The Compression And Transmission Of The Image In Wireless Multimedia Sensor Networks

Wireless Multimedia Sensor Networks(WMSNs) have emerged and shifted the focus from the typical scalar wireless sensor networks to networks with multimedia devices that are capable to retrieve video, audio, images, as well as scalar data, and it makes people can sensor the richer information in digital form in the real physical world. But the ability of the WMSNs nodes is constrained, such as its power energy, the ability of computation and storage space, while the information of the image has more size and needs more computing power compared to the traditional scalar information, thus it is a critical question to the technical staff that how to assure the quality and quantity of the gathered images in the background of the very limited hardware environment.This paper tries to solve the question above by improving the process of the digital image compression and the routing protocol of the WSMNs, and the research object of this paper is the image multimedia data. To the improvement of the digital image compression, this paper brings up two thoughts. The first thought is encoding the Region of Interest(ROI) of the image in high priority, and the second one is only encoding the ROI. In the first thought, we infer the method of the wavelet ROI mask calculation when using the common wavelet to do the Discrete Wavelet Transform, in the second thought we modify the SPIHT algorithm to better accommodate the Shape-Adaptive Discrete Wavelet Transform algorithm. The experiment demonstrates that, when the bit per pixel(bpp) is not greater than 1.0, this first thought can achieve a little higher PSNR than the original algorithm that don't distinguish the ROI; the second thought can greatly decrease the size of the compressed image bit stream and save the energy cost while having a little acceptable PSNR loss compared to the original algorithm which is at the same bpp rate. To the improvement of the routing protocol, this paper reforms the communication mechanism between the cluster head and the sink node by introducing the energy parameter in the 'hop-by-hop' routing instead of the original 'single hop' routing mechanism, and the experiment demonstrates the improved algorithm can increase the total size of the image data which sink node receives, and also prolong the network lifetime of the WMSNs.