Algae Maturity Detection Using UAV-Based Multispectral Imaging

This paper presents a real-time algae maturity detection workflow utilizing UAV-based multispectral imaging for automated monitoring in large-scale algae farms. A MicaSense camera mounted on a UAV captures images across multiple spectral bands, though only the Blue (B), Green (G), and Red (R) bands are analyzed in this study to assess algae health and maturity. The workflow comprises tank detection using the Hough Transform, thresholding for foam segmentation, and average pixel value calculation within foam- free regions. The system achieved high precision 99.69% and recall (97.62% for B-band, 93.60% for G-band, and 95.40% for R-band), effectively identifying algae tanks and providing consistent, quantitative metrics. Misclassification cases, including false positives and false negatives, were analyzed to evaluate algorithm performance and reliability. This workflow enables efficient algae monitoring, delivering timely insights to optimize harvesting and enhance operational efficiency in commercial algae production.