Intelligent video transmission control for mobile cameras

Abstract

Data communication between mobile security robots over the network is an important component in a surveillance system. Because of limited bandwidth, transmission delay, and package loss, real-time video transmission may pose some challenges. In this paper, three steps of image data reduction before transmission are proposed and discussed in detail. The first one is temporal sampling, which selects the clearest frame in sequential ones. The next step is setting region of interest (ROI) for the chosen frame. Finally, a spatial sampling on manipulating the level of detail is performed in the ROI. Experimental results show that the procedures reduce data redundancy and maintain the important information simultaneously.