Design and Implementation of a Dual-Camera Wireless Sensor Network for Object Retrieval

Abstract

This paper presents the design and implementation of a dual-camera sensor network that can be used as a memory assistant tool for assisted living. Our system performs energy-efficient object detection and recognition of commonly misplaced objects. The novelty in our approach is the ability to tradeoff between recognition accuracy and computational efficiency by employing a combination of low complexity but less precise color histogram-based image recognition together with more complex image recognition using SIFT descriptors. In addition, our system can seamlessly integrate feedback from the user to improve the robustness of object recognition. Experimental results reveal that our system is computation-efficient and adaptive to slow changes of environmental conditions

Property management in wireless sensor networks with overcomplete radon bases

“…Such systems can hold up to hundreds of sensor nodes, which can perform lots of sensing and controling tasks such as multitarget tracking [Dutta et al 2006;Liu et al 2011;Zhu et al 2012], intrusion detection [Arora et al 2004], ecosystem surveillance [Mo et al 2009], etc. The classical problem of reliably tracking also connects the vision community if one considers the camera network for object detection and recognition [Xie et al 2008].…”

Share risk and energy: Sampling and communication strategies for multi-camera wireless monitoring networks

“…They were originally intended for gesture recognition, indoor monitoring, and object retrieval [10] [17]. Wireless cameras used for environmental monitoring have also been proposed [7], but cameras are treated as independent nodes.…”

Environmental sensor networks for vegetation, animal and soil sciences

“…Given the high sample rates required to sample these types of signals, more data must be handled than can be achieved with the processors and radios on typical, low-powered wireless sensor platforms. As such, there is increasing focus Greenstein et al, 2006;Xie et al, 2008) on the use of more powerful co-processors which can handle large quantities of audio and video data, including the capture, signal processing and feature extraction from the raw data. The downside to the use of this additional processing power is the significant increase in energy consumption.…”