TY - JOUR
T1 - Real-Time and Robust Compressive Background Subtraction for Embedded Camera Networks
AU - Shen, Yiran
AU - Hu, Wen
AU - Yang, Mingrui
AU - Liu, Junbin
AU - Wei, Bo
AU - Lucey, Simon
AU - Chou, Chun Tung
PY - 2016/2/1
Y1 - 2016/2/1
N2 - Real-time target tracking is an important service provided by embedded camera networks. The first step in target tracking is to extract the moving targets from the video frames, which can be realised by using background subtraction. For a background subtraction method to be useful in embedded camera networks, it must be both accurate and computationally efficient because of the resource constraints on embedded platforms. This makes many traditional background subtraction algorithms unsuitable for embedded platforms because they use complex statistical models to handle subtle illumination changes. These models make them accurate but the computational requirement of these complex models is often too high for embedded platforms. In this paper, we propose a new background subtraction method which is both accurate and computationally efficient. We propose a baseline version which uses luminance only and then extend it to use colour information. The key idea is to use random projection matrics to reduce the dimensionality of the data while retaining most of the information. By using multiple datasets, we show that the accuracy of our proposed background subtraction method is comparable to that of the traditional background subtraction methods. Moreover, to show the computational efficiency of our methods is not platform specific, we implement it on various platforms. The real implementation shows that our proposed method is consistently better and is up to six times faster, and consume significantly less resources than the conventional approaches. Finally, we demonstrated the feasibility of the proposed method by the implementation and evaluation of an end-to-end real-time embedded camera network target tracking application.
AB - Real-time target tracking is an important service provided by embedded camera networks. The first step in target tracking is to extract the moving targets from the video frames, which can be realised by using background subtraction. For a background subtraction method to be useful in embedded camera networks, it must be both accurate and computationally efficient because of the resource constraints on embedded platforms. This makes many traditional background subtraction algorithms unsuitable for embedded platforms because they use complex statistical models to handle subtle illumination changes. These models make them accurate but the computational requirement of these complex models is often too high for embedded platforms. In this paper, we propose a new background subtraction method which is both accurate and computationally efficient. We propose a baseline version which uses luminance only and then extend it to use colour information. The key idea is to use random projection matrics to reduce the dimensionality of the data while retaining most of the information. By using multiple datasets, we show that the accuracy of our proposed background subtraction method is comparable to that of the traditional background subtraction methods. Moreover, to show the computational efficiency of our methods is not platform specific, we implement it on various platforms. The real implementation shows that our proposed method is consistently better and is up to six times faster, and consume significantly less resources than the conventional approaches. Finally, we demonstrated the feasibility of the proposed method by the implementation and evaluation of an end-to-end real-time embedded camera network target tracking application.
KW - compressed sensing
KW - data reduction
KW - embedded systems
KW - image colour analysis
KW - matrix algebra
KW - statistical analysis
KW - target tracking
KW - video cameras
KW - video coding
KW - robust compressive background subtraction method
KW - embedded camera networks
KW - real-time target tracking
KW - video frames
KW - resource constraints
KW - complex statistical models
KW - luminance
KW - colour information
KW - random projection matrices
KW - data dimensionality reduction
KW - Computational modeling
KW - Sensors
KW - Accuracy
KW - Cameras
KW - Compressed sensing
KW - Gaussian distribution
KW - Vectors
KW - Object Tracking
KW - Real-time Performance
KW - Embedded Camera Networks
KW - Background Subtraction
KW - Compressive Sensing
KW - Gaussian Mixture Models
KW - Object tracking
KW - real-time performance
KW - background subtraction
KW - compressive sensing
KW - Gaussian mixture models
U2 - 10.1109/TMC.2015.2418775
DO - 10.1109/TMC.2015.2418775
M3 - Article
VL - 15
SP - 406
EP - 418
JO - IEEE Transactions on Mobile Computing
JF - IEEE Transactions on Mobile Computing
SN - 1536-1233
IS - 2
ER -