• 文献检索
  • 文档翻译
  • 深度研究
  • 学术资讯
  • Suppr Zotero 插件Zotero 插件
  • 邀请有礼
  • 套餐&价格
  • 历史记录
应用&插件
Suppr Zotero 插件Zotero 插件浏览器插件Mac 客户端Windows 客户端微信小程序
定价
高级版会员购买积分包购买API积分包
服务
文献检索文档翻译深度研究API 文档MCP 服务
关于我们
关于 Suppr公司介绍联系我们用户协议隐私条款
关注我们

Suppr 超能文献

核心技术专利：CN118964589B侵权必究
粤ICP备2023148730 号-1Suppr @ 2026

文献检索

告别复杂PubMed语法，用中文像聊天一样搜索，搜遍4000万医学文献。AI智能推荐，让科研检索更轻松。

立即免费搜索

文件翻译

保留排版，准确专业，支持PDF/Word/PPT等文件格式，支持 12+语言互译。

免费翻译文档

深度研究

AI帮你快速写综述，25分钟生成高质量综述，智能提取关键信息，辅助科研写作。

立即免费体验

用于自动驾驶系统的自适应实时目标检测

Adaptive Real-Time Object Detection for Autonomous Driving Systems.

作者信息

Hemmati Maryam, Biglari-Abhari Morteza, Niar Smail

机构信息

Department of Electrical, Computer, and Software Engineering, The University of Auckland, Auckland 1010, New Zealand.

Institut National des Sciences Appliquées (INSA) Hauts-de-France, Université Polytechnique Hauts-de-France, 59300 Valenciennes, France.

出版信息

J Imaging. 2022 Apr 11;8(4):106. doi: 10.3390/jimaging8040106.

DOI:10.3390/jimaging8040106
PMID:35448233
原文链接:https://pmc.ncbi.nlm.nih.gov/articles/PMC9025781/
Abstract

Accurate and reliable detection is one of the main tasks of Autonomous Driving Systems (ADS). While detecting the obstacles on the road during various environmental circumstances add to the reliability of ADS, it results in more intensive computations and more complicated systems. The stringent real-time requirements of ADS, resource constraints, and energy efficiency considerations add to the design complications. This work presents an adaptive system that detects pedestrians and vehicles in different lighting conditions on the road. We take a hardware-software co-design approach on Zynq UltraScale+ MPSoC and develop a dynamically reconfigurable ADS that employs hardware accelerators for pedestrian and vehicle detection and adapts its detection method to the environment lighting conditions. The results show that the system maintains real-time performance and achieves adaptability with minimal resource overhead.

摘要

准确可靠的检测是自动驾驶系统（ADS）的主要任务之一。虽然在各种环境条件下检测道路上的障碍物可提高ADS的可靠性，但这会导致计算量更大且系统更复杂。ADS严格的实时要求、资源限制和能源效率考量增加了设计的复杂性。这项工作提出了一种自适应系统，可在道路上不同光照条件下检测行人和车辆。我们在Zynq UltraScale+ MPSoC上采用硬件-软件协同设计方法，开发了一种动态可重构的ADS，该系统采用硬件加速器进行行人和车辆检测，并根据环境光照条件调整其检测方法。结果表明，该系统保持了实时性能，并以最小的资源开销实现了适应性。

https://cdn.ncbi.nlm.nih.gov/pmc/blobs/f1e2/9025781/ce3031237483/jimaging-08-00106-g019.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/f1e2/9025781/841ab19288d5/jimaging-08-00106-g001.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/f1e2/9025781/8c3edcd096e6/jimaging-08-00106-g002.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/f1e2/9025781/91e0a5b27e78/jimaging-08-00106-g003.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/f1e2/9025781/f8bb452db478/jimaging-08-00106-g004.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/f1e2/9025781/39fb780d17cb/jimaging-08-00106-g005.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/f1e2/9025781/e5dff7a47a51/jimaging-08-00106-g006.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/f1e2/9025781/593ac2dc3a7e/jimaging-08-00106-g007.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/f1e2/9025781/f752c002046e/jimaging-08-00106-g008.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/f1e2/9025781/3c2c2f677e05/jimaging-08-00106-g009.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/f1e2/9025781/b70d7b430170/jimaging-08-00106-g010.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/f1e2/9025781/b69e31da33b1/jimaging-08-00106-g011.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/f1e2/9025781/d13f80b5620f/jimaging-08-00106-g012.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/f1e2/9025781/e896e50b5313/jimaging-08-00106-g013.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/f1e2/9025781/50219351354f/jimaging-08-00106-g014.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/f1e2/9025781/d6ddec262df9/jimaging-08-00106-g015.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/f1e2/9025781/5b2a20d54028/jimaging-08-00106-g016.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/f1e2/9025781/ea3eb0fc3ec5/jimaging-08-00106-g017.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/f1e2/9025781/bdda62765c35/jimaging-08-00106-g018.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/f1e2/9025781/ce3031237483/jimaging-08-00106-g019.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/f1e2/9025781/841ab19288d5/jimaging-08-00106-g001.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/f1e2/9025781/8c3edcd096e6/jimaging-08-00106-g002.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/f1e2/9025781/91e0a5b27e78/jimaging-08-00106-g003.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/f1e2/9025781/f8bb452db478/jimaging-08-00106-g004.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/f1e2/9025781/39fb780d17cb/jimaging-08-00106-g005.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/f1e2/9025781/e5dff7a47a51/jimaging-08-00106-g006.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/f1e2/9025781/593ac2dc3a7e/jimaging-08-00106-g007.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/f1e2/9025781/f752c002046e/jimaging-08-00106-g008.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/f1e2/9025781/3c2c2f677e05/jimaging-08-00106-g009.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/f1e2/9025781/b70d7b430170/jimaging-08-00106-g010.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/f1e2/9025781/b69e31da33b1/jimaging-08-00106-g011.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/f1e2/9025781/d13f80b5620f/jimaging-08-00106-g012.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/f1e2/9025781/e896e50b5313/jimaging-08-00106-g013.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/f1e2/9025781/50219351354f/jimaging-08-00106-g014.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/f1e2/9025781/d6ddec262df9/jimaging-08-00106-g015.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/f1e2/9025781/5b2a20d54028/jimaging-08-00106-g016.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/f1e2/9025781/ea3eb0fc3ec5/jimaging-08-00106-g017.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/f1e2/9025781/bdda62765c35/jimaging-08-00106-g018.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/f1e2/9025781/ce3031237483/jimaging-08-00106-g019.jpg

相似文献

1
Adaptive Real-Time Object Detection for Autonomous Driving Systems.用于自动驾驶系统的自适应实时目标检测
J Imaging. 2022 Apr 11;8(4):106. doi: 10.3390/jimaging8040106.
2
Hardware Trojan Attacks on the Reconfigurable Interconnections of Field-Programmable Gate Array-Based Convolutional Neural Network Accelerators and a Physically Unclonable Function-Based Countermeasure Detection Technique.针对基于现场可编程门阵列的卷积神经网络加速器可重构互连的硬件木马攻击及基于物理不可克隆功能的对策检测技术
Micromachines (Basel). 2024 Jan 19;15(1):149. doi: 10.3390/mi15010149.
3
Hardware-Software Partitioning for Real-Time Object Detection Using Dynamic Parameter Optimization.基于动态参数优化的实时目标检测的软硬件划分。
Sensors (Basel). 2023 May 19;23(10):4894. doi: 10.3390/s23104894.
4
A Conceptual Multi-Layer Framework for the Detection of Nighttime Pedestrian in Autonomous Vehicles Using Deep Reinforcement Learning.一种使用深度强化学习在自动驾驶车辆中检测夜间行人的概念性多层框架。
Entropy (Basel). 2023 Jan 9;25(1):135. doi: 10.3390/e25010135.
5
Pure FPGA Implementation of an HOG Based Real-Time Pedestrian Detection System.基于HOG的实时行人检测系统的纯FPGA实现
Sensors (Basel). 2018 Apr 12;18(4):1174. doi: 10.3390/s18041174.
6
Pedestrian Detection Using Multispectral Images and a Deep Neural Network.基于多光谱图像和深度神经网络的行人检测。
Sensors (Basel). 2021 Apr 4;21(7):2536. doi: 10.3390/s21072536.
7
An Unsupervised Transfer Learning Framework for Visible-Thermal Pedestrian Detection.基于无监督迁移学习的可见光-热行人检测框架。
Sensors (Basel). 2022 Jun 10;22(12):4416. doi: 10.3390/s22124416.
8
Adopting the YOLOv4 Architecture for Low-Latency Multispectral Pedestrian Detection in Autonomous Driving.采用 YOLOv4 架构实现自动驾驶中的低延迟多光谱行人检测。
Sensors (Basel). 2022 Jan 30;22(3):1082. doi: 10.3390/s22031082.
9
Hardware/Software Co-design of Fractal Features based Fall Detection System.基于分形特征的跌倒检测系统的软硬件协同设计。
Sensors (Basel). 2020 Apr 18;20(8):2322. doi: 10.3390/s20082322.
10
A Hybrid Vehicle Detection Method Based on Viola-Jones and HOG + SVM from UAV Images.一种基于Viola-Jones和HOG+SVM的无人机图像混合车辆检测方法。
Sensors (Basel). 2016 Aug 19;16(8):1325. doi: 10.3390/s16081325.

引用本文的文献

1
An Object-Centric Hierarchical Pose Estimation Method Using Semantic High-Definition Maps for General Autonomous Driving.一种用于通用自动驾驶的、基于语义高清地图的以物体为中心的分层姿态估计方法。
Sensors (Basel). 2024 Aug 11;24(16):5191. doi: 10.3390/s24165191.
2
Development of an Autonomous Driving Vehicle for Garbage Collection in Residential Areas.自主驾驶车辆在居民区的垃圾收集应用开发。
Sensors (Basel). 2022 Nov 23;22(23):9094. doi: 10.3390/s22239094.
3
Dynamic Label Assignment for Object Detection by Combining Predicted IoUs and Anchor IoUs.

本文引用的文献

1
Fusing Appearance and Spatio-Temporal Models for Person Re-Identification and Tracking.融合外观与时空模型用于行人重识别与跟踪
J Imaging. 2020 May 1;6(5):27. doi: 10.3390/jimaging6050027.
2
High-Profile VRU Detection on Resource-Constrained Hardware Using YOLOv3/v4 on BDD100K.在资源受限硬件上使用YOLOv3/v4并基于BDD100K进行高姿态VRU检测。
J Imaging. 2020 Dec 19;6(12):142. doi: 10.3390/jimaging6120142.
3
A fast learning algorithm for deep belief nets.一种用于深度信念网络的快速学习算法。
通过结合预测交并比和锚框交并比进行目标检测的动态标签分配
J Imaging. 2022 Jul 11;8(7):193. doi: 10.3390/jimaging8070193.
Neural Comput. 2006 Jul;18(7):1527-54. doi: 10.1162/neco.2006.18.7.1527.
4
Generic object recognition with boosting.基于提升算法的通用目标识别
IEEE Trans Pattern Anal Mach Intell. 2006 Mar;28(3):416-31. doi: 10.1109/TPAMI.2006.54.