Mid- to late-season weeds that escape from the routine early-season weed management threaten agricultural production by creating a large number of seeds for several future growing seasons. Rapid and accurate detection of weed patches in field is the first step of site-specific weed management. In this study, object detection-based convolutional neural network models were trained and evaluated over low-altitude unmanned aerial vehicle (UAV) imagery for mid- to late-season weed detection in soybean fields. The performance of two object detection models, Faster RCNN and the Single Shot Detector (SSD), were evaluated and compared in terms of weed detection performance using mean Intersection over Union (IoU) and inference speed. It was found that the Faster RCNN model with 200 box proposals had similar good weed detection performance to the SSD model in terms of precision, recall, f1 score, and IoU, as well as a similar inference time. The precision, recall, f1 score and IoU were 0.65, 0.68, 0.66 and 0.85 for Faster RCNN with 200 proposals, and 0.66, 0.68, 0.67 and 0.84 for SSD, respectively. However, the optimal confidence threshold of the SSD model was found to be much lower than that of the Faster RCNN model, which indicated that SSD might have lower generalization performance than Faster RCNN for mid- to late-season weed detection in soybean fields using UAV imagery. The performance of the object detection model was also compared with patch-based CNN model. The Faster RCNN model yielded a better weed detection performance than the patch-based CNN with and without overlap. The inference time of Faster RCNN was similar to patch-based CNN without overlap, but significantly less than patch-based CNN with overlap. Hence, Faster RCNN was found to be the best model in terms of weed detection performance and inference time among the different models compared in this study. This work is important in understanding the potential and identifying the algorithms for an on-farm, near real-time weed detection and management.

An electronic, resistance-based sensor array and data acquisition system was developed to measure spray deposition from hydraulic nozzles. The sensor surface consisted of several parallel tin plated copper traces of varying widths with varying gap widths. The system contained an embedded microprocessor to monitor output voltage corresponding to spray deposition every second. In addition, a wireless module was used to transmit the voltage values to a remote laptop. Tests were conducted in two stages to evaluate the performance of the sensor array in an attempt to quantify the spray deposition. Initial tests utilized manual droplet placement on the sensor surface to determine the effects of temperature and droplet size on voltage output. Secondary testing utilized a spray chamber to pass nozzles at different speeds above the sensor surface to determine if output varied based on different application rates or spray droplet classification. Results from this preliminary analysis indicated that manual droplets of 5 and 10 μL resulted in significantly different values from the sensors while temperature did not consistently affect output. Spray chamber test results indicated that different application rates and droplet sizes could be determined using the sensor array.

: With the mechanization of agricultural operations, agricultural machinery management has become an extensive research field. Sizing tractors and implements to provide the most efficient power transfer has become an ongoing process with advances in technology. Utilization of the rotational power transferred through gear trains from the tractor engine to the power take-off (PTO) shaft is one of the most efficient methods of power transfer to an implement. This research used commercially available torque sensors that were installed on a tractor PTO shaft for measuring the torque delivered to an implement. The torque sensor was calibrated using the Nebraska Tractor Test Lab’s (NTTL) dynamometer by following the Organisation for Economic Co-operation and Development (OECD) Code 2 test procedure for varying PTO loads. The calibration of the sensor was verified using the full load at varying speeds test as described in the OECD Code 2. Tractor PTO shaft torque values measured by the torque sensor were compared to the NTTL’s dynamometer torque measurement. Differences in torque values measured between the sensor and the dynamometer ranged from 3 to 23 N·m. Student’s t-test showed no significant difference between the measurements during the full load varying speed tests which demonstrated that the sensor can be mounted on the tractor’s PTO shaft for torque data collection in field operations. Keywords: Data Acquisition, LabVIEW, Power Take-off, Torque, Tractor.