Energy efficient video surveillance approach for multiple camera in wireless sensor networks

Abstract

The advances in the embedded system, computer vision, and wireless transmissions result in cost-effective wireless video surveillance. The digital camera is capable of detecting the moving target according to the control information. Video surveillance in the constrained environment of a wireless sensor network requires to be processed into three steps. Initially, the video is pre-processed after that compression is done and data is transmitted over the WMSN; at the last step, the analysis of the video is done at the receiver site. However, for wireless video-based sensor nodes transmitting a video stream is difficult using the traditional protocols. Camera-based Wireless Multimedia Sensor Network (WMSN) performs relatively in a different way from the traditional setups in the sense that it involves a directional sensing model, a complex in node processing and a need to transfer a large amount of data. In this paper, we suggest a new methodology which integrates the barrier coverage through multiple camera node models for object detection and transmission, instead of sending the complete frame, the objective is to achieve the trade-off between energy consumption and reliability of detection. Since blanket coverage generally results in high transmission time and the communication overhead is significant. To ensure the longevity of applications, reduction in energy consumption is of prime importance. However, the detection of the intruder and its success the rate ratio is highly dependent on the width of the barrier, the speed, and position of the entering event [1, 2, 6].