05-02-2013, 11:33 AM
Moving Objects Detection Based on Embedded Video Surveillance
1Moving Objects.pdf (Size: 313.72 KB / Downloads: 55)
Abstract
Requested by intelligentization of video
surveillance system, the paper proposes a moving
objects detection program based on Embedded video
surveillance. With the help of the transaction of
Servfox streaming media on ARM-Linux operation
platform, the video collection and transmission of
USB camera is achieved. The paper adopts the
improved algorithm of the updated background
model of inter-frame difference to obtain the moving
objects in the client. The method reduces influence of
moving objects to the background model. The
experiment shows that the background image can be
obtained accurately, and the moving objects
recognition is achieved effectively.
INTRODUCTION
With the development of science and
technology ,and the increasing requirement of the
security, video surveillance system has been used
widely in banks, transportations, marketplaces,
intelligent communities and so on. The trend of
video surveillance system is digitized, integrated,
intelligent and networking [1]. Embedded video
surveillance system is Mini size, strong function
and Great real timing.
Moving objects detection is an important part of
intelligent video surveillance system ,its aim is to
Split moving objects from background image in
video sequences. Until now, the main algorithms of
moving objects detection are background difference
method and inter-frame difference method[2][3].
VIDEO ACQUISITION AND
TRANSMISSION
In ARM platform, the USB camera transmits
collected video data through internal bus to
streaming server, which transmits video streaming
to the client by Ethernet based on the TCP/IP
protocol. According to the module of image
transmission in server, it designed network
applications of client using VC6.0 in PC.First, it
created socket, and communicated with video
server, then received video data.
The system selects Servfox to be the embedded
streaming server[5]. It modifies Makefie of Servfox
software package,configures compiler options and
kernel path. It produces executive Servfox program
in embedded platform by executing the command
make .Servfox accomplishes acquisition,
compression and transmission of video signals
through V4L API function, MJPEG/JPEG image
compression and TCP network transfer protocol.
CONCLUSION
The paper transplants Servfox streaming server
in the embedded Linux operating system, and the
remote video surveillance is achieved.VC6.0 is
adopted to compile Video image acquisition and
processing program in the client, which takes full
advantage of OpenCV image processing module.
Moving objects detection takes full advantage of
the interrelation between frame images in video,
and eliminates the influence of complex
background in the nonmoving region. The detection
result indicates that the method has the advantages
of inter-frame difference, it is stability in dynamic
environment. In the mean time, because of
combined with background difference, detected
moving objects have relatively complete
information. which can be applied to moving
objects detection based on embedded video
surveillance.