《计算机应用研究》|Application Research of Computers

基于光流优化的堆叠Unet背景建模神经网络

Optical flow based stacked Unet background modeling network

免费全文下载 (已被下载 次)  
获取PDF全文
作者 陶冶,凌志浩
机构 华东理工大学 化工过程先进控制和优化技术教育部重点实验室,上海 200237
统计 摘要被查看 次,已被下载
文章编号 1001-3695(2020)09-062-2861-05
DOI 10.19734/j.issn.1001-3695.2019.05.0167
摘要 针对现有背景建模算法难以处理复杂前景及间歇性运动前景的问题,提出了一种基于非监督学习的背景建模算法(改进的BM-Unet算法)。该算法结合光流法和Pearson相关系数在视频帧上提取背景关注区域,以此优化网络训练集和损失函数,从而有效提高了该算法在复杂前景情况和前景停留情况下的适应性;在此基础上,为进一步提高背景生成的精确度,又提出了一种堆叠Unet网络架构BM-SUnet(background modelling stacked Unet)。在SBMnet数据集上与现有算法在可视化效果和评估参数两方面的比较结果表明,所提算法在复杂前景和间歇运动前景情况下建模准确性好且鲁棒性高的结论。
关键词 非监督学习; 卷积神经网络; 背景建模; 堆叠Unet; 复杂前景; 间歇运动前景
基金项目 中央高校基本科研业务费专项资金资助项目(222201917006)
本文URL http://www.arocmag.com/article/01-2020-09-062.html
英文标题 Optical flow based stacked Unet background modeling network
作者英文名 Tao Ye, Ling Zhihao
机构英文名 Key Laboratory of Advanced Control & Optimization for Chemical Processes,East China University of Science & Technology,Ministry of Education,Shanghai 200237,China
英文摘要 To tackle the cluttered foreground and the intermittent foreground motion challenges of background modeling algorithms, this paper proposed an unsupervised background modelling method(improved BM-Unet). The proposed method combined optical flow and Pearson correlation coefficient to extract the attention region in frames. In addition, the attention region contributed to modify the training set generation and loss function to improve the ability to adapt to the circumstance including the cluttered and the intermittent motion of foreground. In a further step, this paper proposed a stacked Unet architecture(BM-SUnet) to enhance the accuracy of background generation. Experiments on dataset SBMnet show that the algorithm is not only robust to the cluttered foreground and the intermittent motion challenge, but also outperforms the latest methods in terms of both qualitative and quantitative evaluation.
英文关键词 unsupervised learning; convolutional neural network; background modelling; stacked Unet; cluttered foreground; foreground intermittent motion
参考文献 查看稿件参考文献
 
收稿日期 2019/5/10
修回日期 2019/6/28
页码 2861-2865
中图分类号 TP391
文献标志码 A