低慢小飞行物实时检测原理与实验研究

于凯洋; 张磊; 王克逸

低慢小飞行物实时检测原理与实验研究

中国科学技术大学精密机械与精密仪器系，安徽合肥 230026

基金项目: 国家自然科学基金（61935008）

详细信息

作者简介:
于凯洋（1997—），男，硕士研究生，主要从事FPGA视频图像处理技术研究。E-mail：ykaiyang@mail.ustc.edu.cn

通讯作者:
王克逸（1962—），男，博士，博士生导师，教授，主要从事光电系统、微光学与信息光学技术研究。E-mail：kywang@ustc.edu.cn

中图分类号: TN02;TP277
计量
- 文章访问数: 160
- HTML全文浏览量: 55
- PDF下载量: 40
- 被引次数: 0
出版历程
- 网络出版日期: 2022-07-06

Real-time detection principle and experimental study of low and slow small flying objects

Department of Precision Machinery and Precision Instrumentation, University of Science and Technology of China, Hefei 230026, China

摘要

摘要: 飞鸟撞击飞机与无人机黑飞是威胁航班起降安全的两大隐患，上述的飞鸟和无人机都属于“低慢小”飞行物。为保卫机场净空区的安全，需要研制具有反低慢小功能的预警系统。对此，设计并实现了一套低慢小飞行物实时检测系统，基于FPGA（field programmable gate array）驱动相机阵列实时采集大视场天空视频，将适于硬件操作的奇偶分流算法与帧间差分算法相结合进行运动目标检测，系统帧率达到17 fps@1 024×768 pixel，平均检测准确率为99.69%。采用千兆以太网、光纤和交换机将前端跑道视频传输至后端塔台指挥中心，支持3 km远距离传输。相比于传统基于软件串行处理的方式，该系统具有高实时性、低功耗与小体积的优势，适合部署在实际应用场景中。
- 反低慢小 /
- 运动目标检测 /
- 远距离视频传输 /
- 大视场 /
- FPGA
Abstract: Birds and drones are two major threats to flight safety, and both are "low and slow small" flying objects. In order to protect the safety of the airport clear zone, we need to develop an early warning system with the function of anti-low and slow small flying objects. In this regard, we designed and implemented a real-time detection system for low and slow small flying objects, based on FPGA (Field Programmable Gate Array) driven camera arrays for real-time acquisition of large-field-of-view sky video, combining parity shunt algorithm and inter-frame differential algorithm for hardware operation to detect moving targets, and the system frame rate reached 17 fps@1 024x768 pixel, with an average detection accuracy of 99.69%. Gigabit Ethernet, optical fiber and switch are used to transmit the front-end runway video to the back-end tower command center, supporting 3 km long-distance transmission. Compared with the traditional software-based serial processing, the system has the advantages of high real-time, low power consumption and small size, which is suitable for deployment in practical application scenarios.
- anti-low and slow small /
- motion target detection /
- long distance video transmission /
- large field of view /
- FPGA

HTML全文

图 1 系统整体设计方案

Fig. 1 Overall system design solution

下载: 全尺寸图片幻灯片

图 2 运动目标检测算法流程

Fig. 2 Motion target detection algorithm flow

下载: 全尺寸图片幻灯片

图 3 FPGA中值滤波

Fig. 3 FPGA median filtering

下载: 全尺寸图片幻灯片

图 4 帧间差分仿真图

Fig. 4 Inter-frame differential simulation diagram

下载: 全尺寸图片幻灯片

图 5 系统实物图

Fig. 5 System physical diagram

下载: 全尺寸图片幻灯片

图 6 运动目标检测结果

Fig. 6 Motion target detection results

下载: 全尺寸图片幻灯片

图 7 LCD屏人机交互界面

Fig. 7 LCD screen human-machine interaction interface

下载: 全尺寸图片幻灯片

表 1 检测准确率评价结果

Table 1 Detect accuracy evaluation results %

评价指标	最大值	最小值	中值	平均值
E_drel	2.497 4	0.091 6	0.883 4	0.875 8
R_FP	2.663 7	0.000 0	0.060 9	0.134 4
R_FN	1.239 6	0.000 0	0.093 2	0.141 4
P_WC	2.692 2	0.025 4	0.184 8	0.267 6
R_SP	100.000 0	97.336 3	99.939 1	99.865 6

下载: 导出CSV

表 2 不同平台性能对比

Table 2 Different platform performance comparison

平台	帧率/fps	功耗/W	尺寸/（cm³）
双FPGA	17	19.72	16.04×11.57×7.26
PC机	4	65.09	46.72×18.76×43.45

下载: 导出CSV

参考文献(20)

[1]	王维, 项洪达. 鸟击风险计算方法研究[J]. 中国民航大学学报,2019,37(5):21-24. doi: 10.3969/j.issn.1674-5590.2019.05.005 WANG Wei, XIANG Hongda. Research on the calculation method of bird strike risk[J]. Journal of Civil Aviation University of China,2019,37(5):21-24. doi: 10.3969/j.issn.1674-5590.2019.05.005
[2]	FLOREANO D, WOOD R J. Science, technology and the future of small autonomous drones[J]. Nature,2015,521(7553):460-466. doi: 10.1038/nature14542
[3]	张建伟, 郭会明. 低空慢速小目标拦截系统研究[J]. 计算机工程与设计,2012,33(7):2874-2878. doi: 10.3969/j.issn.1000-7024.2012.07.067 ZHANG Jianwei, GUO Huiming. Research on low altitude slow speed small target interception system[J]. Computer Engineering and Design,2012,33(7):2874-2878. doi: 10.3969/j.issn.1000-7024.2012.07.067
[4]	SHI X, YANG C, XIE W, et al. Anti-drone system with multiple surveillance technologies: architecture implementation and challenges[J]. IEEE Communications Magazine,2018,56(4):68-74. doi: 10.1109/MCOM.2018.1700430
[5]	陈唯实, 黄毅峰, 陈小龙, 等. 机场净空区飞鸟与非合作无人机目标识别[J]. 民航学报,2020,4(3):27-33. CHEN Weishi, HAUNG Yifeng, CHEN Xiaolong, et al. Identification of flying birds and non-cooperative UAV targets in airport clearing areas[J]. Journal of Civil Aviation,2020,4(3):27-33.
[6]	KITTLE D S, MARKS D L, SON H S, et al. A testbed for wide-field, high-resolution, gigapixel-class cameras[J]. Review of Scientific Instruments,2013,84(5):386-389.
[7]	谢家阳, 王行健, 史治国, 等. 动态云台摄像机无人机检测与跟踪算法[J]. 智能系统学报,2021,16(5):858-869. doi: 10.11992/tis.202103032 XIE Jiayang, WANG Xingjian, SHI Zhiguo, et al. Dynamic gimbal camera UAV detection and tracking algorithm[J]. Journal of Intelligent Systems,2021,16(5):858-869. doi: 10.11992/tis.202103032
[8]	常昊. 低空飞行目标检测系统设计[D]. 西安: 西安电子科技大学, 2020. CHANG Hao. Design of low altitude flight target detection system[D]. Xi'an: Xi'an University of Electronic Science and Technology, 2020.
[9]	TANG J W, SHAIKH H N, SHEIKN U U, et al. FPGA-based real-time moving target detection system for unmanned aerial vehicle application[J]. International Journal of Reconfigurable Computing,2016,缺卷期号:1-16.
[10]	AMIR H R, TABERNER A J, NASH M P, et al. Suitability of recent hardware accelerators (DSPs, FPGAs, and GPUs) for computer vision and image processing algorithms[J]. Signal Processing Image Communication,2018,68:101-119. doi: 10.1016/j.image.2018.07.007
[11]	冈萨雷斯, 伍兹. 数字图像处理[M]. 第3版. 阮秋琦, 阮宇智, 等, 译. 北京: 电子工业出版社, 2017: 254-257. RAFAEL C G, RICHARD E W. Digital image processing [M]. 3rd ed. RUAN Qiuqi, RUAN Yuzhi, et al, Transl. Beijing: Electronic Industry Press, 2017: 254-257.
[12]	LI B, ZAN S S, JING Q Y, et al. Detection and tracking method for maneuvering targets using improved gradient-based optical flow field[C]. 缺出版地: 2019 Chinese Control And Decision Conference (CCDC), 2019: 1421-1425.
[13]	POLMOTTAWEGEDARA S S, MUNASINGHE R, DAVARI A. Tracking moving targets[C]// Symposium on System Theory. IEEE, 2006.
[14]	LIPTON A J, FUJIYOSHI H, PATIL R S. Moving target classification and tracking from Real-time video[C]// 缺出版地: Applications of Computer Vision, 1998 WACV '98. Proceedings. Fourth IEEE Workshop on. IEEE, 1998.
[15]	陈科成, 苏成悦, 何雷, 等. 基于ZYNQ架构的运动物体检测系统[J]. 信息技术与信息化,2020(3):63-66. doi: 10.3969/j.issn.1672-9528.2020.03.022 CHEN Kecheng, SU Chengyue, HE Lei, et al. ZYNQ architecture-based moving object detection system[J]. Information Technology and Informatization,2020(3):63-66. doi: 10.3969/j.issn.1672-9528.2020.03.022
[16]	王俊龙. 基于FPGA的运动目标检测系统的设计[D]. 太原: 中北大学, 2021. WANG Junlong. Design of FPGA-based motion target detection system[D]. Taiyuan: North Central University, 2021.
[17]	邢凯. 基于FPGA的多运动目标检测技术研究[D]. 昆明: 昆明理工大学, 2020. XING Kai. Research on FPGA-based multi-motion target detection technology[D]. Kunming: Kunming University of Technology, 2020.
[18]	戴万. 基于FPGA的星载舰船目标检测系统设计与实现[D]. 武汉: 华中科技大学, 2018. DAI Wan. Design and implementation of FPGA-based target detection system for starboard ships[D]. Wuhan: Huazhong University of Science and Technology, 2018.
[19]	张涛. 机场管制塔台通视能力评估方法研究[D]. 天津: 中国民航大学, 2017. ZHANG Tao. Research on the evaluation method of airport control tower visibility capability[D]. Tianjin: Civil Aviation University of China, 2017.
[20]	GOYETTE N, JODOIN P M, PORIKLI F, et al. Changedetection. net: A new change detection benchmark dataset[C]//缺出版地: Computer Vision & Pattern Recognition Workshops. IEEE, 2012.