Sensor Fusion on a Wild Blueberry Harvester for Fruit Yield, Plant Height and Topographic Features Mapping to Improve Crop Productivity
1A. A. Farooque, 1Q. U. Zaman, 2D. Groulx, 3A. W. Schumann, 1T. J. Esau, 1Y. K. Chang
1. NSAC Canada
2. Dalhousie University, Canada
3. University of Florida, USA
Site-specific crop management can improve profitability and environmental risks of wild blueberry crop having large spatial variation in soil/plant characteristics, topographic features which may affect fruit yield. An integrated automated sensor fusion system including an ultrasonic sensor, a digital color camera, a slope sensor, a real-time kinematics GPS (RTK-GPS), custom software and ruggedized computer was developed. The system was incorporated into a commercial wild blueberry harvester, to measure plant height, fruit yield, slope and elevation in a simultaneous manner with harvest. Two wild blueberry fields were selected in central Nova Scotia of Canada to evaluate the performance of the system. The field boundary, bare spots, weeds and grass patches were mapped with a RTK-GPS prior to start the experiment.
Linear regression was used to calibrate the actual fruit yield with the percentage of blue pixels (R2 = 0.91 to 0.92; P < 0.001; n = 40) using 0.91 × 0. 70 m quadrates at selected points from both fields. The output voltage of ultrasonic sensor was significantly correlated with manually measured plant height (R2 = 0.95; P < 0.001; n = 40). Comprehensive surveys were conducted in both fields to measure plant height, fruit yield, slope and elevation rapidly in real time at the same time of harvest. The maps for the measured parameters were developed in ArcGIS 10 showed substantial variability in measured parameters across the fields, suggesting less fruit yield and more plant height in low lying areas and vice versa. A negative relationship among fruit yield and plant height (r = -0.20) also supported the results identified by the maps. Overall, the results of calibration and mapping indicated that the sensor fusion system was accurate, reliable and efficient to map plant height, fruit yield, slope and elevation in real time. This information can be used to develop site-specific nutrition program, based on easily measured field parameters, to optimize productivity while minimizing the environmental impacts.
