基于时空信息的运动目标识别算法
|
摘要
现实场景中人们感兴趣的目标或事件往往与场景中正在移动的物体有关,使得运动目标检测往往成为许多应用中进行信息提取的关键步骤和富有挑战的问题。首先利用单应性变换作为背景运动模型,利用上一帧得到的前景掩码图限制背景运动的估算区域,提取更纯净的背景运动;其次使用具有可变学习速率的双模单高斯模型,获得不被污染的背景模型,融合运动目标的时空信息,生成前景概率图并预测下一帧的检测区域;后依据前景概率图构造自适应的前景自适应判决阈值,进行运动目标检测,并利用核密度估计进一步优化检测结果。实验结果表明,算法可以适用光照变化、复杂运动等场景,具有较低的计算复杂度,查准率、召回率和F值等指标平均提升0.52%、33.37%、20.14%。
The interested objects or events is often related to moving objects in the scene, so moving object detection become a key step and challenging problem. This paper studies the detection of moving objects with a moving camera, and the homography transform is used as the background motion model. Only the background motion is involved in the estimation of the model without the interference of the foreground motion during the estimation of the model parameters. A dual-mode single-Gaussian model is adopted to prevent the background model from being contaminated by foreground pixels and the background model is transferred in continuous frames using the motion-fusion based compensation method. A foreground probability map is also erected according to the temporal and spatial attributes of the foreground object and is intended for an implementation of the adaptive decision threshold constructed to the selected pixels. Compared with other algorithms on a unified video sequence, the experimental results show that the proposed algorithm has a good performance on detection of which is proved to be a real-time approach, and the precision, the recall and the F-measure value increased by 0.52%, 33.37% and 20.14% respectively.
The interested objects or events is often related to moving objects in the scene, so moving object detection become a key step and challenging problem. This paper studies the detection of moving objects with a moving camera, and the homography transform is used as the background motion model. Only the background motion is involved in the estimation of the model without the interference of the foreground motion during the estimation of the model parameters. A dual-mode single-Gaussian model is adopted to prevent the background model from being contaminated by foreground pixels and the background model is transferred in continuous frames using the motion-fusion based compensation method. A foreground probability map is also erected according to the temporal and spatial attributes of the foreground object and is intended for an implementation of the adaptive decision threshold constructed to the selected pixels. Compared with other algorithms on a unified video sequence, the experimental results show that the proposed algorithm has a good performance on detection of which is proved to be a real-time approach, and the precision, the recall and the F-measure value increased by 0.52%, 33.37% and 20.14% respectively.
引文
[1] GUILLOT C, TARON M, SAYD P. Background subtraction adapted to PTZ cameras by keypoint density estimation[C]. British Machine Vision Conference, 2010: 1-10.
[2] ROBINAULT L, BRES S, MIGUET S. Real time foreground object detection using PTZ camera.[C]. Proceedings of the Fourth International Conference on Computer Vision Theory and Applications, 2009: 609-614.
[3] KIM S W, YUN K, YI K M. Detection of moving objects with a moving camera using non-panoramic background model[J]. Machine Vision and Applications, 2013, 24(5): 1-14.
[4] YI K M, YUN K, KIM S W. Detection of moving objects with non-stationary cameras in 5.8 ms: Bringing motion detection to your mobile device[C]. Computer Vision and Pattern Recognition Workshops, 2013: 27-34.
[5] KIM J, WANG X F, WANG H, et al. Fast moving object detection with non-stationary background[J]. Multimedia Tools and Applications, 2013, 67(1): 311-335.
[6] LIM J, HAN B. Generalized background subtraction using super pixels with label integrated motion estimation[C]. European Conference on Computer Vision, Springer, 2014: 173-187.
[7] ZAMALIEVA D, YILMAZ A, DAVIS J W. A multi-transformational model for background subtraction with moving cameras[C]. European Conference on Computer Vision, Springer, 2014: 803-817.
[8] BERGER M, SEVERSKY L M. Subspace tracking under dynamic dimensionality for online background subtraction[C]. IEEE Conference on Computer Vision and Pattern Recognition2014: 1274-1281.
[9] EVANGELIDIS G, DIEGO F, HORAUD R. From video matching to video grounding[C].IEEE International Conference on Computer Vision Workshops, 2013: 608-615.
[10] SHEIKH Y, JAVED O, KANADE T. Background subtraction for freely moving cameras[C].IEEE International Conference on Computer Vision, 2009: 1219-1225.
[11] CUI X, HUANG J, ZHANG S. Background subtraction using low rank and group sparsity constraints[C]. European Conference on Computer Vision, 2012: 612-625.
[12] ELQURSH A. Online moving camera backgroundsubtraction[J]. Lecture Notes in Computer Science,2012, 7577(1): 228-241.
[13] KIM H K, SURYANTO, KIM D H. Fast object detection method for visual surveillance[C] International Technical Conference on Circuits Systems, Computers and Communications, 2008: 709-712.
[14] KWAK S, LIM T, NAM W. Generalized background subtraction based on hybrid inference by belief propagation and bayesian filtering[C]. IEEE International Conference on Computer Vision (ICCV), 2011: 2174-2181.
[15] YUN K, CHOI J Y. Robust and fast moving object detection in a non-stationary camera via foreground probability based sampling[C]. IEEE International Conference on Image Processing, 2015: 4897-4901.
[2] ROBINAULT L, BRES S, MIGUET S. Real time foreground object detection using PTZ camera.[C]. Proceedings of the Fourth International Conference on Computer Vision Theory and Applications, 2009: 609-614.
[3] KIM S W, YUN K, YI K M. Detection of moving objects with a moving camera using non-panoramic background model[J]. Machine Vision and Applications, 2013, 24(5): 1-14.
[4] YI K M, YUN K, KIM S W. Detection of moving objects with non-stationary cameras in 5.8 ms: Bringing motion detection to your mobile device[C]. Computer Vision and Pattern Recognition Workshops, 2013: 27-34.
[5] KIM J, WANG X F, WANG H, et al. Fast moving object detection with non-stationary background[J]. Multimedia Tools and Applications, 2013, 67(1): 311-335.
[6] LIM J, HAN B. Generalized background subtraction using super pixels with label integrated motion estimation[C]. European Conference on Computer Vision, Springer, 2014: 173-187.
[7] ZAMALIEVA D, YILMAZ A, DAVIS J W. A multi-transformational model for background subtraction with moving cameras[C]. European Conference on Computer Vision, Springer, 2014: 803-817.
[8] BERGER M, SEVERSKY L M. Subspace tracking under dynamic dimensionality for online background subtraction[C]. IEEE Conference on Computer Vision and Pattern Recognition2014: 1274-1281.
[9] EVANGELIDIS G, DIEGO F, HORAUD R. From video matching to video grounding[C].IEEE International Conference on Computer Vision Workshops, 2013: 608-615.
[10] SHEIKH Y, JAVED O, KANADE T. Background subtraction for freely moving cameras[C].IEEE International Conference on Computer Vision, 2009: 1219-1225.
[11] CUI X, HUANG J, ZHANG S. Background subtraction using low rank and group sparsity constraints[C]. European Conference on Computer Vision, 2012: 612-625.
[12] ELQURSH A. Online moving camera backgroundsubtraction[J]. Lecture Notes in Computer Science,2012, 7577(1): 228-241.
[13] KIM H K, SURYANTO, KIM D H. Fast object detection method for visual surveillance[C] International Technical Conference on Circuits Systems, Computers and Communications, 2008: 709-712.
[14] KWAK S, LIM T, NAM W. Generalized background subtraction based on hybrid inference by belief propagation and bayesian filtering[C]. IEEE International Conference on Computer Vision (ICCV), 2011: 2174-2181.
[15] YUN K, CHOI J Y. Robust and fast moving object detection in a non-stationary camera via foreground probability based sampling[C]. IEEE International Conference on Image Processing, 2015: 4897-4901.