尺度约束辅助的空对地目标智能检测方法
|
摘要
受视距远、视差小、目标特征单一和背景复杂等因素的影响,空基无人平台对地目标检测作为智能无人平台领域研究的难点问题,得到了越来越多的关注。利用传统的基于深度学习的目标检测算法容易出现错检和漏检,对此,利用单一观测视角下的同类目标成像一致性,定义了空对地区域重叠度(insection of unit,IOU)损失函数,实现了序贯图像同类目标之间的相关性表示;此外,利用空对地场景下目标之间的相关性,建立了基于朴素贝叶斯判据的目标尺度约束辅助检测模型,以提高目标检测的鲁棒性。最后基于公共数据集和自有无人机平台飞行数据,进行了空对地典型目标的检测实验,检测结果证明了上述方法的有效性。
Due to the factors such as long viewing distance,small parallax,single target feature,and complex background,the ground-based object detection by space-based unmanned platforms has attracted more and more attention as a difficult problem in the field of intelligent unmanned platforms. The use of object detection algorithms based on traditional deep learning is prone to be an error detection easily. By using the same target imaging consistency introduced by a single observational perspective,an insection of unit( IOU) loss function is defined,and a correlation between similar objects of the sequential image is realized. Using the correlation between objects in air-to-ground scenarios,a object scale constraint aided detection model based on the naive Bayes criterion is established to improve the robustness of detection.Finally,based on the common dataset and our UAV platform,We conduct air-to-ground typical object detection experiments. The results demonstrate the effectiveness of the above method.
Due to the factors such as long viewing distance,small parallax,single target feature,and complex background,the ground-based object detection by space-based unmanned platforms has attracted more and more attention as a difficult problem in the field of intelligent unmanned platforms. The use of object detection algorithms based on traditional deep learning is prone to be an error detection easily. By using the same target imaging consistency introduced by a single observational perspective,an insection of unit( IOU) loss function is defined,and a correlation between similar objects of the sequential image is realized. Using the correlation between objects in air-to-ground scenarios,a object scale constraint aided detection model based on the naive Bayes criterion is established to improve the robustness of detection.Finally,based on the common dataset and our UAV platform,We conduct air-to-ground typical object detection experiments. The results demonstrate the effectiveness of the above method.
引文
[1] KEMBHAVI A,HARWOOD D,DAVIS L S. Vehicle Detection Using Partial Least Squares[J]. IEEE Transactions on Pattern Analysis&Machine Intelligence,2011,33(6):1250-1265.
[2] SHAO W,YANG W,LIU G,et al. Car Detection from High-Resolution Aerial Imagery Using Multiple Features[C]∥Geoscience and Remote Sensing Symposium.IEEE,2012:4379-4382.
[3] LIU K,MATTYUS G. Fast Multiclass Vehicle Detection on Aerial Images[J]. IEEE Geoscience&Remote Sensing Letters,2015,12(9):1938-1942.
[4] CHEN Z,WANG C,WEN C,et al. Vehicle Detection in High-Resolution Aerial Images via Sparse Representation and Superpixels[J]. IEEE Transactions on Geoscience&Remote Sensing,2015,54(1):103-116.
[5] LINDEBERG T. Scale Invariant Feature Transform[M].Scholarpedia,2012:2012-2021.
[6] DALAL N,TRIGGS B. Histograms of Oriented Gradients for Human Detection[C]∥IEEE Computer Society,2005:886-893.
[7] BAY H,ESS A,TUYTELAARS T,et al. Speeded-up Robust Features(SURF)[J]. Computer Vision and Image Understanding,2008,110(3):346-359.
[8] GIRSHICK R,DONAHUE J,DARRELL T,et al. Rich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation[C]∥Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition,2014:580-587.
[9] GIRSHICK R. Fast R-CNN[C]∥IEEE International Conference on Computer Vision,IEEE,2015:1440-1448.
[10] REN S,HE K,GIRSHICK R,et al. Faster R-CNN:Towards Real-Time Object Detection with Region Proposal Networks[C]∥International Conference on Neural Information Processing Systems,MIT Press,2015:91-99.
[11] REDMON J,DIVVALA S,GIRSHICK R,et al. You only Look Once:Unified,Real-Time Object Detection[C]∥Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2016:779-788.
[12] LIU W,ANGUELOV D,ERHAN D,et al. SSD:Single Shot MultiBox Detector[C]∥European Conference on Computer Vision. Springer International Publishing,2016:21-37.
[13] KONONENKO I. Semi-Naive Bayesian Classifier[C]∥European Working Session on Learning Springer,Berlin,Heidelberg,1991:206-219.
[14] LIU K,MATTYUS G. Fast Multiclass Vehicle Detection on Aerial Images[J]. IEEE Geoscience&Remote Sensing Letters,2015,12(9):1938-1942.
[15] EVERINGHAM M,WINN J. The PASCAL Visual Object Classes Challenge 2010(VOC2010)Development Kit Contents[C]∥International Conference on Machine Learning Challenges:Evaluating Predictive Uncertainty Visual Object Classification. Springer-Verlag,2011:117-176.
[16] KINGA D,ADAM J B. A Method for Stochastic Optimization[J]. Arxiv Preprint Arxiv:1412. 6980,2014.
[17] LIN T Y,MAIRE M,BELONGIE S,et al. Microsoft COCO:Common Objects in Context[C]∥European Conference on Computer Vision,Springer,Cham,2014:740-755.
[2] SHAO W,YANG W,LIU G,et al. Car Detection from High-Resolution Aerial Imagery Using Multiple Features[C]∥Geoscience and Remote Sensing Symposium.IEEE,2012:4379-4382.
[3] LIU K,MATTYUS G. Fast Multiclass Vehicle Detection on Aerial Images[J]. IEEE Geoscience&Remote Sensing Letters,2015,12(9):1938-1942.
[4] CHEN Z,WANG C,WEN C,et al. Vehicle Detection in High-Resolution Aerial Images via Sparse Representation and Superpixels[J]. IEEE Transactions on Geoscience&Remote Sensing,2015,54(1):103-116.
[5] LINDEBERG T. Scale Invariant Feature Transform[M].Scholarpedia,2012:2012-2021.
[6] DALAL N,TRIGGS B. Histograms of Oriented Gradients for Human Detection[C]∥IEEE Computer Society,2005:886-893.
[7] BAY H,ESS A,TUYTELAARS T,et al. Speeded-up Robust Features(SURF)[J]. Computer Vision and Image Understanding,2008,110(3):346-359.
[8] GIRSHICK R,DONAHUE J,DARRELL T,et al. Rich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation[C]∥Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition,2014:580-587.
[9] GIRSHICK R. Fast R-CNN[C]∥IEEE International Conference on Computer Vision,IEEE,2015:1440-1448.
[10] REN S,HE K,GIRSHICK R,et al. Faster R-CNN:Towards Real-Time Object Detection with Region Proposal Networks[C]∥International Conference on Neural Information Processing Systems,MIT Press,2015:91-99.
[11] REDMON J,DIVVALA S,GIRSHICK R,et al. You only Look Once:Unified,Real-Time Object Detection[C]∥Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2016:779-788.
[12] LIU W,ANGUELOV D,ERHAN D,et al. SSD:Single Shot MultiBox Detector[C]∥European Conference on Computer Vision. Springer International Publishing,2016:21-37.
[13] KONONENKO I. Semi-Naive Bayesian Classifier[C]∥European Working Session on Learning Springer,Berlin,Heidelberg,1991:206-219.
[14] LIU K,MATTYUS G. Fast Multiclass Vehicle Detection on Aerial Images[J]. IEEE Geoscience&Remote Sensing Letters,2015,12(9):1938-1942.
[15] EVERINGHAM M,WINN J. The PASCAL Visual Object Classes Challenge 2010(VOC2010)Development Kit Contents[C]∥International Conference on Machine Learning Challenges:Evaluating Predictive Uncertainty Visual Object Classification. Springer-Verlag,2011:117-176.
[16] KINGA D,ADAM J B. A Method for Stochastic Optimization[J]. Arxiv Preprint Arxiv:1412. 6980,2014.
[17] LIN T Y,MAIRE M,BELONGIE S,et al. Microsoft COCO:Common Objects in Context[C]∥European Conference on Computer Vision,Springer,Cham,2014:740-755.