Design of Bandwidth Efficient Compressed Sensing Based Prediction Measurement Encoder for Video Transmission in Wireless Sensor Networks

摘要

Real time transmission of image and video requires a high degree of processing and computing power. A new emerging technique called compressed sensing is used to address this issue and lower the sampling rate of signals. This paper presents an effective compressed sensing based prediction measurement (CSPM) encoder compatible for wireless multimedia sensor networks. CSPM encoding focuses on a significant reduction in data storage and saving in transmission energy. The compression performance of CSPM method is evaluated using metrics such as compression ratio and bit rate. The video is reconstructed by the orthogonal matching pursuit algorithm. The recovered video quality is analyzed by peak signal to noise ratio and structural similarity index. The transmission of encoded data is tested in real time environment using Telos B motes. The experimental results show that the CSPM encoding technique is able to deliver the video at good quality and achieve a high compression ratio of 90.7 % compared to conventional encoders.