基于候选区域列举的红外行人检测研究
|
摘要
针对红外图像中人体成像往往不均匀的问题,提出一类基于图分割和先验知识层级合并的红外行人检测算法.该方法将图像看作一个无向图,图中边的权值则对应所连接连个像素的灰度差异,通过求解最小生成树的方式得到图像中的连通区域.在分析红外行人成像区域特点的基础上,以先验知识为基础对连通区域进行层次合并,得到图像中目标的列举窗口.构建列举区域的积分通道特征进行分类完成行人检测任务.实验结果表明,所研究的方法可以稳健地对各种成像条件下的红外序列进行行人检测.
In the infrared thermal imaging system,the human body regions are often imaging uneven.To address this problem,a graph segmentation and prior information hierarchical merging strategy based infrared pedestrian detection method is proposed.In this method,the image is firstly treated as an undirected graph;the weights of edges in this graph are corresponding to the gray difference of connected pixels. And we obtain the connected regions in this image by constructing the minimum spanning tree of the graph.On the basis of analyzing the imaging characteristics of pedestrian in infrared system,the connected regions are combined hierarchically based on prior knowledge to generate the object windows.The integrated channel features of each proposal window are classified as pedestrian or not.The experimental results show that the proposed method can robustly detect pedestrians in various imaging situations.
In the infrared thermal imaging system,the human body regions are often imaging uneven.To address this problem,a graph segmentation and prior information hierarchical merging strategy based infrared pedestrian detection method is proposed.In this method,the image is firstly treated as an undirected graph;the weights of edges in this graph are corresponding to the gray difference of connected pixels. And we obtain the connected regions in this image by constructing the minimum spanning tree of the graph.On the basis of analyzing the imaging characteristics of pedestrian in infrared system,the connected regions are combined hierarchically based on prior knowledge to generate the object windows.The integrated channel features of each proposal window are classified as pedestrian or not.The experimental results show that the proposed method can robustly detect pedestrians in various imaging situations.
引文
[1]WOJEK C,DOLLAR P,SCHIELE B,et al. Pedestrian detection:an evaluation of the state of the art[J]. IEEE Transactions on Pattern Analysis&Machine Intelligence,2012,34(4):743-761.
[2]MARIN J,VAZQUEZ D,LOPEZ A M,et al. Random forests of local experts for pedestrian detection[C]//International conference on computer vision. IEEE,2013:2592-2599.
[3]WANG G,LIU Q,WU Q. Far-infrared pedestrian detection for advanced driver assistance systems using scene context[J].Optical Engineering,2016,55(4):043105.
[4]GERóNIMO D,LóPEZ A M,SAPPA A D,et al. Survey of pedestrian detection for advanced driver assistance systems[J].IEEE Transactions on Software Engineering,2010,32(7):1239-1258.
[5]JOHN V,MITA S,LIU Z,et al. Pedestrian detection in thermal images using adaptive fuzzy C-means clustering and convolutional neural networks[C]//Iapr International Conference on Machine Vision Applications.Japan:Spring,2015:246-249.
[6]魏丽,丁萌,曾丽君.红外图像中基于似物性与稀疏编码的行人检测[J].红外技术,2016,38(9):752-757.
[7]王国华,刘琼,庄家俊.基于局部特征的车载红外行人检测方法研究[J].电子学报,2015,43(7):1444-1448.
[8]O’MALLEY R,JONES E,GLAVIN M. Detection of pedestrians in far-infrared automotive night vision using region-growing and clothing distortion compensation[J]. Infrared Physics&Technology,2010,53(6):439-449.
[9]HURNEY P,WALDRON P,MORGAN F,et al. Night-time pedestrian classification with histograms of oriented gradientslocal binary patterns vectors[J]. IET Intelligent Transport Systems,2014,9(1):75-85.
[10]QI B,JOHN V,LIU Z,et al. Pedestrian detection from thermal images:a sparse representation based approach[J]. Infrared Physics&Technology,2016,76:157-167.
[11]SUN H,WANG C,WANG B,et al. Pyramid binary pattern features for real-time pedestrian detection from infrared videos[J]. Neurocomputing,2011,74(5):797-804.
[12]FELZENSZWALB P F,HUTTENLOCHER D P. Efficient graph-based image segmentation[J]. International Journal of Computer Vision,2004,59(2):167-181.
[13]UIJLINGS J R,SANDE K E,GEVERS T,et al. Selective search for object recognition[J]. International Journal of Computer Vision,2013,104(2):154-171.
[14]PIOTR D,ZHUO W T,PIETRO P,et al. Integral channel features[C]//Proceedings of british machine vision conference,London. BMVA Press,2009:7-11.
[15]APPEL R,FUCHS T,DOLLáR P,et al. Quickly boosting decision trees-pruning underachieving features early[C]//International conference on machine learning.Atlanta:Spring,2013:594-602.
[2]MARIN J,VAZQUEZ D,LOPEZ A M,et al. Random forests of local experts for pedestrian detection[C]//International conference on computer vision. IEEE,2013:2592-2599.
[3]WANG G,LIU Q,WU Q. Far-infrared pedestrian detection for advanced driver assistance systems using scene context[J].Optical Engineering,2016,55(4):043105.
[4]GERóNIMO D,LóPEZ A M,SAPPA A D,et al. Survey of pedestrian detection for advanced driver assistance systems[J].IEEE Transactions on Software Engineering,2010,32(7):1239-1258.
[5]JOHN V,MITA S,LIU Z,et al. Pedestrian detection in thermal images using adaptive fuzzy C-means clustering and convolutional neural networks[C]//Iapr International Conference on Machine Vision Applications.Japan:Spring,2015:246-249.
[6]魏丽,丁萌,曾丽君.红外图像中基于似物性与稀疏编码的行人检测[J].红外技术,2016,38(9):752-757.
[7]王国华,刘琼,庄家俊.基于局部特征的车载红外行人检测方法研究[J].电子学报,2015,43(7):1444-1448.
[8]O’MALLEY R,JONES E,GLAVIN M. Detection of pedestrians in far-infrared automotive night vision using region-growing and clothing distortion compensation[J]. Infrared Physics&Technology,2010,53(6):439-449.
[9]HURNEY P,WALDRON P,MORGAN F,et al. Night-time pedestrian classification with histograms of oriented gradientslocal binary patterns vectors[J]. IET Intelligent Transport Systems,2014,9(1):75-85.
[10]QI B,JOHN V,LIU Z,et al. Pedestrian detection from thermal images:a sparse representation based approach[J]. Infrared Physics&Technology,2016,76:157-167.
[11]SUN H,WANG C,WANG B,et al. Pyramid binary pattern features for real-time pedestrian detection from infrared videos[J]. Neurocomputing,2011,74(5):797-804.
[12]FELZENSZWALB P F,HUTTENLOCHER D P. Efficient graph-based image segmentation[J]. International Journal of Computer Vision,2004,59(2):167-181.
[13]UIJLINGS J R,SANDE K E,GEVERS T,et al. Selective search for object recognition[J]. International Journal of Computer Vision,2013,104(2):154-171.
[14]PIOTR D,ZHUO W T,PIETRO P,et al. Integral channel features[C]//Proceedings of british machine vision conference,London. BMVA Press,2009:7-11.
[15]APPEL R,FUCHS T,DOLLáR P,et al. Quickly boosting decision trees-pruning underachieving features early[C]//International conference on machine learning.Atlanta:Spring,2013:594-602.