Blueberry Yield Analyzer is a software tool that estimates blueberry yield from field images. Using machine learning and vision-based deep learning models, this tool detects ripe and unripe berries in a field image. In a second step, the tool estimates per-plant yield based on detected berry count and user-provided average berry weight. Currently, the berry detection model is integrated in GDV to count ripe and unripe berries from a single-plant image. The yield prediction model will be available soon. The tool was developed collaboratively by Dr. Yin Bao's lab at the University of Delaware and Dr. Sushan Ru's lab at Auburn University.

Please upload a clear, side-view image of a single, mature blueberry plant. Original photos taken with a smartphone or camera are ideal.

Reference:
Puranjit Singh, Nariman Niknejad, James D Spiers, Yin Bao, Sushan Ru (2025) Development of a Smartphone Application for Blueberry Detection Towards Rapid Yield Estimation. Smart Agricultural Technology 101361. doi.org/10.1016/j.atech.2025.101361