一种基于VMamba的小麦生长期检测模型研究

朱先远 ,  李秀茹

齐鲁工业大学学报 ›› 2026, Vol. 40 ›› Issue (4) : 52 -61.

PDF (10771KB)
齐鲁工业大学学报 ›› 2026, Vol. 40 ›› Issue (4) : 52 -61. DOI: 10.16442/j.cnki.qlgydxxb.2026.04.006
机电与信息工程

一种基于VMamba的小麦生长期检测模型研究

作者信息 +

Research on wheat growth stage detection model based on VMamba

Author information +
文章历史 +
PDF (11029K)

摘要

小麦是一种重要的粮食作物,对其生长期等生长信息准确检测是保障品质和产量的重要基础。近年来,计算机视觉等人工智能技术在农业领域得到广泛应用,但传统卷积神经网络(CNN)在复杂光照、背景变化及长时序特征提取方面存在局限。本文以小麦的“发芽”、“分蘖”、“抽穗”和“成熟”4 个生长期为研究对象,将新型状态空间建模网络 VMamba 引入小麦生长期检测任务中,构建了一种基于 VMamba 的小麦生长期检测模型,并与 ResNet50、MobileNetV3 和 ViT 等主流模型进行对比实验。实验结果表明,VMamba 模型在准确率、F1 值指标上均优于对比模型,且推理速度、参数量方面也具有优势,整体表现出更优的分类检测性能。此外,通过在 Nvidia Jetson Orin 平台实际部署验证了 VMamba 模型在农业视觉检测的可行性与有效性,为智能化作物监测提供了一种实现思路。

Abstract

Wheat is an important cereal crop, and accurate detection of its growth period and other growth information is an important foundation for ensuring quality and yield. In recent years, artificial intelligence technologies such as computer vision have been widely applied in the field of agriculture, but traditional convolutional neural networks (CNNs) have limitations in extracting global features under complex lighting and background variations. This article takes the four growth stages of wheat, namely “ germination,” “ tillering,” “heading,” and “maturity,” as the research objects. A new state space modeling network VMamba is introduced into the wheat growth stage detection task, and a wheat growth stage detection model based on VMamba is constructed. Comparative experiments are conducted with mainstream models such as ResNet50, MobileNetV3, and ViT. The experimental results show that the VMamba model outperforms the comparison model in terms of accuracy and F1 score indicators, and also has advantages in inference speed and parameter quantity. Overall, it exhibits better classification and detection performance. In addition, the feasibility and effectiveness of the VMamba model in agricultural visual inspection were verified through actual deployment on the Nvidia Jetson Orin platform, providing an implementation approach for intelligent crop monitoring.

关键词

卷积神经网络 / VMamba 模型 / 生长期检测 / 计算机视觉 / ResNet50

Key words

convolutional neural network / VMamba model / growth stage detection / computer vision / ResNet50

引用本文

引用格式 ▾
朱先远,李秀茹. 一种基于VMamba的小麦生长期检测模型研究[J]. 齐鲁工业大学学报, 2026, 40(4): 52-61 DOI:10.16442/j.cnki.qlgydxxb.2026.04.006

登录浏览全文

4963

注册一个新账户 忘记密码

参考文献

[1]

扶兰兰, 黄昊, 王恒, 等. 基于 SwinTransformer 模型的玉米生长期分类[J]. 农业工程学报, 2022, 38(14): 191-200.

[2]

徐建鹏, 王杰, 徐祥, 等. 基于 RAdam 卷积神经网络的水稻生育期图像识别[J]. 农业工程学报, 2021, 37(8): 143-150.

[3]

刘平, 刘立鹏, 王春颖, 等. 基于机器视觉的田间小麦开花期判定方法[J]. 农业机械学报, 2022, 53(3): 251-258.

[4]

赵春江, 文朝武, 林森, 等. 基于级联卷积神经网络的番茄花期识别检测方法[J]. 农业工程学报, 2020, 36(24): 143-152.

[5]

Yu Zhenghong, Cao Zhiguo, Wu Xi, et al. Automatic image-based detection technology for two critical growth stages of maize: Emergence and three-leaf stage[J]. Agricultural and Forest Meteorology, 2013, 174/ 175: 65-84.

[6]

韩悦婷, 行鸿彦, 金红伟. 基于 openCV 的玉米出苗期和三叶期自动检测系统的设计[J]. 电子测量与仪器学报, 2017, 31(10): 1574-1581.

[7]

张芸德, 刘蓉, 刘明, 等. 基于深度卷积特征的玉米生长期识别[J]. 电子测量技术, 2018, 41(16): 79-84.

[8]

Zheng Yang, Zhu Qibing, Huang Min, et al. Maize and weed classification using color indices with support vector data description in outdoor fields[J]. Computers and Electronics in Agriculture, 2017, 141: 215-222.

[9]

赵仲文, 张永立, 韩镇宇, 等. 基于改进的 SS-YOLOv8 轻量化鲜食玉米果穗优劣检测模型[J]. 农业工程学报, 2025, 41(11): 183-192.

[10]

温长吉, 王启锐, 陈洪锐, 等. 面向大规模多类别的病虫害识别模型[J]. 农业工程学报, 2022, 38(8): 169-177.

[11]

刘梦姝, 张春琪, 晁金阳, 等. 基于 YOLOv8n 改进的小麦病害检测系统[J]. 农业机械学报, 2024, 55(S1): 280-287.

[12]

Zhang Wang, Ren Ri, Guo Zidi, et al. Improved lightweight convolutional neural network models for the detection and evaluation of Fusarium head blight in wheat[J]. Journal of Zhejiang University-SCIENCE B, 2026, 27(5): 450-465.

[13]

毛锐, 郝思嘉, 李书琴, 等. DSFNet:双流空频协同的无人机小麦黄矮病高效检测模型[J]. 农业工程学报, 2026, 42(6): 197-205.

[14]

Vaswani A, Shazeer N, Parmar N, et al. Attention is all you need[C]// Advances in Neural Information Processing Systems. 2017, 30: 1-11.

[15]

Dosovitskiy A, Beyer L, Kolesnikov A, et al. An image is worth 16x16 words: Transformers for image recognition at scale[PP/OL]. V2.arXiv(2021-06-03)[2025-10-15]. https://iclr.cc/virtual/2021/poster/3013.

[16]

Liu Yue, Tian Yunjie, Zhao Yuzhong, et al. VMamba: Visual state space model[C]// Advances in Neural Information Processing Systems 37.Vancouver, BC, Canada.Neural Information Processing Systems Foundation, Inc.(NeurIPS), 2024: 103031-103063.

基金资助

安徽省教育厅自然科学研究重点项目(2023AH052295)

安徽省教育厅物联网应用技术教学创新团队项目(2024CXTD209)

AI Summary AI Mindmap
PDF (10771KB)

0

访问

0

被引

详细

导航
相关文章

AI思维导图

/

〈 〉