优化搜索策略的KCF目标跟踪算法
|
摘要
针对核相关滤波跟踪算法存在实时性较差的问题,提出了一种优化搜索策略的改进算法。首先,在检测到随机选取的视频某一帧中目标中心位置后,计算该目标图像块的均值和标准差。再设定一个排序队列以及两个自适应阈值来筛除一些特征与目标差异较大的候选块。在视频下一帧中,均值与标准差的差值小于设定阈值的候选块会优先检测并计算响应。实验结果表明,改进后的算法与原算法相比帧率提升可达10%左右,且跟踪精度较KCF、CSK、Struct等其它算法提升2.2%、14.4%和24.9%。
To improve the real-time performance of kernelized correlation filters(KCF)algorithm,we proposedan optimized searching strategy for KCF tracking algorithm. The center position of target in a randomly selectedvideo frame was detected,and the mean value and standard deviation of the target image patch were calculatedrespectively. Then a sort queue and two adaptive thresholds were set to discard unsuitable patches with certainfeatures that differ greatly from the target patch. It was realized that the candidate patch in the next frame wasdetected and calculated with priority when whose mean values and standard deviations were both within acertain margin of the target patch. Experimental results show that the proposed algorithm increases the framerate about 10% than the original KCF algorithm,and its tracking accuracy is about 2.2%,14.4% and 24.9%higher than that of other algorithms such as KCF,circulant structure of tracking-by-detection with Kernels andStruct,respectively.
To improve the real-time performance of kernelized correlation filters(KCF)algorithm,we proposedan optimized searching strategy for KCF tracking algorithm. The center position of target in a randomly selectedvideo frame was detected,and the mean value and standard deviation of the target image patch were calculatedrespectively. Then a sort queue and two adaptive thresholds were set to discard unsuitable patches with certainfeatures that differ greatly from the target patch. It was realized that the candidate patch in the next frame wasdetected and calculated with priority when whose mean values and standard deviations were both within acertain margin of the target patch. Experimental results show that the proposed algorithm increases the framerate about 10% than the original KCF algorithm,and its tracking accuracy is about 2.2%,14.4% and 24.9%higher than that of other algorithms such as KCF,circulant structure of tracking-by-detection with Kernels andStruct,respectively.
引文
[1]张微,康宝生.相关滤波目标跟踪进展综述[J].中国图象图形学报,2017,22(8):1017-1033.
[2]尹宏鹏,陈波,柴毅,等.基于视觉的目标检测与跟踪综述[J].自动化学报,2016,42(10):1466-1489.
[3]BOLME D S,BEVERIDGE J R,DRAPER B A,et al.Visual object tracking using adaptive correlation filters[C]//Proceedings of IEEE Conference on Computer Vision and Pattern Recognition.San Francisco:IEEE,2010:2544-2550.
[4]HENRIQUES J F,CASEIRO R,MARTINS P,et al.Exploiting the circulant structure of tracking-bydetection with kernels[C]//Proceedings of 2012European Conference on Computer Vision.Heidelberg Berlin:Springer,2012:702-715.
[5]HENRIQUES J F,CASEIRO R,MARTINS P,et al.High-speed tracking with kernelized correlation filters[J].IEEE Transactions on Pattern Analysis&Machine Intelligence,2015,37(3):583-596.
[6]DANELLJAN M,BHAT G,KHAN F S,et al.ECO:efficient convolution operators for tracking[C]//Proceedings of 2017 IEEE Conference on Computer Vision and Pattern Recognition.Honolulu:IEEE,2017:6931-6939.
[7]DANELLJAN M,BHAT G,KHAN F S,et al.Learning spatially regularized correlation filters for visual tracking[C]//Proceedings of 2015 IEEE International Conference on Computer Vision.Santiago:IEEE,2016:4310-4318.
[8]LI Y,ZHU J K,HOI S C H.Reliable patch trackers:robust visual tracking by exploiting reliable patches[C]//Proceedings of 2015 IEEE Conference on Computer Vision and Pattern Recognition.Boston:IEEE,2015:353-361.
[9]XU Y L,WANG J B,LI H,et al.Patch-based scale calculation for real-time visual tracking[J].IEEE Signal Processing Letters,2016,23(1):40-44.
[10]孙健,向伟,谭舒昆,等.改进的核相关滤波跟踪算法[J].计算机工程与应用,2018,54(9):178-182.
[11]单玲玉,闵锋,李延达.全局阈值与局部阈值相结合的视网膜血管分割方法[J].武汉工程大学学报,2015,37(3):62-67.
[12]张雷,王延杰,孙宏海,等.采用核相关滤波器的自适应尺度目标跟踪[J].光学精密工程,2016,24(2):448-459.
[13]WU Y,LIM J,YANG M H.Online object tracking:a benchmark[C]//Proceedings of IEEE Conference on Computer Vision and Pattern Recognition.Portland:IEEE,2013:2411-2418.
[14]HARE S,SAFFARI A,TORR P H S.Struct:structured output tracking with kernels[C]//Proceedings of 2011 IEEE International Conference on Computer Vision.Barcelona:IEEE,2012:263-270.
[15]KALAL Z,MATAS J,MIKOLAJCZYK K.Trackinglearning-detection[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2012,34(7):1409-1422.
[2]尹宏鹏,陈波,柴毅,等.基于视觉的目标检测与跟踪综述[J].自动化学报,2016,42(10):1466-1489.
[3]BOLME D S,BEVERIDGE J R,DRAPER B A,et al.Visual object tracking using adaptive correlation filters[C]//Proceedings of IEEE Conference on Computer Vision and Pattern Recognition.San Francisco:IEEE,2010:2544-2550.
[4]HENRIQUES J F,CASEIRO R,MARTINS P,et al.Exploiting the circulant structure of tracking-bydetection with kernels[C]//Proceedings of 2012European Conference on Computer Vision.Heidelberg Berlin:Springer,2012:702-715.
[5]HENRIQUES J F,CASEIRO R,MARTINS P,et al.High-speed tracking with kernelized correlation filters[J].IEEE Transactions on Pattern Analysis&Machine Intelligence,2015,37(3):583-596.
[6]DANELLJAN M,BHAT G,KHAN F S,et al.ECO:efficient convolution operators for tracking[C]//Proceedings of 2017 IEEE Conference on Computer Vision and Pattern Recognition.Honolulu:IEEE,2017:6931-6939.
[7]DANELLJAN M,BHAT G,KHAN F S,et al.Learning spatially regularized correlation filters for visual tracking[C]//Proceedings of 2015 IEEE International Conference on Computer Vision.Santiago:IEEE,2016:4310-4318.
[8]LI Y,ZHU J K,HOI S C H.Reliable patch trackers:robust visual tracking by exploiting reliable patches[C]//Proceedings of 2015 IEEE Conference on Computer Vision and Pattern Recognition.Boston:IEEE,2015:353-361.
[9]XU Y L,WANG J B,LI H,et al.Patch-based scale calculation for real-time visual tracking[J].IEEE Signal Processing Letters,2016,23(1):40-44.
[10]孙健,向伟,谭舒昆,等.改进的核相关滤波跟踪算法[J].计算机工程与应用,2018,54(9):178-182.
[11]单玲玉,闵锋,李延达.全局阈值与局部阈值相结合的视网膜血管分割方法[J].武汉工程大学学报,2015,37(3):62-67.
[12]张雷,王延杰,孙宏海,等.采用核相关滤波器的自适应尺度目标跟踪[J].光学精密工程,2016,24(2):448-459.
[13]WU Y,LIM J,YANG M H.Online object tracking:a benchmark[C]//Proceedings of IEEE Conference on Computer Vision and Pattern Recognition.Portland:IEEE,2013:2411-2418.
[14]HARE S,SAFFARI A,TORR P H S.Struct:structured output tracking with kernels[C]//Proceedings of 2011 IEEE International Conference on Computer Vision.Barcelona:IEEE,2012:263-270.
[15]KALAL Z,MATAS J,MIKOLAJCZYK K.Trackinglearning-detection[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2012,34(7):1409-1422.