|
|
论文:2020,Vol:38,Issue(6):1345-1351 |
|
|
引用本文： |
|
|
杨锋平, 马博迪, 王金荣, 高红岗, 刘贞报. 基于深度去噪自动编码器的无人机航空影像目标检测[J]. 西北工业大学学报 |
|
|
YANG Fengping, MA Bodi, WANG Jinrong, GAO Honggang, LIU Zhenbao. Target Detection of UAV Aerial Image Based on Rotational Invariant Depth Denoising Automatic Encoder[J]. Northwestern polytechnical university |
|
|
|
|
|
|
|
| 基于深度去噪自动编码器的无人机航空影像目标检测 |
|
| 杨锋平1, 马博迪2, 王金荣3, 高红岗2, 刘贞报2 |
|
1. 中国石油集团石油管工程技术研究院, 陕西 西安 710077;
2. 西北工业大学 民航学院, 陕西 西安 710072;
3. 中航油彭州管道运输有限公司, 四川 成都 610202 |
| 摘要: |
| 利用无人机航拍获取目标场景影像信息的方式，具有可低空作业、覆盖面积广、机动性强、效率高、不受地势环境阻碍等优点，广泛应用于军民用领域，军事领域包括威胁目标空中监视、目标搜索、目标打击，民用领域包括交通监测、灾难营救、管线巡检、区域勘测、边境巡逻等方面。无人机航空影像目标检测过程中，针对待识别目标具有多个角度、成像像素尺寸小、机体震动干扰强等困难，提出一种基于深度去噪自动编码器的目标检测模型。该模型通过进行选择性搜索，提取航空影像感兴趣区域，计算感兴趣区域的径向梯度特征，得到旋转不变特征向量，利用深度去噪自动编码器滤掉原始数据中的噪声，并提取特征向量的深层特征。在国际无人机低空航空影像标准数据集UAV123以及德国宇航院的慕尼黑无人机航空影像集DLR 3K上开展了识别实验，结果表明，针对航空影像目标包括地面车辆、行人、海面船只等，所提方法能够达到90%以上的识别精度，在精准率、召回率、F1调和值等指标上领先于现有方法。 |
| 关键词:
无人机航空影像
目标检测
深度去噪自动编码器
旋转不变性
|
|
| Target Detection of UAV Aerial Image Based on Rotational Invariant Depth Denoising Automatic Encoder |
|
| YANG Fengping1, MA Bodi2, WANG Jinrong3, GAO Honggang2, LIU Zhenbao2 |
|
1. CNPC Tubular Goods Research Institute, Xi'an 710077, China;
2. School of Civil Aviation, Northwestern Polytechnical University, Xi'an 710072, China;
3. China National Aviation Fuel Pengzhou Pipeline Transportation Company Limited, Chengdu 610202, China |
| Abstract: |
| The method of using unmanned aerial vehicle (UAV) to obtain aerial image information of target scene has the characteristics of wide coverage, strong mobility and high efficiency, which is widely used in urban traffic monitoring, vehicle detection, oil pipeline inspection, regional survey and other aspects. Aiming at the difficulties of the object to be detected in the process of aerial image object detection, such as multiple orientations, small image pixel size and UAV body vibration interference, a novel aerial image object detection model based on the rotation-invariant deep denoising auto encoder is proposed in this paper. Firstly, the interest region of the aerial image is extracted by the selective search method, and the radial gradient of interest region is calculated. Then, the rotation invariant feature descriptor is obtained from the radial gradient feature, and the noise in the original data is filtered out by the deep denoising automatic encoder and the deep feature of the feature descriptors is extracted. Finally, the experimental results show that this method can achieve high accuracy for aerial image target detection and has good rotation invariance. |
| Key words:
UAV aerial image
target detection
deep denoising automatic encoder
rotation-invariant
model
simulation experiment
|
|
| 收稿日期: 2020-03-20
修回日期:
|
| DOI: 10.1051/jnwpu/20203861345 |
| 基金项目: 国家自然科学基金（52072309）、陕西省重点研发计划（2019ZDLGY14-02-01）、深圳市基础研究（JCYJ20190806152203506）、航空科学基金（ASFC-2018ZC53026）与国家留学基金创新型人才国际合作培养项目（201906290246）资助 |
| 通讯作者: 马博迪(1993-),西北工业大学博士研究生,主要从事无人机影像目标检测与跟踪控制技术研究。e-mail:maboid@mail.nwpu.edu.cn
Email：maboid@mail.nwpu.edu.cn |
| 作者简介: 杨锋平(1982-),中国石油集团石油管工程技术研究院高级工程师,主要从事油气管道完整性评价研究。
|
|
|
|
|
|
|
|
参考文献: |
|
|
[1] REDMON J, DIVVALA S, GIRSHICK R, et al. You Only Look Once:Unified, Real-Time Object Detection[C]//IEEE Conference on Computer Vision and Pattern Recognition, Las Vegas, 2016:779-788
[2] REN S, HE K, GIRSHICK R, et al. Faster RCNN:Towards Real-Time Object Detection with Region Proposal Networks[J]. IEEE Trans on Pattern Analysis and Machine Intelligence, 2017, 39(6):1137-1149
[3] ZHOU Z, FENG J. Deep Forest:Towards an Alternative to Deep Neural Networks[C]//Proc 26th Int Joint Conf Artif Intell, Melbourne, Australia, 2017:3553-3559
[4] UIJLINGS J R R, VANDESANDE K E A, GEVERS T, et al. Selective Search for Object Recognition[J]. Int J Comput Vis, 2013, 104(2):154-171
[5] MUELLER M, SMITH N, GHANEM B. A Benchmark and Simulator for UAV Tracking[C]//European Conference on Computer Vision, 2016
[6] LIU K, MATTYUS G. DLR 3K Munich Vehicle Aerial Image Dataset[EB/OL]. (2015-12-31)[2020-01-05]. http://pbafreesoftware.eoc.dlr.de/3K_VehicleDetection_dataset.zip
[7] LIU K, MATTYUS G. Fast Multiclass Vehicle Detection on Aerial Images[J]. IEEE Geosci Remote Sens Lett, 2015, 12(9):1938-1942
[8] DENG Z, HAO S, ZHOU S, et al. Toward Fast and Accurate Vehicle Detection in Aerial Images Using Coupled Region-Based Convolutional Neural Networks[J]. IEEE J Sel Topics Appl Earth Observ Remote Sens, 2017, 10(8):3652-3664
[9] LOGOGLU K B, et al. Feature-Based Efficient Moving Object Detection for Low-Altitude Aerial Platforms[C]//IEEE International Conference on Computer Vision Workshops, 2017:2119-2128
[10] TANG T, ZHOU S, DENG Z, et al. Vehicle Detection in Aerial Images Based on Region Convolutional Neural Networks and Hard Negative Example Mining[J]. Sensors, 2017, 17(2):336-352 |
|
|
|
|
|
|
|
