浏览全部资源
扫码关注微信
收稿：2017-07-19，
修回：2017-9-30，
纸质出版：2018-02-16
移动端阅览
目的
2
动态场景图像中所存在的静态目标、背景纹理等静态噪声，以及背景运动、相机抖动等动态噪声，极易导致运动目标检测误检或漏检。针对这一问题，本文提出了一种基于运动显著性概率图的目标检测方法。
方法
2
该方法首先在时间尺度上构建包含短期运动信息和长期运动信息的构建时间序列组；然后利用TFT（temporal Fourier transform）方法计算显著性值。基于此，得到条件运动显著性概率图。接着在全概率公式指导下得到运动显著性概率图，确定前景候选像素，突出运动目标的显著性，而对背景的显著性进行抑制；最后以此为基础，对像素的空间信息进行建模，进而检测运动目标。
结果
2
对提出的方法在3种典型的动态场景中与9种运动目标检测方法进行了性能评价。3种典型的动态场景包括静态噪声场景、动态噪声场景及动静态噪声场景。实验结果表明，在静态噪声场景中，
$$ {\mathit{F}_{{\rm{score}}}}$$
提高到92.91%，准确率提高到96.47%，假正率低至0.02%。在动态噪声场景中，
$$ {\mathit{F}_{{\rm{score}}}}$$
提高至95.52%，准确率提高到95.15%，假正率低至0.002%。而在这两种场景中，召回率指标没有取得最好的性能的原因是，本文所提方法在较好的包络目标区域的同时，在部分情况下易将部分目标区域误判为背景区域的，尤其当目标区域较小时，这种误判的比率更为明显。但是，误判的比率一直维持在较低的水平，且召回率的指标也保持在较高的值，完全能够满足于实际应用的需要，不能抵消整体性能的显著提高。另外，在动静态噪声场景中，4种指标均取得了最优的性能。因此，本文方法能有效地消除静态目标干扰，抑制背景运动和相机抖动等动态噪声，准确地检测出视频序列中的运动目标。
结论
2
本文方法可以更好地抑制静态背景噪声和由背景变化（水波荡漾、相机抖动等）引起的动态噪声，在复杂的噪声背景下准确地检测出运动目标，提高了运动目标检测的鲁棒性和普适性。
Objective
2
Moving target detection is an important research content of image analysis technology. The purpose of it is to remove the background interference through a series of operations to extract and detect moving targets. It can be applied to video surveillance
image retrieval and motion analysis etc. The classical methods of moving object detection are mainly realized by extracting motion information of inter frames
detecting optical flow changes
or background modeling. But in dynamic scenarios
such as the scenarios which are affected by static noise (background is similar to moving target) and dynamic noise (noise which is caused by the branches
ripple and camera jitter)
accuracy and robustness of moving target detection is greatly reduced. For this reason
many improved methods have been put forward. Some achievements have been made in static background noise scene or dynamic background noise scene. However
it is a pity that few methods achieve perfect results in both two situations. Aiming to solve the problems of false detection caused by static noise(background is similar to moving target) and dynamic noise (background changes or camera shaking etc.)
this paper proposes a moving object detection method by using the motion saliency probability map and compares it with 9 moving object detection methods in 3 typical dynamic scenes.
Method
2
The motion saliency probability map enhances the saliency of the moving targets at the current frame by using the motion information in a long time and weakens the saliency of background and moving targets in historical frames. Therefore
the maximum of the probability value corresponds to the motion saliency of the current frame. The dynamic noise in the image (such as background branches
ripple and camera jitter) often causes a large number of mistakes in detection results. These pixels are not only gathered together
but also partially salient in movement
and can not be eliminated by morphological or noise filtering methods. However
the correlation between adjacent pixels can provide a high level of detection accuracy in dynamic background scenes
and can effectively suppress dynamic noise. In this method
the paper firstly constructs a set of time series including short-term motion information and long-term motion information on a time scale. Then the saliency value is calculated by the TFT method. Based on this
the conditional motion saliency probability map is obtained. Next
the motion saliency probability map is given under the guidance of the full probability formula
which can eliminate static noise
small dynamic noise and the influence of historical frame on moving targets. By segmenting the motion saliency probability map
the saliency of the moving target is highlighted
and the saliency of the background is suppressed. Finally
the spatial information of the pixel is modeled based on the motion saliency probability map to optimize the result
which can eliminate significant dynamic noise. The process of modeling is divided into two steps
including the calculation of shift probability map and component shift probability map. Foreground target can be extracted by binaryzating the component shift probability map. Generally speaking
the main innovation of this paper lies in:a saliency probability model is proposed. The paper combines the saliency value with the probability of occurrence of the moving targets. Besides
it fuses saliency detection model and probability model to construct a saliency probability model creatively.
Result
2
In this paper
the proposed method is compared with 9 moving object detection methods in 3 typical dynamic scenes
including static background noise scene
dynamic background noise scene caused by camera shake and dynamic background noise scene caused by water ripple. The experimental results show that in the static noise scene
$$ {\mathit{F}_{{\rm{score}}}}$$
is increased to 92.91%
precision rate is increased to 96.47% and false positive rate is as low as 0.02%. In the dynamic noise scene caused by camera shake
$$ {\mathit{F}_{{\rm{score}}}}$$
is in creased to 95.52%
precision is increased to 95.15%
and false positive rate is as low as 0.002%. In these two scenarios
the index recall can not achieve the best performance
because in some cases
the method proposed in this paper is easy to misjudge some of the target areas as the background area while it can better envelope target region
especially when the target region is small
and in this case
ratio of this misjudgment is more obvious. However
false positive rate has been maintained at a low level
and recall rate is kept at a higher value
which can fully meet the needs of practical applications. So while recall rate doesn't perform best
this can not offset the significant improvement in overall performance. In addition
in the dynamic background noise scene caused by water ripple
the four indexes all achieve the best performance. In general
the proposed method can eliminate static object interference and suppress dynamic background noise
accurately detecting the moving object in video sequence.
Conclusion
2
To solve the problems of false detection caused by static noise and dynamic noise
a novel target detection method is proposed in this paper. In general
the method can be divided into three parts. Firstly
the paper construct a set of time series. Then the saliency value is calculated and we can get the conditional motion saliency probability map and the saliency probability map respectively. Finally
the spatial information of the pixel is modeled
including calculating the shift probability map and component shift probability map
and the moving object is detected. In the paper
nine methods (GMM
KDE
ViBe
PQFT
DiscSal
ROSL
RMAMR
ManhNMF and Deep-Semi-NMF) are compared with the proposed method in 3 typical dynamic scenes. Experiments show that the proposed method can better suppress the static background noise and the dynamic noise caused by background change (such as water ripple
camera shake
etc.)
and accurately detect moving objects.
Jain R, Nagel H H. On the analysis of accumulative difference pictures from image sequences of real world scenes[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1979, 1(2):206-214.[DOI:10.1109/TPAMI.1979.4766907]
Berthold K.P.Horn, Brian G.Schunck. Determining optical flow[J].Artificial Intelligence, 1981, 17:185-203.
Babacan S D, Pappas T N. Spatiotemporal algorithm for background subtraction[C]//Proceedings of 2007 IEEE International Conference on Acoustics, Speech and Signal Processing. Honolulu, HI, USA: IEEE, 2007, 1: Ⅰ-1065-Ⅰ-1068. [ DOI:10.1109/ICASSP.2007.366095 http://dx.doi.org/10.1109/ICASSP.2007.366095 ]
Liu G, Ning S K, You Y G, et al. An improved moving objects detection algorithm[C]//Proceedings of 2013 International Conference on Wavelet Analysis and Pattern Rec ognition. Tianjin, China: IEEE, 2013: 96-102. [ DOI:10.1109/ICWAPR.2013.6599299 http://dx.doi.org/10.1109/ICWAPR.2013.6599299 ]
Hao J Y, Li C, Kim Z, et al. Spatio-temporal traffic scene modeling for object motion detection[J]. IEEE Transactions on Intelligent Transportation Systems, 2013, 14(1):295-302.[DOI:10.1109/TITS.2012.2212432]
Guo C L, Ma Q, Zhang L M. Spatio-temporal saliency detection using phase spectrum of quaternion Fourier transform[C]//Proceedings of 2008 IEEE Conference on Computer Vision and Pattern Recognition. Anchorage, AK, USA: IEEE, 2008: 1-8. [ DOI:10.1109/CVPR.2008.4587715 http://dx.doi.org/10.1109/CVPR.2008.4587715 ]
Cui X Y, Liu Q S, Zhang S T, et al. Temporal Spectral Residual for fast salient motion detection[J]. Neurocomputing, 2012, 86:24-32.[DOI:10.1016/j.neucom.2011.12.033]
Wang Z H, Liao K, Xiong J L, et al. Moving object detection based on temporal information[J]. IEEE Signal Processing Letters, 2014, 21(11):1403-1407.[DOI:10.1109/LSP.2014.2338056]
Shah V R, Maru S V, Jhaveri R H. An obstacle detection scheme for vehicles in an intelligent transportation system[J]. International Journal of Computer Network and Information Security, 2016, 8(10):23-28.[DOI:10.5815/ijcnis.2016.10.03]
Chen Z, Wang X, Sun Z, et al. Motion saliency detection using a temporal fourier transform[J]. Optics&Laser Technology, 2016, 80:1-15.[DOI:10.1016/j.optlastec.2015.12.013]
Nobuyuki Otsu. A threshold selection method from gray-level histograms[J]. IEEE Transactions on Systems, Man, and Cybernetics, 1979, 9(1):62-66.
Elgammal A M, Harwood D, Davis L S. Non-parametric model for background subtraction[C]//Proceedings of the 6th European Conference on Computer Vision. Dublin, Ireland: Springer-Verlag, 2000: 751-767. [ DOI:10.1007/3-540-45053-X_48 http://dx.doi.org/10.1007/3-540-45053-X_48 ]
Stauffer C, Grimson W E L. Adaptive background mixture models for real-time tracking[C]//Proceedings of 1999 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. Fort Collins, CO, USA: IEEE, 1999: 252. [ DOI:10.1109/CVPR.1999.784637 http://dx.doi.org/10.1109/CVPR.1999.784637 ]
Barnich O, Van Droogenbroeck M. ViBe:a universal background subtraction algorithm for video sequences[J]. IEEE Transactions on Image Processing, 2011, 20(6):1709-1724.[DOI:10.1109/TIP.2010.2101613]
Mahadevan V, Vasconcelos N. Spatiotemporal saliency in dynamic scenes[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2010, 32(1):171-177.[DOI:10.1109/TPAMI.2009.112]
Shu X B, Porikli F, Ahuja N. Robust orthonormal subspace learning: efficient recovery of corrupted low-rank matrices[C]//Proceedings of 2014 IEEE Conference on Computer Vision and Pattern Recognition. Columbus, OH, USA: IEEE, 2014: 3874-3881. [ DOI:10.1109/CVPR.2014.495 http://dx.doi.org/10.1109/CVPR.2014.495 ]
Ye X C, Yang J Y, Sun X, et al. Foreground-background separation from video clips via motion-assisted matrix restoration[J]. IEEE Transactions on Circuits and Systems for Video Technology, 2015, 25(11):1721-1734.[DOI:10.1109/TCSVT.2015.2392491]
Guan N Y, Tao D C, Luo Z G, et al. MahNMF: manhattan non-negative matrix factorization[J]. arXiv: 1207. 3438, 2012.
Trigeorgis G, Bousmalis K, Zafeiriou S, et al. A deep Semi-NMF model for learning hidden representations[C]//Proceedings of the 31st International Conference on Machine Learning. Beijing, China: JMLR, 2014: 1692-1700.
相关作者
相关机构
京公网安备11010802024621
