Abstract
This paper describes a scalable video streaming system based on JPEG2000 with various modes of streaming. A core function of the proposed system is a low-complexity transcoding technique that adapts the quality and resolution of the scene based on available bandwidth. One key feature of this technique that is important for surveillance applications is that interesting regions of the scene are assigned higher quality than background regions. To cope with varying network conditions, we also present a rate control algorithm that adaptively transcodes stored JPEG2000 frames. The proposed algorithm is designed to improve overall quality over a uniform rate control method by increasing bandwidth utilization, while satisfying buffer constraints and maintaining consistent quality over time. Simulation results confirm the effectiveness of the proposed system and rate control algorithm in terms of both objective measures and subjective evaluation. Complexity is also evaluated.
Highlights
The JPEG2000 standard [1] is becoming an increasing popular coding format for a variety of applications that require efficient storage and scalable transmission of images and video
We describe a video surveillance system based on JPEG2000 that allows for transmission of the scene over limited bandwidth networks
A third method that we introduce is mosaic streaming (MS), which is similar to background refresh (BR) mode, but it superimposes successive ROI images on a background in a mosaic style
Summary
The JPEG2000 standard [1] is becoming an increasing popular coding format for a variety of applications that require efficient storage and scalable transmission of images and video. Several ROI coding techniques for JPEG2000 images have been proposed in the recent years. These methods can be classified into two categories: static and dynamic ROI coding. In static ROI coding, the ROI is selected and defined during the encoding process. The main problem with such static schemes is that they are not suitable for interactive environments in which the ROI is defined after encoding. One dynamic ROI coding method is described in [6] This method allows for the definition of ROI in an interactive environment and handles the ROI by dynamically inserting layers. The dynamic layer insertion in this scheme reencodes the packet header, which requires rate-distortion recalculation and is an undesirable for real-time applications
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
