PulseExploreJournal ClubResearchersJournals
Instagram
HomeJournal ClubExplore
Synapse
⌘+K
Synapse
May 7, 2026AgronomyOpen Access

Tomato Ripeness Detection Model Based on Improved RT-DETR Lightweight Model

View Full Paper
Ask AI
Bookmark
Share

Authors

GYGuoliang YangDWDali WengZLZhiteng Li

Discussion

Loading...

Member takes

Overview

This technology improves tomato ripeness detection in greenhouses, suggesting enhanced automated harvesting practices.

Key Points

  • Develop a lightweight model for accurate tomato ripeness detection in challenging greenhouse conditions.
  • Proposed CFD-DETR model based on RT-DETR architecture.
  • Incorporated CAEfficientViT for multi-scale feature extraction.
  • Implemented FEAA mechanism for local ripening trait capture.
  • Utilized DwDySample for semantic feature preservation.
  • Designed Wise-SIoU loss function to enhance robustness against noise.
  • Achieved 90.2% mAP@0.5, outperforming baseline by 2.1 percentage points.
  • Reduced parameter count by 47.2% and computational complexity by 52.5%.
  • Confirmed superior generalization on Laboro Tomato and RaUTD datasets.

Cite This Study

Yang et al. (2026) studied this question.

synapsesocial.com/papers/69fbe2b3164b5133a91a21c8https://doi.org/10.3390/agronomy16090932
View Full Paper
Ask AI
Bookmark
Share