Towards a Visual Sensor Network Built from Smartphones

摘要

With the developments of the last years, smartphones became suitable to be used as smart cameras in a visual sensor network. They have the necessary cameras with an acceptable quality of lenses and CCD, the required processing power, and WiFi or Bluetooth interfaces for communication. This paper describes the architecture of the SMEyeL (Smart Mobile Eyes for Localization) system. It is designed to utilize both smartphones and PC-s as processing nodes of a visual sensor network. It aims to achieve high spatial and temporal resolution, energy consumption optimization via computation offloading and distributed processing. The SMEyeL project is open source: sourcecode and measurement data sets are available on the web. We present the detailed description of the architecture and exprimental results related to timing accuracy and localization precision.