Leveraging ResNet-50 for Precision Toxicity Classification in Plants: A Vision-Based Approach to Safeguard Public Health

2024 · doi:10.32942/x28w4k
preprint OA: closed
Full text JSON View at publisher
⚙ AI-generated summary by claude@2026-07, 2026-07-05 ⓘ

This study developed a vision-based approach using ResNet-50 for precise classification of plant toxicity, aiming to enhance public health safety by identifying hazardous plants.

One-sentence paraphrase of the abstract; not a substitute for reading it. No clinical advice. How this works

⚙ AI-generated deep summary by claude@2026-07, 2026-07-05 · read from full text ⓘ

The paper investigates using a ResNet-50 deep learning model, with transfer learning, to classify plants as toxic versus non-toxic for public-safety applications in agriculture and food contexts. Using a vision-based approach, the authors report performance metrics of 89.6% accuracy, 87.4% precision, 91.1% recall, and an 89.2% F1 score, and they state that the method maintains high performance with limited data. The main limitation noted is that results may depend on the available dataset, with future work proposed to expand plant species coverage and test other models. The paper does not explicitly discuss endometriosis or adenomyosis; it was included in the corpus via a keyword match in the upstream search index.

Read from the paper's body, not the abstract. Not a substitute for reading the paper. No clinical advice. How this works

Abstract

The classification of toxic and non-toxic plants plays an important role in ensuring public safety, especially in agriculture, food safety, and health. Correct identification of these plants can prevent accidental poisoning and promote ecological protection. In this paper, we investigate the application of the ResNet-50 model for the classification of toxic and non-toxic plants. Leveraging the powerful feature extraction techniques of the ResNet-50 architecture, the model achieved 89.6% accuracy, 87.4% precision, 91.1% recall, and an 89.2% F1 score, demonstrating the model’s effectiveness. Transfer learning proved effec-tive with limited data while maintaining high performance metrics in the classification task. Future research could focus on expanding the dataset to include more plant species and exploring other state-of-the-art models to improve classification accuracy. Addition-ally, integrating these models with mobile applications or monitoring systems could pro-vide solutions for business and public use, enhancing environmental protection and pub-lic safety.
Full text 1,557 characters · extracted from oa-doi-fallback · 2 sections · click to expand

Abstract

The classification of toxic and non-toxic plants plays an important role in ensuring public safety, especially in agriculture, food safety, and health. Correct identification of these plants can prevent accidental poisoning and promote ecological protection. In this paper, we investigate the application of the ResNet-50 model for the classification of toxic and non-toxic plants. Leveraging the powerful feature extraction techniques of the ResNet-50 architecture, the model achieved 89.6% accuracy, 87.4% precision, 91.1% recall, and an 89.2% F1 score, demonstrating the model’s effectiveness. Transfer learning proved effec-tive with limited data while maintaining high performance metrics in the classification task. Future research could focus on expanding the dataset to include more plant species and exploring other state-of-the-art models to improve classification accuracy. Addition-ally, integrating these models with mobile applications or monitoring systems could pro-vide solutions for business and public use, enhancing environmental protection and pub-lic safety. DOI https://doi.org/10.32942/X28W4K Subjects Environmental Engineering, Plant Sciences

Keywords

Poisonous plants, Non-poisonous plants, ResNet-50, Plant classification, Deep learning, transfer learning, accuracy Dates Published: 2024-10-04 03:25 Last Updated: 2025-02-06 23:56 Older Versions License CC BY Attribution 4.0 International Additional Metadata Data and Code Availability Statement: Available on Kaggle Language: English

Text is read by the "Ask this paper" AI Q&A widget below. Extraction quality varies by source — PMC NXML preserves structure cleanly, OA-HTML may include some navigation residue, and OA-PDF can have broken hyphenation. The publisher copy (via DOI) is the canonical version.

My notes (saved in your browser only)

⚙ Ask this paper AI returns verbatim quotes from the full text · source: oa-doi-fallback ⓘ

Answers must be backed by verbatim quotes from this paper's full text. Hallucinated quotes are dropped automatically; if no verbatim passage answers the question, we say so. How this works

Citation neighborhood (no data yet)

We don't have any in-corpus citations linked to this paper yet. This is a recent paper (2024) — citers typically take a year or two to land, and the OpenAlex reference graph may still be filling in.

Source provenance

europepmc
last seen: 2026-05-20T01:45:00.602351+00:00