基于深度学习的小目标检测综述  

A review of deep learning based on small target detection

作　　者：王聪敏 李秀丽 Wang Congmin;Li Xiuli(School of Information Engineering,North China University of Water Resources and Electric Power,Zhengzhou 450000,China;Engineering Training Center,North China University of Water Resources and Electric Power,Zhengzhou 450000,China)

机构地区：[1]华北水利水电大学信息工程学院,郑州450000 [2]华北水利水电大学工程训练中心,郑州450000

出　　处：《现代计算机》2024年第7期59-63,共5页Modern Computer

摘　　要：该研究综述了基于深度学习的小目标检测技术,首先,梳理介绍了关键技术,对不同方法进行了比较分析,评估了它们的优劣势。其次,讨论了相关数据集和评估指标,为算法性能评估提供了坚实基础。此外,探讨了小目标检测在智能交通和安全监控等领域的各种应用场景。最后,提出了小目标检测的未来发展方向,强调了提高检测精度和速度的必要性。This study reviews the deep learning⁃based small target detection techniques.Firstly,combing and introducing the key techniques,comparing and analyzing the different methods,and evaluating their advantages and disadvantages.Secondly,rel⁃evant datasets and evaluation metrics are discussed in the article,providing a solid foundation for algorithm performance evalua⁃tion.In addition,the article discusses various application scenarios of small target detection in fields such as intelligent transporta⁃tion and security monitoring.Finally,the article proposes the future development direction of small target detection and emphasizes the necessity of improving detection accuracy and speed.

关 键 词：小目标检测 深度学习 目标检测 

分 类 号：TP3[自动化与计算机技术—计算机科学与技术]