The Distributed Fusion Of Visual Network Of Things

With the development of multimedia technology, visual network of things has become an important direction of IOT development. However, due to a large amount of multimedia data and computational processing requirements, its energy consumption is much higher than the normal transmission of sensor nodes; but visual sensor node local storage capacity, computing power and node energy are limited. How to use limited resources to maintain the efficient operation of the network, balance the network load, reduce node energy consumption and prolong the network lifetime became key issues in visual network of things research. Distributed data fusion technology that is an information-processing approach can balance the network load. Local computing and storage capacity is underutilized in the process of sensor nodes collecting visual information from the network, sensor nodes are used to process the collected information, eliminating redundant information and reduce unnecessary duplication of data transmission; and then, another node will analyze and fuse this data collected by different sense nodes, on the basis of further reductions of redundant information in the network, monitoring information with higher reliability and accuracy has been obtained.Based on the full analysis IOT, sensor network and related knowledge of information fusion technology, we propose a distributed wavelet image fusion algorithm based on clustering structure and its simulation implementation. The algorithm uses the background subtraction method to removing redundant information in the image data, and using SURF algorithm for image registration, in this way, the rate of image processing is greatly accelerated, and the network traffic data is decreased; In addition, the image fusion algorithm and distributed processing mechanisms combine to make the integration task shared equally on multiple nodes and improve the image fusion of computing speed and network energy balance.