Search research articles

关于 JoVE

概览领导团队博客 JoVE 帮助中心

作者

出版流程编辑委员会范围与政策同行评审常见问题投稿

图书馆员

用户评价订阅访问资源图书馆顾问委员会常见问题

研究

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments 存档

教育

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual 教师资源中心教师网站

使用条款与条件

相关概念视频

您也可能阅读

相关文章

通过共同作者、期刊和引用图与本文相关的文章。

排序

Same author

Prognostic Implications of the Periaortic Fat Attenuation Index in Uncomplicated Type B Aortic Dissection.

Radiology. Cardiothoracic imaging·2025

Same author

Potential value of saline-induced Pd/Pa ratio in patients with coronary artery stenosis.

Frontiers in cardiovascular medicine·2023

Same author

An Autopsy Case of Pulmonary Veno-Occlusive Disease Complicated with Chronic Obstructive Pulmonary Disease and Severe Pulmonary Hypertension.

International heart journal·2021

Same author

Images in cardiovascular medicine. Pulmonary embolism due to popliteal venous aneurysm.

Circulation·2008

Same author

Images in cardiovascular medicine. Silent double aortic arch found in an elderly man.

Circulation·2006

Same author

[Complete left bundle branch block recorded in an emergency patient with shortness of breath].

Journal of cardiology·2005

Search research articles

相关实验视频

Updated: May 10, 2025

Author Spotlight: Enhancing PSC-to-Functional Cell Differentiation Using ML Models Based on Live-Cell Bright-Field Imaging

Author Spotlight: Enhancing PSC-to-Functional Cell Differentiation Using ML Models Based on Live-Cell Bright-Field Imaging

Published on: October 4, 2024

对于黄瓜分类的RGB色彩空间增强训练数据生成

Hotaka Hoshino¹, Takuya Shindo¹, Takefumi Hiraguri¹

¹Nippon Institute of Technology, 4-1 Gakuendai, Miyashiro, Saitama 345-8501, Japan.

Journal of imaging

|April 25, 2025

概括

这项研究开发了一个使用卷积神经网络 (CNN) 来分类黄瓜的AI系统,通过将物理属性编码为图像背景来提高准确性. 新方法的准确度达到79.1%,超过了传统方法.

科学领域:

农业技术农业技术
计算机视觉计算机视觉计算机视觉
机器学习机器学习

背景情况:

黄瓜分类依赖于专业知识,限制了未经培训的劳动力参与.
收获旺季需要高效,可扩展的黄瓜分类解决方案.
目前的手动分类方法耗时且容易出现不一致.

研究的目的:

开发一个自动化的黄瓜分类系统,可供没有先前专业知识的个人使用.
提高农业环境中黄瓜分类的准确性和效率.
利用深度学习进行客观和一致的收获产品质量评估.

主要方法:

一个有11个层的卷积神经网络 (CNN) (2个卷积,2个聚合,3个密集,4个掉落) 设计用于基于图像的分类.
一种新的数据增强技术将黄瓜的物理属性 (长度,曲,厚度) 嵌入到训练图像背景的RGB颜色空间中.
使用标准的多类分类指标 (准确性,回忆,精度,F-测量) 对现实世界合作标准进行绩效评估.

主要成果:

拟议的方法,将属性信息嵌入RGB背景中,实现了79.1%的分类准确度.
在没有RGB色彩空间嵌入的情况下,基线方法实现了70.1%的准确性.
改进的方法比传统方法提高了1.1倍的性能,验证了背景编码技术的有效性.

关键词:

这就是为什么物联网物联网物联网.黄瓜黄瓜的使用方法机器学习是机器学习.智能农业智能农业

更多相关视频

Tomato Analyzer: A Useful Software Application to Collect Accurate and Detailed Morphological and Colorimetric Data from Two-dimensional Objects

Tomato Analyzer: A Useful Software Application to Collect Accurate and Detailed Morphological and Colorimetric Data from Two-dimensional Objects

Published on: March 16, 2010

Combining Eye-tracking Data with an Analysis of Video Content from Free-viewing a Video of a Walk in an Urban Park Environment

Combining Eye-tracking Data with an Analysis of Video Content from Free-viewing a Video of a Walk in an Urban Park Environment

Published on: May 7, 2019

相关实验视频

Last Updated: May 10, 2025

Author Spotlight: Enhancing PSC-to-Functional Cell Differentiation Using ML Models Based on Live-Cell Bright-Field Imaging

Author Spotlight: Enhancing PSC-to-Functional Cell Differentiation Using ML Models Based on Live-Cell Bright-Field Imaging

Published on: October 4, 2024

Tomato Analyzer: A Useful Software Application to Collect Accurate and Detailed Morphological and Colorimetric Data from Two-dimensional Objects

Tomato Analyzer: A Useful Software Application to Collect Accurate and Detailed Morphological and Colorimetric Data from Two-dimensional Objects

Published on: March 16, 2010

Combining Eye-tracking Data with an Analysis of Video Content from Free-viewing a Video of a Walk in an Urban Park Environment

Combining Eye-tracking Data with an Analysis of Video Content from Free-viewing a Video of a Walk in an Urban Park Environment

Published on: May 7, 2019

结论:

通过RGB色彩空间将物理黄瓜属性嵌入到图像背景中,可以显著提高基于CNN的分类准确性.
开发的系统提供了一个自动化黄瓜分类的实用解决方案,减少对专家知识的依赖.
这种方法有可能简化农业收获后的流程,提高整体运营效率.