Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
March 18, 2026HorticulturaeOpen Access

Linking Cucumber Surface Color to Internal Hydration Level Using Deep Learning for Freshness Classification

View Full Paper
Ask AI
Bookmark
Share

Authors

ATAmin Taheri-GaravandTMTheodora MakrakiOAOmidali Akbarpour

Discussion

Loading...

Member takes

Overview

Demonstrates deep learning linking cucumber surface color to hydration levels, suggesting improved assessment methods.

Key Points

  • The aim is to develop a non-destructive method for determining cucumber freshness based on surface characteristics and internal hydration levels.
  • Collected 4160 RGB images of cucumbers alongside gravimetric measurement of relative water content (RWC).
  • Utilized a convolutional neural network (CNN) trained on standardized images to classify freshness categories.
  • Evaluated classification accuracy using an independent test set and Cohen's Kappa coefficient.
  • Conducted interpretability analyses to identify key surface features related to dehydration.
  • Achieved an overall classification accuracy of 91.35%.
  • Cohen’s Kappa coefficient was 0.875, indicating strong agreement with actual freshness categories.
  • F1-scores exceeded 0.94 for both Very Fresh and Spoiled categories, while intermediate classes showed greater overlap.

Cite This Study

Taheri-Garavand et al. (2026) studied this question.

synapsesocial.com/papers/69ba424e4e9516ffd37a26d6https://doi.org/10.3390/horticulturae12030357
View Full Paper
Ask AI
Bookmark
Share