基于障碍物和车位检测的单阶段多任务YOLO-Parking算法研究
Research on single-stage multi-task YOLO-Parking algorithm based on obstacle and parking space detection
张炳力 1王焱辉 1潘泽昊 1王怿昕 1杨程磊 1王欣雨1
作者信息
- 1. 合肥工业大学汽车与交通工程学院,安徽合肥 230009;安徽省智能汽车工程实验室,安徽合肥 230009
- 折叠
摘要
文章提出一种基于YOLOv4的端到端多任务网络模型用于自动泊车系统中的感知任务,以环视图像(around view monitor,AVM)作为网络输入,基于卷积网络提取图像特征信息,通过YOLO和DMPR-PS(di-rectional marking-point regression-parking slot)检测头实现停车位与障碍物并行检测.在PS 2.0公开数据集上进行验证的结果表明,所提出的多任务检测方法能够同时检测停车位和障碍物,障碍物识别平均精度均值达到89.72％,车位识别查准率达到93.53％,网络检测速率为34.0帧/s,在满足自动泊车感知任务需求的同时提升了系统的检测效率.该文研究成果对自动泊车感知技术的发展具有一定的意义.
Abstract
This paper proposes an end-to-end multi-task network model based on YOLOv4 for percep-tion tasks in automatic parking systems.The model takes around view monitor(AVM)images as net-work inputs and extracts image feature information using convolutional networks.By utilizing YOLO and directional marking-point regression-parking slot(DMPR-PS)detection heads,it achieves parallel detection of parking spaces and obstacles.Validation on the PS 2.0 public dataset demonstrates that the proposed multi-task detection method can simultaneously detect parking spaces and obstacles,with an average precision of 89.72％for obstacle recognition and a precision rate of 93.53％for park-ing space recognition.The network achieves a detection rate of 34.0 frames per second,meeting the requirements of automatic parking perception tasks while improving detection efficiency.This research holds implications for the development of automatic parking perception technology.
关键词
自动泊车/环视图像(AVM)/多任务网络/障碍物识别/停车位识别Key words
automatic parking/around view monitor(AVM)/multi-task network/obstacle recogni-tion/parking space recognition引用本文复制引用
基金项目
长三角科技创新共同体联合攻关专项资助项目(2022CSJGG1501)
安徽省科技重大专项资助项目(202203a05020008)
安徽省发展和改革委员会2021新能源汽车产业创新发展资助项目(wfgcyh2021439)
中国声谷创新发展关键核心技术揭榜挂帅攻关资助项目(2108-340161-04-01-727575)
合肥市关键共性技术研发和重大科技成果工程化资助项目(2021CG003)
出版年
2024
NETL
NSTL
维普
万方数据