Low-Cost Prioritization of Microorganism Image Data in Wireless Visual Sensor Network

Survivability is the primary problem in wireless visual sensor networks (WVSNs) when they are utilized for sewage monitoring with limited resources. In this paper, we propose a low-cost data prioritization framework for sewage microorganism microscopic images to overcome this frangibility of WVSNs. Firstly, an improved fast visual background extraction (Vibe) algorithm is proposed to rapidly detect indicative motion objects in wastewater. Then, based on the detection results and central weighted criteria, each image macro block is classified into different priority levels. Finally, we propose a flexible dynamic routing protocol depending on image data priorities to keep good balance between transmission delay and energy consumption. Experiment results demonstrate the efficiency and effectiveness in terms of microorganism detection and transmission cost. Compared with other state-of-the-art approaches, the whole network lifetime achieves significant increase under the same condition of network delay.