Communications and Signal Processing Seminar

Fast, Efficient, Scalable and Accessible Image Compression

Aous NamanSenior Research AssociateUniversity of New South Wales (UNSW)

This talk gives an overview of some recent research activities in our research group at UNSW, Sydney, Australia. We start with the vision of our group, giving a demonstration of what we strive to achieve. Then, we explore two recent research topics, namely: Compression of high density camera array (HDCA) and the faster block coder for JPEG2000.
The HDCA is proposed as a way of capturing the lightfield emanating from a scene. A set of HDCA images comprises a 2D grid of views for a static scene, each shot from a slightly different vantage point. Here, we explore how such a set can be compressed in an accessible and scalable format. Our contributions are in geometrically consistent depth inference at different locations from a single depth field, disparity compensation, and the inter-view transform.
The JPEG2000 image compression format has many desirable features such as high dynamic range (or bit depth), accessibility, support for lossy and lossless compression, and resolution and quality scalability. The format has been adopted for medical imaging, digital cinema, and as an intermediate format for video exchange by movie studios such as Netflix. Although JPEG2000 is more efficient than JPEG in both subjective and objective tests, it is yet to see wide consumer adoption. One main complain has been the complexity of the block coder. In this talk, I will give a short overview of JPEG2000, and show how the proposed fast block coder can be as fast as JPEG if not faster, while maintaining the JPEG2000 compression efficiency.

Aous Thabit Naman received the B.Sc. degree in Electronics and Telecommunication Engineering from Al-Nahrain University, Baghdad, Iraq, in 1994, the M.Eng.Sc. degree in Engineering from University of Malaya, Kuala Lumpur, Malaysia, in 2000, and the Ph.D. degree in Electrical Engineering from the University of New South Wales (UNSW), Sydney, Australia, in 2011.
He is currently working as a senior research associate at the School of Electrical Engineering and Telecommunications, the University of New South Wales (UNSW), Sydney, Australia. Research Interests include image and video compression, inverse problems in image and video, and transmission of video.

Sponsored by


Faculty Host

Dave Neuhoff