北京航空航天大学学报 ›› 2020, Vol. 46 ›› Issue (3): 634-642.doi: 10.13700/j.bh.1001-5965.2019.0279

• 论文 • 上一篇    下一篇

基于视频帧间运动估计的无人机图像车辆检测

陈映雪1, 丁文锐2, 李红光2, 王蒙1, 王旭3   

  1. 1. 北京航空航天大学 电子信息工程学院, 北京 100083;
    2. 北京航空航天大学 无人系统研究院, 北京 100083;
    3. 合一智芯科技有限公司, 北京 100083
  • 收稿日期:2019-06-03 发布日期:2020-03-28
  • 通讯作者: 李红光 E-mail:lihongguang@buaa.edu.cn
  • 作者简介:陈映雪,女,硕士研究生。主要研究方向:遥感图像目标检测算法及应用;丁文锐,女,博士,研究员,博士生导师。主要研究方向:多源图像信息处理、视觉目标检测与跟踪;李红光,男,博士,高级工程师,硕士生导师。主要研究方向:无人系统光学图像智能处理及边缘计算应用。
  • 基金资助:
    国防基础科研计划(JCKY2017601C006);武汉大学测绘遥感信息工程国家重点实验室开放基金(17E01)

Vehicle detection in UAV image based on video interframe motion estimation

CHEN Yingxue1, DING Wenrui2, LI Hongguang2, WANG Meng1, WANG Xu3   

  1. 1. School of Electronic and Information Engineering, Beihang University, Beijing 100083, China;
    2. Institute of Unmanned System, Beihang University, Beijing 100083, China;
    3. Heyintelligence Technology Limited Company, Beijing 100083, China
  • Received:2019-06-03 Published:2020-03-28
  • Supported by:
    National Defense Basic Scientific Research Program of China (JCKY2017601C006); Open Fund of State Key Laboratory of Information Engineering in Surveying, Mapping and Remote Sensing, Wuhan University (17E01)

摘要: 基于人工智能(AI)芯片搭建轻量化深度神经网络,可以在无人机(UAV)机载端实现视频中车辆目标的自动检测,具有重要的应用前景。为此,提出了一种针对无人机图像车辆目标的检测方法,并在AI芯片上进行部署与测试。方法具体包括:结合无人机图像中车辆目标的尺寸范围,对MobileNet-SSD网络进行裁剪,构建轻量化单帧图像检测器;为解决小目标特性在轻量网络框架下引发的检测性能下降问题,引入帧间运动矢量估计,根据相邻帧信息辅助预测当前帧丢失目标的位置范围,并利用检测结果进行修正,实现丢失目标的再召回。通过对多个数据集进行融合与自动补充标注,搭建了一个高质量的无人机图像车辆目标数据集;同时将方法在基于RK3399芯片计算的嵌入式开发平台上进行实验验证,结果表明:搭建的网络能够显著减少存储资源占用,具有轻量化的特点;同时相比于单帧检测法,引入视频帧间运动估计方法可以有效提高检测精度,并在AI芯片上实现125.3 ms/帧的检测速度。

关键词: 无人机(UAV), 目标检测, 轻量化神经网络, 人工智能(AI)芯片, 运动估计

Abstract: The lightweight neural network embedded on artificial intelligence (AI) chips can realize the onboard automatic detection of vehicle objects in unmanned aerial vehicle (UAV) videos, which is important in practical applications. In this paper, a vehicle object detection algorithm in UAV videos is proposed, and then deployed and tested on AI chips. For the proposed detection algorithm, firstly, the MobileNet-SSD network is clipped based on the range of vehicle objects' size in UAV images to construct a lightweight single-frame object detector. Secondly, the interframe motion estimation was introduced to improve the poor detection performance which is usually caused by small object characteristics and lightweight network. Thirdly, the position range of missing objects in the current frame is predicted according to the information of adjacent frames. Finally, the predicted position is corrected by detection results, and the recall of lost objects is realized. Additionally, a high-quality UAV image vehicle dataset was built by fusion and automatic supplementary annotation of multiple datasets. The proposed algorithm is verified on the embedded development platform based on RK3399 chip. The results show that the network with the proposed algorithm can significantly reduce the occupation of storage resources with the lightweight characteristics. Compared to the traditional single-image detection algorithm, the proposed algorithm can effectively improve the detection accuracy. Moreover, detection speed can be as low as 125.3 ms per frame on the AI chip.

Key words: unmanned aerial vehicle (UAV), object detection, lightweight neural network, artificial intelligence (AI) chip, motion estimation

中图分类号: 


版权所有 © 《北京航空航天大学学报》编辑部
通讯地址:北京市海淀区学院路37号 北京航空航天大学学报编辑部 邮编:100191 E-mail:jbuaa@buaa.edu.cn
本系统由北京玛格泰克科技发展有限公司设计开发