Jove
Visualize
联系我们
JoVE
x logofacebook logolinkedin logoyoutube logo
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 JournalMethods CollectionsJoVE Encyclopedia of Experiments存档
教育
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab Manual教师资源中心教师网站
使用条款与条件
隐私政策
政策

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

RETRACTED: Martínez-Revuelta et al. Numerical Study of a Solar Cell to Achieve the Highest InGaN Power Conversion Efficiency for the Whole In-Content Range. <i>Micromachines</i> 2022, <i>13</i>, 1828.

Micromachines·2025
Same author

Plant Stress Detection Using a Three-Dimensional Analysis from a Single RGB Image.

Sensors (Basel, Switzerland)·2024
Same author

Non-Invasive Fish Biometrics for Enhancing Precision and Understanding of Aquaculture Farming through Statistical Morphology Analysis and Machine Learning.

Animals : an open access journal from MDPI·2024
Same author

Theoretical and Computational Analysis of a Wurtzite-AlGaN DUV-LED to Mitigate Quantum-Confined Stark Effect with a Zincblende Comparison Considering Mg- and Be-Doping.

Nanomaterials (Basel, Switzerland)·2022
Same author

RETRACTED: Numerical Study of a Solar Cell to Achieve the Highest InGaN Power Conversion Efficiency for the Whole In-Content Range.

Micromachines·2022
Same author

ECG-Based Identification of Sudden Cardiac Death through Sparse Representations.

Sensors (Basel, Switzerland)·2021

相关实验视频

Updated: May 10, 2025

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
03:31

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

Published on: December 15, 2023

430

城市空中图像中的三维着陆区细分从深度信息使用深度神经网络-超像素方法.

N A Morales-Navarro1, J A de Jesús Osuna-Coutiño1, Madaín Pérez-Patricio1

  • 1Department of Science, Tecnológico Nacional de México/IT de Tuxtla Gutiérrez, Carr. Panamericana Km. 1080, Tuxtla Gutiérrez 29050, Chiapas, Mexico.

Sensors (Basel, Switzerland)
|April 26, 2025
PubMed
概括

本研究引入了一种用于自动驾驶飞行器的新型3D着陆区细分方法,通过评估地面可达性来提高安全性. 深度神经网络 (DNN) 方法显著提高了登陆区检测的准确性.

关键词:
深度学习是一种深度学习.登陆区检测 登陆区检测超像素细分的超像素细分.三维的语义细分是三维的语义细分.

更多相关视频

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
04:48

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Published on: November 30, 2022

2.6K
Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique
04:48

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

319

相关实验视频

Last Updated: May 10, 2025

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
03:31

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

Published on: December 15, 2023

430
Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
04:48

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Published on: November 30, 2022

2.6K
Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique
04:48

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

319

科学领域:

  • 机器人技术 机器人技术 机器人技术
  • 计算机视觉 计算机视觉
  • 人工智能的人工智能

背景情况:

  • 目前,自动驾驶飞行器的着陆区检测主要使用RGB摄像头,缺乏深度感知和表面可访问性评估.
  • 基于RGB的方法可以识别由于不规则或难以进入的地形而不可行的着陆区域.
  • 使用3D深度信息在正确解释深度模糊性方面存在挑战.

研究的目的:

  • 为自动驾驶飞行器开发一个3D着陆区细分方法.
  • 通过结合深度感知和可访问性分析,提高着陆区检测的准确性和安全性.
  • 解决RGB-only方法在确定合适的着陆区域方面的局限性.

主要方法:

  • 一个DNN-超像素方法用于3D着陆区细分.
  • 使用超像素对区域进行细分和划分,聚类深度信息.
  • 使用边界框从相邻对象中提取特征.
  • 深度神经网络 (DNN) 用于根据可访问性将3D区域分类为可着陆或不可着陆的.

主要成果:

  • 拟议的方法实现了0.953的平均回忆,正确识别了95.3%的着陆区像素.
  • 平均精度为0.949,表明94.9%的细分着陆区是准确的.
  • 实验结果证明了DNN-超像素方法在3D着陆区检测中的可行性和前景.

结论:

  • DNN-超像素方法有效地对3D着陆区域进行细分,考虑到自动驾驶飞行器的表面可访问性.
  • 这种方法通过整合深度信息来克服仅使用RGB系统的局限性.
  • 高回忆和精度值表明着陆区检测准确性和安全性得到了显著改善.