Enhancing citrus fruit yield investigations through flight height optimization with UAV imaging

Area of Science:

• Agricultural Science
• Remote Sensing
• Horticulture

Background:

• Citrus fruit yield estimation is crucial for market stability but traditionally complex and labor-intensive.
• Accurate yield data enables better production planning and resource allocation in the citrus industry.

Purpose of the Study:

• To determine the optimal altitude for unmanned aerial vehicle (UAV) imaging for estimating Citrus unshiu fruit yield.
• To evaluate image processing techniques, such as histogram equalization, for improving yield estimation accuracy.

Main Methods:

• Collected images of Citrus unshiu at five different altitudes (30m, 50m, 70m, 90m, 110m).
• Assessed image resolution requirements for accurate fruit size estimation (approx. 5 pixels/cm for 46.7mm fruit diameter).
• Compared fruit count estimations from normal and histogram-equalized images, and analyzed vegetation indices (IPCA, I1).

Main Results:

• A resolution of approximately 5 pixels/cm is necessary for reliable fruit size estimation.
• Histogram equalization significantly improved fruit count accuracy compared to untreated images.
• Vegetation Index IPCA showed similar estimation to histogram equalization, while I1 had a notable gap to actual yields.

Conclusions:

• Optimal UAV imaging altitudes and image processing techniques can enhance the accuracy of citrus fruit yield estimation.
• UAV-based remote sensing offers a cost-effective and scalable solution for regional crop yield monitoring.
• This study provides a valuable dataset for future UAV applications in precision agriculture for citrus.