for practical farming applications.
Mobile Robots for Digital Farming explores topics such as Robot Operating Systems (ROS), dynamic simulation, Artificial Intelligence (AI), image processing, and Machine Learning. Additionally, it features multiple case studies from funded projects and real-field trials.
Simulation environments such as Robot Operating System (ROS), MatLAB, and CoppeliaSim provide designers with a fexible approach for experimenting and implementing IBVS algorithms in order to optimize plant/fruit localization, improve plant/fruit scanning, and develop strategies for finding collision-free paths. Published studies demonstrate a completely simulated workspace environment, including a replica of the robot manipulator in Inverse Kinematic mode, object tracking, and the orchard or the field with bushes and plants. It should be noted that any visual servo system must be capable of tracking image features in a sequence of images. In an actual experiment with one or more RGB-D cameras, feature-based and correlation-based methods (as well as Artificial Intelligence and Deep Learning training methods) are used to improve the robot’s image classification for tracking. These studies use image data taken from the robot camera as the input to the IBVS control algorithm in MatLAB. The extracted information is then fed back to the simulated workspace for IK calculation and for determining the trajectory path to the fruit. In most cases, ROS is used for bi-directional communication between the simulated environment and the robot via its publish and subscribe architecture. The proposed approach allows researchers to review, approve, and execute different trajectories for placing the sensor in the most desired positions, as well as providing a fexible framework for evaluating different sense-think-act scenarios to verify the functional performance of future manipulators with zero risk to the robots and operators.
This book will be useful for professors and academics in various engineering disciplines (mechanical, robotics, control, electrical, computer, and agricultural), graduate and undergraduate students, farmers, commercial growers, startups, private companies, consultancy agencies, equipment suppliers, and agricultural policymakers