基于CNN和RNN的像素级视频目标跟踪算法

崔家梁; 冯朝晖; 李芹; 赵红颖

doi:DOI:10.13442/j.gnss.1008-9268.2019.03.001

基于CNN和RNN的像素级视频目标跟踪算法

doi: DOI:10.13442/j.gnss.1008-9268.2019.03.001

北京大学地球与空间科学学院, 北京 100871

计量
- 文章访问数: 517
- HTML全文浏览量: 64
- PDF下载量: 131
- 被引次数: 0
出版历程
- 刊出日期: 2019-06-15

CNN and RNN-based pixel-wise video object tracking algorithm

School of Earth and Space Sciences, Peking University,Beijing100871,China

摘要

摘要: 影像目标跟踪定位技术是当前计算机视觉领域的研究热点,目标跟踪算法也是现阶段将视频结果用于定位的薄弱环节之一．本文分析了像素级目标跟踪存在的问题,根据深度学习在图像领域的最新研究成果与视频跟踪需求,结合最新的图像分割、卷积神经网络(CNN)、循环神经网络(RNN)和加密解码结构等方法提出了一种像素级视频目标跟踪算法．使用公开数据集实现算法并设计了定量评价指标．实验结果表明该算法具有较强的像素级视频目标跟踪定位能力
- 视频处理 /
- 目标跟踪定位 /
- 像素级 /
- 深度学习
Abstract: Video object tracking is a research hotpot of computer vision, but the object tracking algorithm is one of weaknesses in localization with video. In this paper, we introduced the pixel-wise video object tracking problem. We proposed a pixel-wise video object tracking algorithm based on convolution and recurrent neural network of deep learning technology and the latest research results in image segmentation to solve the problem. By designed and conducted an experiment and evaluation on a public dataset, the algorithm shows abilities on obtaining the boundary of objects in videos.
- video processing /
- object tracking /
- pixel-wise /
- deep learning

HTML全文

参考文献(10)

[1]	SCHALKOFF R J, MCVEY E S. A model and tracking algorithm for a class of video targets［J］. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1982, 4(1):2.
[2]	KIM C, HWANG J N. Fast and automatic video object segmentation and tracking for content-based applications［J］. IEEE Transactions on Circuits and Systems for Video Technology, 2002, 12(2):122-129.
[3]	RONNEBERGER O, FISCHER P, BROX T. UNet: Convolutional networks for biomedical image segmentation［C］／／International Conference on Medical Image Computing and ComputerAssisted Intervention, 2015.
[4]	BADRINARAYANAN V, KENDALL A, CIPOLLA R. SegNet: A deep convolutional encoder-decoder architecture for image segmentation［J］.IEEE transactions on pattern analysis and machine intelligence, 2017, 39(12): 2481-2495.
[5]	HUA C, WU H Y, CHEN Q, et al. A pixel-wise object tracking algorithm with target and background sample［C］／／International Conference on Pattern Recognition, 2006.
[6]	SONF Y L, LI C, YAO W. Pixel-wise object tracking［J］. CoPR abs/1711.07377 ,2017: 265-283.
[7]	HOCHREITER S, SCHMIDHUBER J. Long shortterm memory［J］. Neural Computation, 1997, 9(8):1735-1780.
[8]	HAN J, MORAGA C. The influence of the sigmoid function parameters on the speed of backpropagation learning［C］／／International Workshop on Artificial Neural Networks. Springer, Berlin, Heidelberg, 1995: 195-201.
[9]	VOJIR T, MATAS J. Pixel-wise object segmentations for the VOT2016 dataset［R］. Research Report CTU-CMP-2017-01, Center for Machine Perception, Czech Technical University, Prague, Czech Republic, 2017.
[10]	ABADI M, BARHAM P, CHEN J, et al. TensorFlow: A system for large-scale machine learning［C］//12th {USENIX} Symposium on Operating Systems Design and Implementation ({OSDI} 16), 2016: 265-283.