Search

In this study, a lightweight phenotyping system that combined the advantages of both deep learning-based panicle detection and the photogrammetry based on light consumer-level UAVs was proposed. A two-year experiment was conducted to perform data collection and accuracy validation. A deep learning model, named Mask Region-based Convolutional Neural Network (Mask R-CNN), was trained to detect panicles in complex scenes of paddy fields. A total of 13 857 images were fed into Mask R-CNN, with 80% used for training and 20% used for validation. Scores, precision, recall, Average Precision (AP), and F1-score of the Mask R-CNN, were 82.46%, 80.60%, 79.46%, and 79.66%, respectively. A complete workflow was proposed to preprocess flight trajectories and remove repeated detection and noises. Eventually, the evident changed in rice growth during the heading stage was visualized with geographic distributions, and the total number of panicles was predicted before harvest. The average error of the predicted amounts of panicles was 33.98%. Experimental results showed the feasibility of using the developed system as the high-throughput phenotyping approach.

This research introduces a new inclination correction method with increased accuracy applied to the guidance system of an agricultural vehicle. The method considers the geometry of a robot tractor and uses an Inertial Measurement Unit (IMU) to correct the lateral error of the RTK-GPS antenna measurements raised by the tractor's inclinations. A parameters optimization experiment and an automatic guidance experiment under real working conditions were used to compare the accuracy of a traditional correction method with the new correction method, by calculating the RMSE of the lateral error and the error reduction percentage. An additional tuned correction method was found by using a simple analytical method to find the optimal variables that reduced the lateral error to a minimum. The results indicate that the new correction method and the tuned correction method display a significant error reduction percentage compared to the traditional correction method. The methods could correct the GPS lateral error caused by the roll inclinations in real-time. The resulting lateral deviation caused by the tractor's inclinations could be reduced up to 23% for typical travelling speeds.