基于图像目标识别算法的车位识别系统研究
Research on Parking Space Recognition System Based on Image Object Recognition Algorithm
马骁 1陈瑞 1马毓辰 1杨景哲 1赵冬辉2
作者信息
- 1. 辽宁工程技术大学 辽宁阜新 123000
- 2. 承德医学院 河北承德 067000
- 折叠
摘要
随着社会经济的迅猛发展,我国私人汽车保有量逐年提高,停车位的供给已无法满足人们的需求.为解决当前小城市及欠发达地区停车场空余车位信息无法获取这一问题,通过采用无人机飞行器搭载倾斜摄影技术来捕获停车相关信息,基于Paddle Detection的图像目标识别算法建立FairMOT模型,最后借助多终端服务系统发布实时数据,实现空余车位识别,提高车位识别的精确率和出行效率.
Abstract
With the rapid development of urban social economy,the number of private cars in China has increased year by year,and the supply of parking space has been unable to meet people's needs.In order to solve the problem that available parking space information can not be obtained in small cities and underdeveloped areas,this system uses tilt photography technology carried by unmanned aerial vehicles to capture related parking information,and es-tablishes a FairMOT model based on the Image Object Recognition Algorithm of Paddle Detection.Finally,it achives available parking space recognition by issuing real-time data with the multi-terminal service system to im-prove the accuracy of parking space recognition and travel efficiency.
关键词
智能交通/车位识别/物联网/图像目标识别算法Key words
Intelligent transportation/Parking space recognition/Internet of Things/Image Object Recognition Algorithm引用本文复制引用
出版年
2024
NETL
NSTL
万方数据