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

Deconvolution

Deconvolution

Deconvolution, also known as inverse filtering, is the process of extracting the impulse response from known input and output signals. This technique is vital in scenarios where the system's characteristics are unknown, and they must be inferred from the observable signals.
Deconvolution involves several mathematical techniques to derive the impulse response. One common approach is polynomial division. In this method, the input and output sequences are treated as coefficients of...

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

Efficient Data Aggregation in Smart Grids: A Personalized Local Differential Privacy Scheme.

Sensors (Basel, Switzerland)·2026

Same author

Roles of miR-200 family members in lung cancer: more than tumor suppressors.

Future oncology (London, England)·2018

Same author

Alterations and structural resilience of the gut microbiota under dietary fat perturbations.

The Journal of nutritional biochemistry·2018

Same author

Development and Feasibility Testing of an mHealth (Text Message and WeChat) Intervention to Improve the Medication Adherence and Quality of Life of People Living with HIV in China: Pilot Randomized Controlled Trial.

JMIR mHealth and uHealth·2018

Same author

Meta-Analysis of Preclinical Studies of Fibrinolytic Therapy for Acute Lung Injury.

Frontiers in immunology·2018

Same author

Corrigendum to: The calcium sensor TaCBL4 and its interacting protein TaCIPK5 are required for wheat resistance to stripe rust fungus.

Journal of experimental botany·2018

Search research articles

相关实验视频

Updated: Mar 15, 2026

Whole-cell Super-Resolution Imaging via DNA-PAINT on a Spinning Disk Confocal with Optical Photon Reassignment

Whole-cell Super-Resolution Imaging via DNA-PAINT on a Spinning Disk Confocal with Optical Photon Reassignment

Published on: January 6, 2026

由可变形卷积和多平面尺度约束驱动:一个模糊的图像脱-拼接系统.

Sheng Hu¹, Han Xiao¹, Cong Liu¹

¹School of Electrical and Electronic Engineering, Hubei University of Technology, Wuhan 430068, China.

Sensors (Basel, Switzerland)

|March 14, 2026

概括

本研究介绍了一种使用可变形卷积v4 (DCNv4) 和Retinex启发的变压器的新型非均除雾方法,以提高雾中高级驾驶辅助系统 (ADAS) 的图像清晰度. 这种方法可以提高在恶劣天气下感知和准确度.

关键词:

深度学习是一种深度学习.可以变形的卷积卷积.功能匹配的功能匹配.图像去染图像去染图像去染图像拼接图像拼接图像拼接

更多相关视频

Digital Inline Holographic Microscopy DIHM of Weakly-scattering Subjects

Digital Inline Holographic Microscopy DIHM of Weakly-scattering Subjects

Published on: February 8, 2014

Automated Joint Space Detection Improves Bone Segmentation Accuracy

Automated Joint Space Detection Improves Bone Segmentation Accuracy

Published on: November 28, 2025

相关实验视频

Last Updated: Mar 15, 2026

Whole-cell Super-Resolution Imaging via DNA-PAINT on a Spinning Disk Confocal with Optical Photon Reassignment

Whole-cell Super-Resolution Imaging via DNA-PAINT on a Spinning Disk Confocal with Optical Photon Reassignment

Published on: January 6, 2026

Digital Inline Holographic Microscopy DIHM of Weakly-scattering Subjects

Digital Inline Holographic Microscopy DIHM of Weakly-scattering Subjects

Published on: February 8, 2014

Automated Joint Space Detection Improves Bone Segmentation Accuracy

Automated Joint Space Detection Improves Bone Segmentation Accuracy

Published on: November 28, 2025

科学领域:

计算机视觉计算机视觉
人工智能的人工智能
自动驾驶自动驾驶的自动驾驶

背景情况:

恶劣的天气,如雾,降低了图像质量,影响了对高级驾驶辅助系统 (ADAS) 至关重要的深度学习算法.
由于传感器差异和低质感特征,现有的除方法与不均的雾作斗争,导致背景信息丢失和ADAS场景中的图像拼接差.

研究的目的:

开发一种先进的非均的脱和图像接方法,以在雾条件下为自动驾驶提供强大的环境感知.
在具有挑战性的ADAS场景中,提高图像清晰度,特征匹配精度和拼接质量.

主要方法:

一个基于Deformable Convolution v4 (DCNv4) 的变形类网络被设计用于远程依赖和自适应空间聚合.
一个轻量级的Retinex灵感的变压器被集成用于色彩校正和结构改进.
采用使用LightGlue网络的多平面尺度约束模块和自适应融合接方法来提高匹配和接精度.

主要成果:

拟议的方法显著提高了特征匹配的准确性和同谱矩阵估计精度.
与现有方法相比,实现了22.78dB (NH-HAZE) 和24.34dB (BRAS) 的高峰信号噪声比 (PSNR).
在拼接图像中有效消除文物和过渡区域.

结论:

开发的非均的除雾和接技术为在雾条件下自动驾驶提供了可靠的环境感知解决方案.
验证了该方法在改善图像质量和ADAS感知方面的有效性和实用性.