Mobile Multispectral Video Streaming

The Multi-Spectral (MS) image provides 10x or 100x extra information compared with the RGB content. It plays an important role in various applications such as material identification, object tracking, environment monitoring, etc. However, complex and bulky optical subsystem of the acquisition setup and ultra high volume of the spectral data have severely hindered the widespread adoption of MS image/video based applications. This paper presents a mobile MS video streaming system, including the compact hand-held MS video acquisition, transmission, and rendering in the back-end servers. All of these subsystems are realized using low-cost off-the shelf optical and electrical components, resulting in a small form factor mobile device. It enlarges the potential of applying the MS videos in different areas. Towards this purpose, we have demonstrated two real-time simulations, including environment monitor using material discrimination and criminal investigation using human-face tracking, by leveraging the spectral characteristics of the MS video that are not contained in conventional RGB content.