Logo PTI Logo FedCSIS

Proceedings of the 21st Conference on Computer Science and Intelligence Systems (FedCSIS)

Annals of Computer Science and Information Systems, Volume 47

Multispectral Analysis of Olive Fruits from UAV-based imagery

, , ,

DOI: http://dx.doi.org/10.15439/2026F0774

Citation: Ana Šarić Gudelj, , ,

Full text

Abstract. In this work, an automated approach for analysing olive fruit ripening using multispectral UAV imagery at the level of individual fruits is presented. RGB and multispectral images were collected from July to October using a DJI Mavic 3 Multispectral unmanned aerial vehicle. Fruits were first detected in RGB images using a YOLOv11 model finetuned for olive fruit detection, after which the corresponding bounding boxes were geometrically mapped to the G, R, RE, and NIR multi- spectral channels using calibration and homographic parameters extracted from the image metadata. After patch extraction, a filtering procedure was applied to retain representative and uniformly illuminated samples suitable for spectral analysis. NDVI and NDRE vegetation indices, as well as reflectance values in the NIR and RE channels, were analysed over time and separately for the cultivars Istarska bjelica (Croatian olive cultivar) and Leccino. The results showed statistically significant temporal differences for all analysed indicators, with NDRE exhibiting greater sensitivity to ripening-related changes than NDVI. Differences in the dynamics of change were also observed between the studied cultivars, indicating cultivar-specific ripen- ing patterns. The relationship between NDRE values and the laboratory-determined oil content suggests that NDRE reflects physiological changes associated with the ripening process rather than oil accumulation directly. The obtained results confirm the potential of combining deep learning and multispectral UAV imaging for non-destructive monitoring of olive fruit ripening in precision agriculture applications.

References

  1. C. Mena, A. Z. González, R. Olivero-David, and M. Brenes, “Characterization of ‘castellana’ virgin olive oils with regard to olive ripening,” HortTechnology, vol. 28, no. 1, pp. 48–57, 2018.
  2. C. Skodra, V. S. Titeli, and M. Michailidis, “Olive fruit development and ripening: Break on through to the ‘-omics’ side,” International Journal of Molecular Sciences, vol. 22, no. 11, p. 5806, 2021.
  3. M. Á. Delgado Noguera, B. Millán, A. Aquino, J. A. Diago, and J. Tardaguila, “Methodology for olive fruit quality assessment by means of a low-cost multispectral device,” Agronomy, vol. 12, no. 5, p. 979, 2022.
  4. M. Roca and M. I. Mínguez-Mosquera, “Changes in chloroplast pigments of olive varieties during fruit ripening,” Journal of Agricultural and Food Chemistry, vol. 49, no. 2, pp. 832–839, 2001.
  5. A. I. de Castro, Y. Shi, J. M. Maja, and C. Peña, “Uavs for vegetation monitoring: Overview and recent scientific contributions,” Remote Sensing, vol. 13, no. 11, p. 2139, 2021.
  6. J. M. Jurado, L. Ortega, J. J. Cubillas, and F. R. Feito, “Multispectral mapping on 3d models and multi-temporal monitoring for individual characterization of olive trees,” Remote Sensing, vol. 12, no. 7, p. 1106, 2020.
  7. G. Caruso, G. Palai, F. P. Marra et al., “High-resolution uav imagery for field olive (olea europaea l.) phenotyping,” Horticulturae, vol. 7, no. 8, p. 258, 2021.
  8. A. Carella, R. Massenti, F. P. Marra et al., “Combining proximal and remote sensing to assess ‘calatina’ olive water status,” Frontiers in Plant Science, vol. 15, 2024.
  9. G. Agati, P. Pinelli, and S. Cortés Ebner, “Nondestructive evaluation of anthocyanins in olive (olea europaea) fruits by in situ chlorophyll fluorescence spectroscopy,” Journal of Agricultural and Food Chemistry, vol. 53, no. 5, pp. 1354–1363, 2005.
  10. J. Ma, M. Li, W. Fan, and J. Liu, “State-of-the-art techniques for fruit maturity detection,” Agronomy, vol. 14, no. 12, p. 2783, Nov. 2024.
  11. J. Dell’Amico, A. Moriana, M. Corell, I. F. Girón, D. Morales, A. Torrecillas, and F. Moreno, “Low water stress conditions in table olive trees (olea europaea l.) during pit hardening produced a different response of fruit and leaf water relations,” Agricultural Water Management, vol. 114, pp. 11–17, Nov. 2012.
  12. Agroklub, “Istarska bjelica je najranija sorta masline za berbu u hrvatskoj,” 2026, accessed: 2026-0409. [Online]. Available: https://www.agroklub.com/vocarstvo/ istarska-bjelica-je-najranija-sorta-masline-za-berbu-u-hrvatskoj/63601/
  13. M. Distefano et al., “Vegetation indices for predicting ripeningassociated changes in chlorophyll and polyphenol content: A multicultivar assessment in olive germplasm,” Remote Sensing, vol. 18, no. 2, 2026.