Faster-RCNN和Level-Set结合的高分遥感影像建筑物提取
|
摘要
目前Level-Set图像分割方法存在初始轮廓的确定受人为因素影响较大的问题,对目标被遮盖和目标与背景灰度值相近无法达到理想的分割效果。针对此问题,本文提出了利用Faster-RCNN网络模型确定目标初始轮廓和区域信息的先验水平集图像分割方法,搭建Caffe深度学习框架训练Faster-RCNN网络模型;通过有监督学习的方式在IAILD数据集上训练模型,检测出目标建筑物并初步提取建筑物的轮廓,并将其与形状先验的Level-Set算法结合。对比实验结果表明,本文方法解决了Level-Set算法中图像分割结果初始轮廓受人为标记框选的影响较大的问题,能够更好地完成被遮挡建筑物的分割,对于目标建筑和背景灰度值相近也能达到更好的分割效果。
At present,the Level-Set image segmentation method has the problem that the initial contour is greatly affected by human factors.The target segmentation and the target and background gray values are close to each other,and the ideal segmentation effect cannot be achieved.Aiming at this problem,this paper proposes a priori level set image segmentation method based on the Faster-RCNN network model to determine the initial contour and region information of the target,and builds the Caffe deep learning framework to train the Faster-RCNN network model.The IAILD data is obtained through supervised learning.The training model is assembled,the target building is detected and the outline of the building is initially extracted and combined with the shape-priority Level-Set algorithm.The experimental results show that the proposed method solves the problem that the initial contour of the image segmentation result is greatly influenced by the artificial marker frame selection in the Level-Set algorithm.It can better complete the segmentation of the occluded building,and achieve better segmentation effect for the target building and the background gray value.
At present,the Level-Set image segmentation method has the problem that the initial contour is greatly affected by human factors.The target segmentation and the target and background gray values are close to each other,and the ideal segmentation effect cannot be achieved.Aiming at this problem,this paper proposes a priori level set image segmentation method based on the Faster-RCNN network model to determine the initial contour and region information of the target,and builds the Caffe deep learning framework to train the Faster-RCNN network model.The IAILD data is obtained through supervised learning.The training model is assembled,the target building is detected and the outline of the building is initially extracted and combined with the shape-priority Level-Set algorithm.The experimental results show that the proposed method solves the problem that the initial contour of the image segmentation result is greatly influenced by the artificial marker frame selection in the Level-Set algorithm.It can better complete the segmentation of the occluded building,and achieve better segmentation effect for the target building and the background gray value.
引文
[1]刘纯,洪亮,楚森寐,等.高分辨率遥感影像分割方法研究[J].测绘与空间地理信息,2014,(10):44-49.LIU C,HONG L,CHU S M,et al.Research on high resolution remote sensing image segmentation method[J].Mapping and Spatial Geographic Information,2014,(10):44-49.
[2]郑东玉.多尺度分割框架下的面向对象高分辨率遥感影像变化检测[D].成都:西南交通大学,2018.ZHENG D Y.Object-oriented high resolution remote sensing image change detection in multi-scale segmentation framework[D].Chengdu:Southwest Jiaotong University,2018.(in Chinese)
[3]陶超,谭毅华,蔡华杰,等.面向对象的高分辨率遥感影像城区建筑物分级提取方法[J].测绘学报,2010,39(01):39-45.TAO C,TAN Y H,CAI H J,et al.Object-oriented hierarchical extraction of urban buildings from high resolution remote sensing images[J].Journal of Surveying and Mapping,2010,39(1):39-45.
[4]KARANTZALOS K,PARAGIOS N.Recognition-driven two-dimensional competing priors toward automatic and accurate building detection[J].IEEE Transactions on Geoscience and Remote Sensing,2009,47(1):133-144.
[5]梁华,宋玉龙,钱锋,等.基于深度学习的航空对地小目标检测[J].液晶与显示,2018,33(09):793-800.LIANG H,SONG Y L,QIAN F,et al.Aeronautical small target detection to ground based on deep learning[J].Chinese Journal of Liquid Crystals and Displays,2018,33(09):793-800.
[6]周筑博,高佼,张巍,等.基于深度卷积神经网络的输电线路可见光图像目标检测[J].液晶与显示,2018,33(04):317-325.ZHOU Z B,GAO J,ZHANG W,et al.Target detection of transmission line visible image based on deep convolution neural network[J].Chinese Journal of Liquid Crystals and Displays,2018,33(04):317-325.
[7]李宇,刘雪莹,张洪群,等.基于卷积神经网络的光学遥感图像检索[J].光学精密工程,2018,26(01):200-207.LI Y,LIU X Y,ZHANG H Q,et al.Optical remote sensing image retrieval based on convolutional neural network[J].Optics and Precision Engineering,2018,26(01):200-207.
[8]李伟山,卫晨,王琳.改进的Faster RCNN煤矿井下行人检测算法[J/OL].计算机工程与应用:1-16[2018-11-25].http://kns.cnki.net/kcms/detail/11.2127.TP.20180522.0944.002.html.LI W S,WEI C,WANG L.Improved faster rcnn algorithm for underground pedestrian detection in coal mine[J/OL].Computer engineering and application:1-16[2018-11-25].http://kns.cnki.net/kcms/detail/11.2127.TP.20180522.0944.002.html.
[9]何传江,李梦,詹毅.用于图像分割的自适应距离保持水平集演化[J].软件学报,2008,19(12):3161-3169.HE C J,LI M,ZHAN Y.Evolution of adaptive distance-preserving level set for image segmentation[J].Journal of Software,2008,19(12):3161-3169.
[10]耿秀秀,潘振宽,魏伟波,等.整体曲率变分水平集方法的曲面去噪模型[J].计算机工程与应用,2008,44(35):185-187.GENG X X,PAN Z K,WEI W B,et al.Surface denoising model of global curvature variational level set method[J].Computer Engineering and Applications,2008,44(35):185-187.
[11]鲁圆圆,强静仁,汪朝.基于改进CV模型的图像分割算法[J/OL].现代电子技术,2018(21):71-75[2018-11-18].LU Y Y,QIANG J R,WANG C.Image segmentation algorithm based on improved CV model[J/OL].Modern electronic technology,2018(21):71-75[2018-11-18].
[12]席志红,赵春梅.基于变分水平集理论的水下图像分割方法[J/OL].应用科技:1-6[2018-11-25].http://kns.cnki.net/kcms/detail/23.1191.U.20181023.1520.002.html.XI Z H,ZHAO C M.Underwater image segmentation method based on variational level set theory[J/OL].Applied science and technology:1-6[2018-11-25].http://kns.cnki.net/kcms/detail/23.1191.U.20181023.1520.002.html.
[13]唐利明,王洪珂,陈照辉,等.基于变分水平集的图像模糊聚类分割[J].软件学报,2014,25(07):1570-1582.TANG L M,WANG H K,CHEN Z H,et al.Image fuzzy clustering segmentation based on variational level sets[J].Journal of Software,2014(07):1570-1582.
[14]ZHANG J H,KONG F T,WU J Z,et al.Automatic image segmentation method for cotton leaves with disease under natural environment[J].Journal of Integrative Agriculture,2018,17(08):1800-1814.
[15]戴陈卡,李毅.基于Faster RCNN以及多部件结合的机场场面静态飞机检测[J].计算机应用,2017,37(S2):85-88.DAI C K,LI Y.Airfield static aircraft detection based on faster rcnn and multi-component combination[J].Computer Applications,2017,37(S2):85-88.
[16]IVANOVSKA T,LAQUA R,WANG L,et al.Fast Implementations of the Levelset segmentation method with bias field correction in MR images:full domain and mask-based versions[C]//Proceedings of the 6th Iberian Conference on Pattern Recognition and Image Analysis.Funchal,Madeira,Portugal:Springer,2013:674-681.
[2]郑东玉.多尺度分割框架下的面向对象高分辨率遥感影像变化检测[D].成都:西南交通大学,2018.ZHENG D Y.Object-oriented high resolution remote sensing image change detection in multi-scale segmentation framework[D].Chengdu:Southwest Jiaotong University,2018.(in Chinese)
[3]陶超,谭毅华,蔡华杰,等.面向对象的高分辨率遥感影像城区建筑物分级提取方法[J].测绘学报,2010,39(01):39-45.TAO C,TAN Y H,CAI H J,et al.Object-oriented hierarchical extraction of urban buildings from high resolution remote sensing images[J].Journal of Surveying and Mapping,2010,39(1):39-45.
[4]KARANTZALOS K,PARAGIOS N.Recognition-driven two-dimensional competing priors toward automatic and accurate building detection[J].IEEE Transactions on Geoscience and Remote Sensing,2009,47(1):133-144.
[5]梁华,宋玉龙,钱锋,等.基于深度学习的航空对地小目标检测[J].液晶与显示,2018,33(09):793-800.LIANG H,SONG Y L,QIAN F,et al.Aeronautical small target detection to ground based on deep learning[J].Chinese Journal of Liquid Crystals and Displays,2018,33(09):793-800.
[6]周筑博,高佼,张巍,等.基于深度卷积神经网络的输电线路可见光图像目标检测[J].液晶与显示,2018,33(04):317-325.ZHOU Z B,GAO J,ZHANG W,et al.Target detection of transmission line visible image based on deep convolution neural network[J].Chinese Journal of Liquid Crystals and Displays,2018,33(04):317-325.
[7]李宇,刘雪莹,张洪群,等.基于卷积神经网络的光学遥感图像检索[J].光学精密工程,2018,26(01):200-207.LI Y,LIU X Y,ZHANG H Q,et al.Optical remote sensing image retrieval based on convolutional neural network[J].Optics and Precision Engineering,2018,26(01):200-207.
[8]李伟山,卫晨,王琳.改进的Faster RCNN煤矿井下行人检测算法[J/OL].计算机工程与应用:1-16[2018-11-25].http://kns.cnki.net/kcms/detail/11.2127.TP.20180522.0944.002.html.LI W S,WEI C,WANG L.Improved faster rcnn algorithm for underground pedestrian detection in coal mine[J/OL].Computer engineering and application:1-16[2018-11-25].http://kns.cnki.net/kcms/detail/11.2127.TP.20180522.0944.002.html.
[9]何传江,李梦,詹毅.用于图像分割的自适应距离保持水平集演化[J].软件学报,2008,19(12):3161-3169.HE C J,LI M,ZHAN Y.Evolution of adaptive distance-preserving level set for image segmentation[J].Journal of Software,2008,19(12):3161-3169.
[10]耿秀秀,潘振宽,魏伟波,等.整体曲率变分水平集方法的曲面去噪模型[J].计算机工程与应用,2008,44(35):185-187.GENG X X,PAN Z K,WEI W B,et al.Surface denoising model of global curvature variational level set method[J].Computer Engineering and Applications,2008,44(35):185-187.
[11]鲁圆圆,强静仁,汪朝.基于改进CV模型的图像分割算法[J/OL].现代电子技术,2018(21):71-75[2018-11-18].LU Y Y,QIANG J R,WANG C.Image segmentation algorithm based on improved CV model[J/OL].Modern electronic technology,2018(21):71-75[2018-11-18].
[12]席志红,赵春梅.基于变分水平集理论的水下图像分割方法[J/OL].应用科技:1-6[2018-11-25].http://kns.cnki.net/kcms/detail/23.1191.U.20181023.1520.002.html.XI Z H,ZHAO C M.Underwater image segmentation method based on variational level set theory[J/OL].Applied science and technology:1-6[2018-11-25].http://kns.cnki.net/kcms/detail/23.1191.U.20181023.1520.002.html.
[13]唐利明,王洪珂,陈照辉,等.基于变分水平集的图像模糊聚类分割[J].软件学报,2014,25(07):1570-1582.TANG L M,WANG H K,CHEN Z H,et al.Image fuzzy clustering segmentation based on variational level sets[J].Journal of Software,2014(07):1570-1582.
[14]ZHANG J H,KONG F T,WU J Z,et al.Automatic image segmentation method for cotton leaves with disease under natural environment[J].Journal of Integrative Agriculture,2018,17(08):1800-1814.
[15]戴陈卡,李毅.基于Faster RCNN以及多部件结合的机场场面静态飞机检测[J].计算机应用,2017,37(S2):85-88.DAI C K,LI Y.Airfield static aircraft detection based on faster rcnn and multi-component combination[J].Computer Applications,2017,37(S2):85-88.
[16]IVANOVSKA T,LAQUA R,WANG L,et al.Fast Implementations of the Levelset segmentation method with bias field correction in MR images:full domain and mask-based versions[C]//Proceedings of the 6th Iberian Conference on Pattern Recognition and Image Analysis.Funchal,Madeira,Portugal:Springer,2013:674-681.