DefoliationCAM: An Edge-AI System for Real-Time Insect Defoliation Severity Assessment in Grapevines
Quick Answer
This paper shows that DefoliationCAM (D-CAM) is a solar-powered edge AI system utilizing YOLOX-Nano and MobileNet for real-time insect defoliation assessment in grapevines, achieving a macro-F1 score of 0.838.
Quick Take
The system processes 4K images in ~1.5 seconds, enabling timely pest management decisions with a 29% action threshold for intervention.
Key Points
- D-CAM uses YOLOX-Nano for 4K image localization and MobileNet for classification.
- Class imbalance reduced from 40:1 to 8.8:1 by merging damage classes.
- Synthetic data augmentation improved macro-F1 score from 0.768 to 0.838.
- The system processes images in ~1.5 seconds with a full cycle of 57-60 seconds.
- Field tests demonstrated feasibility of autonomous operation and remote monitoring.
DeepSignal Analysis
What happened
DefoliationCAM (D-CAM) is a solar-powered AI system designed for assessing insect-induced defoliation in grapevines. It utilizes YOLOX-Nano for localization and MobileNet for classification, achieving a macro-F1 score of 0.838. The system processes 4K images in about 1.5 seconds, facilitating timely pest management decisions.
Key evidence
- D-CAM processes 4K images in approximately 1.5 seconds, allowing for rapid assessment of defoliation severity.
- The system achieved a macro-F1 score of 0.838, outperforming ResNet-50 while using significantly fewer parameters (1.08 M compared to 23.52 M).
- A five-week deployment of two field units demonstrated the feasibility of autonomous operation and remote monitoring via LoRa communication.
Why it matters
The introduction of D-CAM addresses the challenge of subjective manual scouting for defoliation assessment, which can lead to inconsistent pest management decisions. By providing objective, real-time data, the system enables more effective interventions against pests, potentially improving grapevine health and yield. This technology could influence broader agricultural practices by integrating AI into pest management.
Paper Resources
📖 Reader Mode
~2 min readAbstract:Insect-induced defoliation threatens grapevine health and yield, yet severity assessment relies on subjective manual scouting. We developed DefoliationCAM (D-CAM), a solar-powered edge AI system that estimates canopy defoliation and transmits results via LoRa to a web dashboard. The pipeline uses YOLOX-Nano to localize grapevine cordons in 4K images and a MobileNet-based classifier with lightweight channel attention to classify 320 x 320 pixel patches. The 29,674 field-collected patches were labeled as background or one of four damage levels: no to negligible, mild, moderate, and severe. Merging moderate and severe classes reduced the class imbalance ratio from approximately 40:1 to 8.8:1. Synthetic augmentation increased macro-F1 from 0.768 to 0.838. The classifier outperformed ResNet-50 in macro-F1 with fewer parameters (1.08 M versus 23.52 M). Patch predictions were aggregated into a canopy-level Non-linear Global Defoliation Severity (NGDS) score and validated against an entomologist's ratings of 86 held-out canopy images. Quadratic discriminant analysis identified boundaries of 9% and 29% separating negligible damage, mild damage requiring monitoring, and action-level damage warranting on-site scouting and insecticide treatment consideration. The 29% boundary aligned with the 25%-30% action threshold range for Japanese beetle management in grapevine. Using TensorRT FP16 on an NVIDIA Jetson Orin Nano, the pipeline processed each 4K image in approximately 1.5 s, with a complete operating cycle of 57-60 s. A five-week deployment of two field units demonstrated the feasibility of autonomous solar-powered operation, LoRa communication, and remote defoliation monitoring. D-CAM supports frequent, objective canopy damage assessment for targeted scouting and timely pest-management decisions.
| Subjects: | Computer Vision and Pattern Recognition (cs.CV) |
| Cite as: | arXiv:2610.20879 [cs.CV] |
| (or arXiv:2610.20879v1 [cs.CV] for this version) | |
| https://doi.org/10.48550/arXiv.2610.20879 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Wei-Zhen Liang [view email]
[v1]
Wed, 14 Oct 2026 20:12:17 UTC (48,858 KB)
— Originally published at arxiv.org
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.
More from arXiv cs.CV
See more →ProMoE-FL: Prototype-conditioned Mixture of Experts for Multimodal Federated Learning with Missing Modalities
ProMoE-FL introduces a Prototype-conditioned Mixture-of-Experts framework for multimodal federated learning, effectively addressing missing modalities. It outperforms existing methods on four chest X-ray datasets, demonstrating superior feature synthesis capabilities in both homogeneous and heterogeneous settings.