论文解读

AI 辅助整理

方法: 主动分离与近距成像装置、3077张6类叶龄图像、RGB与相对深度四通道早融合YO
发现: 掩膜精确率91.42%、召回率77.61%，较基线提升3.57个百分点。
机会点: 可探索多生育期、多品种泛化及轻量化部署，结合时序影像实现叶龄连续估计。

论文摘要

研制主动分离与近距成像装置，构建覆盖6个叶龄类别、3077张图像的数据集，提出RGB与相对深度四通道早融合YOLO-PFL模型。四次随机种子运行下掩膜精确率91.42%、召回率77.61%，较基线提升3.57个百分点。

全文

Agriculture (MDPI)

You are currently on the new version of our website. Access the old version here.

Journals
Topics

Our Topics
* All Topics
* About Topics
* Topics Awards
* Subject Editors
* Propose a Topic

By Subjects

*   [Biology & Life Sciences](https://www.mdpi.com/topics?facets=NobwRAlgJmBcYGcCuAjAVgUwMYBcFgBowA3AQwBskM4wBGQsc0lDcmgIQgHtyuBzAJ4ACAGRCAMhABmGIQGUsEDADssGfAF8AukA)
*   [Chemistry & Materials](https://www.mdpi.com/topics?facets=NobwRAlgJmBcYGcCuAjAVgUwMYBcFgBowA3AQwBskM4wBmQsc0lDcmgYQAsMBbCBHACcAngAIAZKICypHBkEQKCUQGUsEDADss1AL4BdIA)
*   [Computer Science & Mathematics](https://www.mdpi.com/topics?facets=NobwRAlgJmBcYGcCuAjAVgUwMYBcFgBowA3AQwBskM4wAWQsc0lDcmgYQHsBbAByRwYATgAIAylggYAdlgwiAZCICypHAAsM3NRCz4AvgF0gA)
*   [Engineering](https://www.mdpi.com/topics?facets=NobwRAlgJmBcYGcCuAjAVgUwMYBcFgBowA3AQwBskM4wBWQsc0lDcmgUQDsBzCTjDACc%2B3MAF8AukA)
*   [Environmental & Earth Sciences](https://www.mdpi.com/topics?facets=NobwRAlgJmBcYGcCuAjAVgUwMYBcFgBowA3AQwBskM4wA2Qsc0lDcmgUQDtiIAnAe04BbDJxwUABADIJ7UrxwALCQGUsEUVgz4AvgF0gA)

*   [Medicine & Pharmacology](https://www.mdpi.com/topics?facets=NobwRAlgJmBcYGcCuAjAVgUwMYBcFgBowA3AQwBskM4wB2Qsc0lDcmgWQygiwgDsMAAgBkggAoALUgCcAtqSwB7cooDmATzABfALpA)
*   [Public Health & Healthcare](https://www.mdpi.com/topics?facets=NobwRAlgJmBcYGcCuAjAVgUwMYBcFgBowA3AQwBskM4wBGABkLHNJQ3JoAVVyIsACABIYKOABb8AZEJHlxWUgCdqAXwC6QA)
*   [Social Sciences, Arts & Humanities](https://www.mdpi.com/topics?facets=NobwRAlgJmBcYGcCuAjAVgUwMYBcFgBowA3AQwBskM4wBOQsc0lDcmgZQHssIKACdjwwA7LBgQE%2BAQQBOePqWFQ%2BACSQBbRRBwRxYAL4BdIA)

Author Services
Information

General Content
* Open Access Policy
* Editorial Process
* Publication Ethics
* Special Issues Guidelines
* Article Processing Charge

Guidelines

*   [For Authors](https://www.mdpi.com/authors)
*   [For Reviewers](https://www.mdpi.com/reviewers)
*   [For Editors](https://www.mdpi.com/editors)
*   [For Librarians](https://www.mdpi.com/librarians)

Partnerships

*   [Institutional Open Access Program](https://www.mdpi.com/ioap)
*   [Societies](https://www.mdpi.com/societies)
*   [Conferences](https://www.mdpi.com/partnerships/conferences)
*   [Publishing Services](https://www.mdpi.com/publishing_services)

About

Company
* About Us
* Mission
* Impact
* History
* Leadership

*   [Office Locations](https://www.mdpi.com/about/office-locations)
*   [Careers](https://careers.mdpi.com/)
*   [Awards](https://www.mdpi.com/awards)
*   [Products](https://www.mdpi.com/about/products-solutions)
*   [Accessibility](https://www.mdpi.com/accessibility)

Media

*   [News](https://www.mdpi.com/news)
*   [Blog](https://blog.mdpi.com/)
*   [Press](https://www.mdpi.com/about/press)

Contact

*   [Support](https://www.mdpi.com/support)
*   [Send Feedback](https://www.mdpi.com/support/feedback)

This Journal

AdvancedToggle search box

Search Sign in

Submit

Open mobile navigation

Agriculture Agriculture

Cite Cite

Share Share

Get Alerted Get Alerted

Download PDF PDF

Abstract Article Metrics

This is an early access version, the complete PDF, HTML, and XML versions will be available soon.
Article

22 October 2026

36 Pages

Rice Seedling Leaf-Age Recognition Using Human-Inspired Active Separation for Imaging and YOLO-PFL

Haoyu Cui 1,†,

Di Jiao 2,†,

Jinyang Li 1,

Ziyi Zhu 3,

Bo Zhang 1,* and

Huayang Wang 4,*

College of Information and Electrical Engineering, Heilongjiang Bayi Agricultural University, Daqing 163319, China

Heilongjiang Academy of Land Reclamation Sciences, Harbin 150038, China

College of Agriculture, Heilongjiang Bayi Agricultural University, Daqing 163319, China

Beijing Simulation Center, Beijing 100854, China

Agriculture2026, 16(21), 2311;https://doi.org/10.3390/agriculture16212311

This article belongs to the Section Artificial Intelligence and Digital Agriculture

Version Notes

Order Reprints

Article Metrics

Article Views 37

Academic Editor

IN In Seop Na

In Seop Na

Publication History

Received: 9 September 2026
Revised: 14 October 2026
Accepted: 20 October 2026
Published: 22 October 2026

A Stage-Aware Cascaded Detection–Segmentation Framework for Leaf Phenotyping and Leaf Dry Biomass Estimation of Pepper Seedlings

Han Li et al. Plants, 20 June 2026

Using Channel and Network Layer Pruning Based on Deep Learning for Real-Time Detection of Ginger Images

Lifa Fang et al. Agriculture, 25 November 2021

Improved YOLOv8 Model for Phenotype Detection of Horticultural Seedling Growth Based on Digital Cousin

Yuhao Song et al. Agriculture, 26 December 2024

Abstract

Rice seedling leaf age is important for determining transplanting time, but manual assessment is laborious, and leaf overlap and subtle differences between adjacent stages hinder automated recognition. This study aimed to improve leaf-age recognition in dense rice seedling nurseries by integrating active seedling separation with RGB and relative depth information. A device for human-inspired active seedling separation and close-up imaging was developed, and a dataset comprising 3077 images across six leaf-age categories was constructed. The proposed YOLO-PFL model was based on YOLO26n-seg and used four-channel early fusion of RGB images and spatially aligned relative depth maps estimated using Depth Anything V2. CKConvC3k2, C2PSA_CirculantAtt, and DPCF were incorporated to enhance morphological feature extraction for slender leaves, spatial relationship modeling, and multiscale detail fusion, respectively. Across four runs with random seeds 42, 43, 44, and 45, YOLO-PFL achieved mean test-set instance-mask precision, recall, F1 score, mAP@0.5, and mAP@0.5:0.95 of 91.42%, 77.61%, 83.95%, 86.05%, and 54.42%, respectively. Compared with YOLO26n-seg using the same four-channel input, these mean values increased by 3.57, 2.13, 2.75, 4.09, and 2.53 percentage points, respectively. These findings support the feasibility of class-aware instance segmentation for rice seedling leaf-age recognition under the evaluated greenhouse and post-separation imaging conditions.

Keywords:

rice seedlings; greenhouse nursery monitoring; instance segmentation; monocular depth estimation; early fusion; close-range imaging