Robust detection of small and dense objects in images from autonomous aerial vehicles

作     者：Lee, Joo Chan Yoo, JeongYeop Kim, Yongwoo Moon, SungTae Ko, Jong Hwan 

作者机构：Sungkyunkwan Univ Dept Artificial Intelligence Suwon South Korea Sungkyunkwan Univ Dept Elect & Comp Engn Suwon South Korea Sangmyung Univ Dept Syst Semicond Engn Cheonan South Korea Korea Aerosp Res Inst Artificial Intelligence Res Div Daejeon South Korea Sungkyunkwan Univ Coll Informat & Commun Engn Suwon South Korea 

出 版 物：《ELECTRONICS LETTERS》 (电子学快报)

年 卷 期：2021年第57卷第16期

页      面：611-613页

核心收录：

学科分类：0808[工学-电气工程] 0809[工学-电子科学与技术（可授工学、理学学位）] 08[工学] 

基　　金：KARI Institutional Program NRF grant [2019R1F1A1048115] IITP grant - Ministry of Science and ICT (MSIT) of the Korea government [IITP-2019-0-00421] 

主　　题：Optical, image and video signal processing Aerospace control Mobile robots Computer vision and image processing techniques 

摘      要：Aerial images obtained from autonomous aerial vehicles have lots of small and densely distributed objects because of the capture distance. This paper proposes a deep neural network architecture and training/inference techniques for robust detection of objects in the aerial images. Based on cascade R-CNN, the proposed model adopts the recursive feature pyramid and switchable atrous convolution for robust detection of dense objects. A patch-level division and multi-scale inference techniques are applied to effectively detect small objects. The results show that the proposed approach achieves the highest performance on the VisDrone test-dev dataset, in the official ECCV VisDrone2020-DET challenge.