Research On Energy-efficient JPEG Image Coding Algorithm In Wireless Multimedia Sensor Network

Abstract:There are a large number of image data have to be processed and transmitted in resource-constrained wireless multimedia sensor network. In order to achieve maximum utilization of the resources on wireless multimedia sensor network, it is very significant to design an image coding scheme which synthetically considers performance requirements of energy consumption, compression rate and image quality.Based on the study of compression of video-surveillance image sequences collected by a wireless multimedia sensor network, then an energy-efficient modified JPEG algorithm based on change detection is proposed. In this scheme, Change detection is used to locate the region of interest, and the region of non-interest only keep DC coefficient, the data for transmission is cut down, the compression ratio is increased which is suitable for wireless multimedia sensor network. The clipping parameter having been quoted in the region of interest, the two clipping algorithms based on the square and triangle are proposed. Based on the interest region hierarchical algorithm, different interest grade uses different clipping parameter. Simulation and experimental results show that the modified method can effectively reduce the communication data in wireless sensor nodes and the energy consumption during computation in the case of ensuring the application of the desired image quality.For the sudden change scenarios, a single target interest region selection algorithm based on historical frames is proposed in this paper. The scope of single target interest region is determined by computing the area of binary images, and the region of interest in the current frame can be forecasted based on the ranges of history single target interest region. Simulation and experimental results show that the modified method will not be affected by scenarios change and the single target interest region can be predicted accurately. Integrating the improved image encoding scheme and single target interest region selection algorithm will contribute to better balance performance requirements of energy consumption, compression rate and image quality in wireless multimedia sensor networks. Therefore, this study is of great significance to promote the development and application of wireless sensor network technology.