Reconstructing the Plenoptic function from wireless multimedia sensor networks

The development of low-cost image/video sensors has enabled wireless multimedia sensor networks (WMSN). Equipped with the ability to collect images/video streams from surrounding environments, many interesting applications will emerge through the use of WMSN. One of the key targets of a WMSN application is reconstructing the Plenoptic function by using collected images. However, due to the characteristics of wireless channels, this task is challenging. In this paper, a reliable transport protocol for WMSN is presented. This protocol utilizes JPEG stream semantics to distinguish the importance of different parts within the stream, and schedules the transmission based on the importance of the information. In order to investigate the performance of the proposed protocol, a novel image mosaicking algorithm is used as a sample WMSN application. The experimental results show that the proposed protocol and application can offer snapshots of the interested environment with larger field-of-view (FOV) and with little delay. The quality of the view can be improved gradually with the arrival of the data representing the higher frequency part of the images.