首页 > 论文 > 光学学报 > 38卷 > 12期(pp:1215008--1)

场景耦合的空对地多任务遥感影像智能检测算法

Scene-Coupled Intelligent Multi-Task Detection Algorithm for Air-to-Ground Remote Sensing Image

  • 摘要
  • 论文信息
  • 参考文献
  • 被引情况
  • PDF全文
分享:

摘要

在空对地遥感检测中,目标所占视场比例小、视角单一、易受背景干扰且视场高度变化大,这给传统深度学习检测算法带来了挑战。针对该问题,提出一种场景耦合的多任务目标检测算法。首先,设计了一种新的场景耦合目标检测网络结构,将场景分类特征图和目标检测特征图在同一尺度上进行镜像融合,丰富了网络特征描述的细粒度;其次,设计了差异化激活模块,实现特征通道的重要性筛选;然后,推导了多任务耦合的网络优化函数,实现了目标检测损失和场景分类损失的同步优化;最后,建立了空对地目标检测多任务数据集,对所提方法的有效性进行验证。实验证明,本文算法有效提升了空对地小目标检测的精度和稳健性,同时能够自适应不同高度的识别检测多任务需求,为空基无人平台对地智能检测提供了新的思路和方法。

Abstract

In air-to-ground remote sensing detection, the object has the characteristics of small field of view and single viewing angle, which is susceptible to background interference. At the same time, the height of the field of view varies greatly, which brings challenges to the traditional deep learning detection algorithm. To solve the problem, a scene-coupled multi-task object detection algorithm is proposed. First, a new scene-coupled object detection network structure is designed, which mirrors and fuses the scene classification feature map and the object detection feature map on the same scale to enrich the fine-grain of the feature description. Second, a differentiated activation module is designed to realize the importance screening of feature channels. Then, the optimization function of multi-task coupling is derived, which can simultaneously optimize the scene classification loss and object detection loss. Finally, an air-to-ground detection multi-task dataset is established to verify the effectiveness of proposed method. The experimental results show that the proposed algorithm effectively improves the accuracy and robustness of air-to-ground small object detection, and can adapt to different heights to identify multi-task requirements, which provides a new idea and method for space-based unmanned platform intelligent detection.

Newport宣传-MKS新实验室计划
补充资料

中图分类号:TP751.1

DOI:10.3788/aos201838.1215008

所属栏目:机器视觉

基金项目:国家自然科学基金(61673017)、陕西省自然基金(2017JM6077)、陕西省重点研发计划项目(2018ZDXM-GY-039)

收稿日期:2018-06-28

修改稿日期:2018-07-19

网络出版日期:2018-08-07

作者单位    点击查看

刘星:火箭军工程大学导弹工程学院, 陕西 西安 710025
陈坚:火箭军工程大学导弹工程学院, 陕西 西安 710025
杨东方:火箭军工程大学导弹工程学院, 陕西 西安 710025
贺浩:火箭军工程大学导弹工程学院, 陕西 西安 710025

联系人作者:杨东方(yangdf301@163.com)

【1】Lindeberg T. Scale invariant feature transform[J]. Scholarpedia, 2012, 7(5): 10491.

【2】Dalal N, Triggs B. Histograms of oriented gradients for Human detection[C]∥2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR''05), 2005: 886-893.

【3】Bay H, Ess A, Tuytelaars T, et al. Speeded-up robust features (SURF)[J]. Computer Vision and Image Understanding, 2008, 110(3): 346-359.

【4】Krizhevsky A, Sutskever I, Hinton G E. ImageNet classification with deep convolutional neural networks[J]. Communications of the ACM, 2017, 60(6): 84-90.

【5】Pfister T, Simonyan K, Charles J, et al. Deep convolutional neural networks for efficient pose estimation in gesture videos[C]∥Asian Conference on Computer Vision, 2014: 538-552.

【6】Szegedy C, Liu W, Jia Y Q, et al. Going deeper with convolutions[C]∥2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2015: 1-9.

【7】He K M, Zhang X Y, Ren S Q, et al. Deep residual learning for image recognition[C]∥2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2016: 770-778.

【8】Howard A G, Zhu M, Chen B, et al. MobileNets: efficient convolutional neural networks for mobile vision applications[J]. arXiv preprint arXiv: 1704.04861, 2017.

【9】Girshick R, Donahue J, Darrell T, et al. Rich feature hierarchies for accurate object detection and semantic segmentation[C]∥2014 IEEE Conference on Computer Vision and Pattern Recognition, 2014: 580-587.

【10】Girshick R. Fast R-CNN[C]∥2015 IEEE International Conference on Computer Vision (ICCV), 2015: 1440-1448.

【11】Ren S Q, He K M, Girshick R, et al. Faster R-CNN: towards real-time object detection with region proposal networks[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017, 39(6): 1137-1149.

【12】Redmon J, Divvala S, Girshick R, et al. You only look once: unified, real-time object detection[C]∥2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2016: 779-788.

【13】Liu W, Anguelov D, Erhan D, et al. SSD: Single shot multibox detector[C]∥European Conference on Computer Vision, 2016: 21-37.

【14】Deng Z P, Sun H, Zhou S L, et al. Toward fast and accurate vehicle detection in aerial images using coupled region-based convolutional neural networks[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2017, 10(8): 3652-3664.

【15】Xin P, Xu Y L, Tang H, et al. Fast airplane detection based on multi-layer feature fusion of fully convolutional networks[J]. Acta Optica Sinica, 2018, 38(3): 0315003.
辛鹏, 许悦雷, 唐红, 等. 全卷积网络多层特征融合的飞机快速检测[J]. 光学学报, 2018, 38(3): 0315003.

【16】Hou Y Q Y, Quan J C, Wei Y M. Valid aircraft detection system for remote sensing images based on cognitive models[J]. Acta Optica Sinica, 2018, 38(1): 0111005.
侯宇青阳, 全吉成, 魏湧明. 基于认知模型的遥感图像有效飞机检测系统[J]. 光学学报, 2018, 38(1): 0111005.

【17】Liu D W, Han L, Han X Y. High spatial resolution remote sensing image classification based on deep learning[J]. Acta Optica Sinica, 2016, 36(4): 0428001.
刘大伟, 韩玲, 韩晓勇. 基于深度学习的高分辨率遥感影像分类研究[J]. 光学学报, 2016, 36(4): 0428001.

【18】Chu X, Ouyang W, Yang W, et al. Multi-task recurrent neural network for immediacy prediction[C]∥2015 IEEE International Conference on Computer Vision (ICCV), 2015: 3352-3360.

【19】Long M, Wang J. Learning multiple tasks with deep relationship networks[J]. Computer Science, 2017, arXiv: 1506.02117v1.

【20】Teichmann M, Weber M, Zoellner M, et al. MultiNet: real-time joint semantic reasoning for autonomous driving[J]. Computer Vision and Pattern Recognition, 2018, arXiv: 1612.07695.

【21】Li Z, Zhou F. FSSD: feature fusion single shot multibox detector[J]. Computer Vision and Pattern Recognition, 2017, arXiv: 1712.00960.

【22】Lin T Y, Dollár P, Girshick R, et al. Feature pyramid networks for object detection[C]∥2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2017: 936-944.

【23】Hu J, Shen L, Sun G. Squeeze-and-excitation networks[J]. Computer Vision and Pattern Recognition, 2018, arXiv: 1709.01507.

【24】Liu K, Mattyus G. Fast multiclass vehicle detection on aerial images[J]. IEEE Geoscience and Remote Sensing Letters, 2015, 12(9): 1938-1942.

【25】Dagan I, Glickman O, Magnini B. The PASCAL recognising textual entailment challenge[M]. Heidelberg: Springer, 2006: 177-190.

引用该论文

Liu Xing,Chen Jian,Yang Dongfang,He Hao. Scene-Coupled Intelligent Multi-Task Detection Algorithm for Air-to-Ground Remote Sensing Image[J]. Acta Optica Sinica, 2018, 38(12): 1215008

刘星,陈坚,杨东方,贺浩. 场景耦合的空对地多任务遥感影像智能检测算法[J]. 光学学报, 2018, 38(12): 1215008

您的浏览器不支持PDF插件,请使用最新的(Chrome/Fire Fox等)浏览器.或者您还可以点击此处下载该论文PDF