BerryPortraits: Phenotyping Of Ripening Traits in cranberry (Vaccinium macrocarpon Ait.) with YOLOv8
2024

BerryPortraits: Software for Analyzing Cranberry Ripening Traits

Sample size: 512 publication 20 minutes Evidence: high

Author Information

Author(s): Loarca Jenyne, Wiesner-Hanks Tyr, Lopez-Moreno Hector, Maule Andrew F., Liou Michael, Torres-Meraz Maria Alejandra, Diaz-Garcia Luis, Johnson-Cicalese Jennifer, Neyhart Jeffrey, Polashock James, Sideli Gina M., Strock Christopher F., Beil Craig T., Sheehan Moira J., Iorizzo Massimo, Atucha Amaya, Zalapa Juan

Primary Institution: University of Wisconsin-Madison

Hypothesis

BerryPortraits software can accurately measure berry color, size, and shape traits in cranberries.

Conclusion

BerryPortraits is an effective and open-source tool for high-throughput phenotyping of cranberry traits.

Supporting Evidence

  • BerryPortraits was trained on 512 postharvest images of diverse cranberry populations.
  • The software allows for rapid detection and segmentation of berries.
  • BerryPortraits can measure traits like berry color, size, and shape accurately.
  • Statistical analyses showed high correlation between BerryPortraits and previous software GiNA.
  • Imaging under controlled conditions improved measurement accuracy.
  • BerryPortraits is open-source and user-friendly, lowering barriers for researchers.
  • The software has potential applications for other berry crops.

Takeaway

BerryPortraits is a computer program that helps scientists quickly measure how cranberries look, like their color and size, using pictures.

Methodology

The study used a YOLOv8 model for image segmentation and analysis of cranberry traits from 512 images.

Potential Biases

Potential bias in color measurement due to variations in lighting and background.

Limitations

The software may struggle with very light berries against a white background and requires careful lighting for accurate color measurement.

Participant Demographics

The study involved a cranberry population developed at Rutgers University, consisting of 171 genotypes.

Statistical Information

P-Value

p<0.05

Confidence Interval

95%

Statistical Significance

p<0.05

Digital Object Identifier (DOI)

10.1186/s13007-024-01285-1

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication