基于多方向显著子区域置信表决的多目标检测
|
摘要
多目标检测是智能交通监控系统的一项重要和挑战性的工作。提出一种基于多方向显著子区域置信表决的多目标检测方法。利用Gabor滤波提取目标四方向显著特征像素;根据多阈值条件确定每个目标的有效子区域及尺度;判断有效交集子区域和交集状态;加权融合交集面积占空比和有效交集子区域相对面积占空比,获得各个有效交集子区域置信表决系数;根据最大置信表决系数获得目标最优检测。采用实际交通场景采集的视频图像序列进行实验,结果表明该方法在邻近目标交互干扰和遮挡、目标阴影、非均匀光照及背景中相似色彩干扰等情况下,均具有鲁棒、准确的检测性能。
Multi-object detection under the smart traffic surveillance system is an important and challenging task. This paper presents a multi-object detection algorithm based on multi-orientation saliency subregion voting by confidence. Fourorientation saliency feature pixels are firstly extracted by Gabor filters. The valid subregions and scales of each object are further confirmed in terms of the multi-threshold criteria. And then valid intersection subregions and intersection states are confirmed. The duty ratios of intersection area and the relative duty ratios of valid intersection subregions are weighted and integrated to acquire the confidence vote coefficient for every subregion. The optimal detection is finally derived according to the maximum confidence vote. Experiments conducted on real traffic image sequences demonstrate that the proposed method is robust and accurate in multi-object detection under mutual disturbance and occlusion among adjacent objects, object shadow, non-uniform illumination and similar color disturbance in the background etc.
Multi-object detection under the smart traffic surveillance system is an important and challenging task. This paper presents a multi-object detection algorithm based on multi-orientation saliency subregion voting by confidence. Fourorientation saliency feature pixels are firstly extracted by Gabor filters. The valid subregions and scales of each object are further confirmed in terms of the multi-threshold criteria. And then valid intersection subregions and intersection states are confirmed. The duty ratios of intersection area and the relative duty ratios of valid intersection subregions are weighted and integrated to acquire the confidence vote coefficient for every subregion. The optimal detection is finally derived according to the maximum confidence vote. Experiments conducted on real traffic image sequences demonstrate that the proposed method is robust and accurate in multi-object detection under mutual disturbance and occlusion among adjacent objects, object shadow, non-uniform illumination and similar color disturbance in the background etc.
引文
[1] Yu H,Qin L,Huang Q,et al.Online multiple object tracking via exchanging object context[J].Neurocomputing,2018,292:28-37.
[2] Wang X,Turetken E,Fleuret F,et al.Tracking interacting objects using intertwined flows[J].IEEE Trans Pattern Anal Mach Intell,2016,38(11):2312-2326.
[3] Lu H,Zou W,Li H,et al.Edge and color contexts based object representation and tracking[J].Optik,2015,126(1):148-152.
[4] Plant C,Zherdin A,Sorg C,et al.Mining interaction patterns among brain regions by clustering[J].IEEE Trans Knowl Data Eng,2014,26(9):2237-2249.
[5]王松,党建武,王阳萍,等.实时动作识别方法研究[J].计算机工程与应用,2017,53(3):28-31.
[6]杨世强,弓逯琦,乔丹.人机交互装配中手部指形轮廓信息检测方法[J].计算机工程与应用,2018,54(8):226-230.
[7] St-Charles P L,Bilodeau G A,Bergevin R.Universal background subtraction using word consensus models[J].IEEE Trans Image Process,2016,25(10):4768-4781.
[8]肖乔,裴继红,王荔霞,等.基于多通道Gabor滤波模糊融合的遥感图像舰船检测[J].山东大学学报(工学版),2015,45(5):29-35.
[9] Zhang Y,Tong X,Yang T,et al.Multi-model estimation based moving object detection for aerial video[J].Sensors,2015,15(4):8214-8231.
[10] Tsai D M,Chiu W Y.Motion detection using Fourier image reconstruction[J].Pattern Recogn Lett,2008,29(16):2145-2155.
[11] Lu H,Gu K,Lin W,et al.Tracking based on stable feature mining using intraframe clustering and interframe association[J].IEEE Access,2017(5):4690-4703.
[12] Beyan C,Temizel A.Adaptive mean-shift for automated multi object tracking[J].IET Comput Vis,2012,6(1):1-12.
[13] Lipton A J,Fujiyoshi H,Patil R S.Moving target classification and tracking from real-time video[C]//Proc of IEEE Workshop on Applications of Computer Vision,1998:8-14.
[14] Denman S,Chandran V,Sridharan S.An adaptive optical flow technique for person tracking system[J].Pattern Recogn Lett,2007,28(10):1232-1239.
[15]周洋,何永健,唐向宏,等.融合双目多维感知特征的立体视频显著性检测[J].中国图象图形学报,2017,22(3):305-314.
[16] Xin B,Tian Y,Wang Y Z,et al.Background Subtraction via generalized fused lasso foreground modeling[C]//Proc of IEEE Conf on Computer Vision and Pattern Recognition,2015:4676-4684.
[17] Wu B,Nevatia R.Tracking of multiple,partially occluded humans based on static body part detection[C]//Proc of IEEE Conf on Computer Vision and Pattern Recognition,2006:951-958.
[18] Elharrouss O,Moujahid D,Tairi H.Motion detection based on the combining of the background subtraction and the structure-texture decomposition[J].Optik,2015,126(24):5992-5997.
[19] Wang X,Fang L,Cong Z.Research on real-time multitarget tracking algorithm based on MSPF[J].Acta Automatica Sinica,2012,38(1):139-144.
[20] Brenning A,Long S,Fieguth P.Detecting rock glacier flow structures using Gabor filters and IKONOS imagery[J].Remote Sens Environ,2012,125:227-237.
[21] Dahmane M,Cossette S,Meunier J.Conditional Gabor phase-based disparity estimation applied to facial tracking for person-specific facial action recognition:a preliminary study[J].Multimed Tools Appl,2015,74(17):7111-7130.
[22]卢健,黄杰,潘峰.第二小方向导数信息熵的兴趣点检测[J].中国图象图形学报,2016,21(1):8-16.
[23] Lee T S.Image representation using 2D Gabor wavelets[J].IEEE Trans Pattern Anal Mach Intell,1996,18(10):959-971.
[24] Sen D,Kankanhalli M.A bio-inspired center-surround model for salience computation in images[J].Journal of Visual Communication and Image Representation,2015,30:277-288.
[2] Wang X,Turetken E,Fleuret F,et al.Tracking interacting objects using intertwined flows[J].IEEE Trans Pattern Anal Mach Intell,2016,38(11):2312-2326.
[3] Lu H,Zou W,Li H,et al.Edge and color contexts based object representation and tracking[J].Optik,2015,126(1):148-152.
[4] Plant C,Zherdin A,Sorg C,et al.Mining interaction patterns among brain regions by clustering[J].IEEE Trans Knowl Data Eng,2014,26(9):2237-2249.
[5]王松,党建武,王阳萍,等.实时动作识别方法研究[J].计算机工程与应用,2017,53(3):28-31.
[6]杨世强,弓逯琦,乔丹.人机交互装配中手部指形轮廓信息检测方法[J].计算机工程与应用,2018,54(8):226-230.
[7] St-Charles P L,Bilodeau G A,Bergevin R.Universal background subtraction using word consensus models[J].IEEE Trans Image Process,2016,25(10):4768-4781.
[8]肖乔,裴继红,王荔霞,等.基于多通道Gabor滤波模糊融合的遥感图像舰船检测[J].山东大学学报(工学版),2015,45(5):29-35.
[9] Zhang Y,Tong X,Yang T,et al.Multi-model estimation based moving object detection for aerial video[J].Sensors,2015,15(4):8214-8231.
[10] Tsai D M,Chiu W Y.Motion detection using Fourier image reconstruction[J].Pattern Recogn Lett,2008,29(16):2145-2155.
[11] Lu H,Gu K,Lin W,et al.Tracking based on stable feature mining using intraframe clustering and interframe association[J].IEEE Access,2017(5):4690-4703.
[12] Beyan C,Temizel A.Adaptive mean-shift for automated multi object tracking[J].IET Comput Vis,2012,6(1):1-12.
[13] Lipton A J,Fujiyoshi H,Patil R S.Moving target classification and tracking from real-time video[C]//Proc of IEEE Workshop on Applications of Computer Vision,1998:8-14.
[14] Denman S,Chandran V,Sridharan S.An adaptive optical flow technique for person tracking system[J].Pattern Recogn Lett,2007,28(10):1232-1239.
[15]周洋,何永健,唐向宏,等.融合双目多维感知特征的立体视频显著性检测[J].中国图象图形学报,2017,22(3):305-314.
[16] Xin B,Tian Y,Wang Y Z,et al.Background Subtraction via generalized fused lasso foreground modeling[C]//Proc of IEEE Conf on Computer Vision and Pattern Recognition,2015:4676-4684.
[17] Wu B,Nevatia R.Tracking of multiple,partially occluded humans based on static body part detection[C]//Proc of IEEE Conf on Computer Vision and Pattern Recognition,2006:951-958.
[18] Elharrouss O,Moujahid D,Tairi H.Motion detection based on the combining of the background subtraction and the structure-texture decomposition[J].Optik,2015,126(24):5992-5997.
[19] Wang X,Fang L,Cong Z.Research on real-time multitarget tracking algorithm based on MSPF[J].Acta Automatica Sinica,2012,38(1):139-144.
[20] Brenning A,Long S,Fieguth P.Detecting rock glacier flow structures using Gabor filters and IKONOS imagery[J].Remote Sens Environ,2012,125:227-237.
[21] Dahmane M,Cossette S,Meunier J.Conditional Gabor phase-based disparity estimation applied to facial tracking for person-specific facial action recognition:a preliminary study[J].Multimed Tools Appl,2015,74(17):7111-7130.
[22]卢健,黄杰,潘峰.第二小方向导数信息熵的兴趣点检测[J].中国图象图形学报,2016,21(1):8-16.
[23] Lee T S.Image representation using 2D Gabor wavelets[J].IEEE Trans Pattern Anal Mach Intell,1996,18(10):959-971.
[24] Sen D,Kankanhalli M.A bio-inspired center-surround model for salience computation in images[J].Journal of Visual Communication and Image Representation,2015,30:277-288.