Unmanned aerial vehicle-based remote sensing technology has recently been widely applied to crop monitoring due to the rapid development of unmanned aerial vehicles, and these technologies have considerable potential in smart agriculture applications. Field phenotyping using remote sensing is mostly performed using unmanned aerial vehicles equipped with RGB cameras or multispectral cameras. For accurate field phenotyping for precision agriculture, images taken from multiple perspectives need to be simultaneously collected, and phenotypic measurement errors may occur due to the movement of the drone and plants during flight. In this study, to minimize measurement error and improve the digital surface model, we proposed a collaborative driving system that allows multiple UAVs to simultaneously acquire images from different viewpoints. An integrated navigation system based on MAVSDK is configured for the attitude control and position control of unmanned aerial vehicles. Based on the leader–follower-based swarm driving algorithm and a long-range wireless network system, the follower drone cooperates with the leader drone to maintain a constant speed, direction, and image overlap ratio, and to maintain a rank to improve their phenotyping. A collision avoidance algorithm was developed because different UAVs can collide due to external disturbance (wind) when driving in groups while maintaining a rank. To verify and optimize the flight algorithm developed in this study in a virtual environment, a GAZEBO-based simulation environment was established. Based on the algorithm that has been verified and optimized in the previous simulation environment, some unmanned aerial vehicles were flown in the same flight path in a real field, and the simulation and the real field were compared. As a result of the comparative experiment, the simulated flight accuracy (RMSE) was 0.36 m and the actual field flight accuracy was 0.46 m, showing flight accuracy like that of a commercial program.
High-precision position estimations of agricultural mobile robots (AMRs) are crucial for implementing control instructions. Although the global navigation satellite system (GNSS) and real-time kinematic GNSS (RTK-GNSS) provide high-precision positioning, the AMR accuracy decreases when the signals interfere with buildings or trees. An improved position estimation algorithm based on multisensor fusion and autoencoder neural network is proposed. The multisensor, RTK-GNSS, inertial-measurement-unit, and dual-rotary-encoder data are fused with Extended Kalman filter (EKF). To optimize the EKF noise matrix, the autoencoder and radial basis function (ARBF) neural network was used for modeling the state equation noise and EKF measurement equation. A multisensor AMR test platform was constructed for static experiments to estimate the circular error probability and twice-the-distance root-mean-squared criteria. Dynamic experiments were conducted on road, grass, and field environments. To validate the robustness of the proposed algorithm, abnormal working conditions of the sensors were tested on the road. The results showed that the positioning estimation accuracy was improved compared to the RTK-GNSS in all three environments. When the RTK-GNSS signal experienced interference or rotary encoders failed, the system could still improve the position estimation accuracy. The proposed system and optimization algorithm are thus significant for improving AMR position prediction performance.