Abstract:
The wireless multimedia sensor networks is becoming popular and is used in many important
applications like surveillance systems, health monitoring system, maintaining law and order
systems monitoring and control of industrial processes. These applications require good
quality of information to be transmitted and received by multimedia sensors. For wireless
sensor networks multimedia applications are complex and require extra memory and
processing power of the sensors. Since sensors are tiny devices, they have very limited
memory and processing power. The lifetime of sensor depends on its remaining power,
which is less likely to be recharged. Therefore multimedia applications, which need to be run
on sensor networks, should consider the limitation of such networks. On wireless sensor
networks, multi-view video processing is an important application. As sensors impose
constraints like limited memory, storage and battery power, therefore, these limitations
should be considered while designing multi view applications for such networks.
The aim of this research is to study the existing correlation models in multi view videos and
propose an approach which reduces battery consumption by minimizing the compression
time of video frames based on the specific correlation model. A novel OPI model is
discussed that is designed based on the decision of correlation value of sensor nodes.
Algorithms for the extraction of overlapping part from the frame at the decoder and fusion of
overlapping part with the modified frame at the decoder are discussed. Experimental results
have been derived using OPI model and the results depict that by using this approach
compression time of video frames is reduced as compared to originally encoded frames.
Results show that there is remarkable decrease in compression time of processed frame using
OPI model as compared to the compression time of original frame captured by the camera
i.e. compression time of frames decreases from 10 to 20 percent and quality of image is also
maintained as shown by PSNR values. PSNR values of reconstructed images range from 40
DBs to 45 DBs, which shows that quality of image is not degraded.
