Marion Benetiere

A fully scalable 3D subband video codec

Multimedia transmissions over heterogeneous networks require a high degree of flexibility from video compression systems. They are expected to be fully scalable, that is to say to be able to partly decode a video bitstream and to get a reconstruction quality proportional to the received amount of information. To achieve this functionality, we propose a video codec based on 2D+t subband decomposition. Groups of frames are first temporally filtered using motion compensation and then spatially decomposed with wavelets. The spatiotemporal coefficients are further scanned and compressed using a new SPIHT-like strategy, namely fully scalable zerotree coding, together with arithmetic encoding, which provides a combination of temporal, spatial and SNR scalability. In addition, scalable motion vector coding ensures a fully progressive bitstream.

Scalable compression of 3D medical datasets using a (2D+T) wavelet video coding scheme

Nowadays the all-digital solution in hospitals is becoming widespread. The formidable increase of medical data to be processed, transmitted and stored, requires some efficient compression systems and innovative tools to improve data access, while ensuring a sufficient visualization quality for diagnosis. Medical image sequences can benefit from advanced video coding techniques when adapted to their specific constraints. Scalability, or the capability to partly decode a video bitstream and to get a reconstruction quality proportional to the decoded amount of information, is a key functionality. We have developed a video codec based on a 3D motion-compensated subband decomposition, which provides a combination of temporal, spatial and SNR scalabilities together with a very competitive compression ratio. We show that, when applied to medical sequences, it outperforms JPEG-2000 on coding efficiency aspects and offers new functionalities.

Embedded color coding for scalable 3D wavelet video compression

With the recent expansion of multimedia applications, video coding systems are expected to become highly scalable, that is to allow partial decoding of the compressed bit-stream. Encoding techniques based on subband/wavelet decompositions offer a natural hierarchical representation for still pictures and their high efficiency in progressively encoding images yields a scalable representation. The multiscale representation can be extended to video data, by a 3D (or 2D+t) wavelet analysis, which includes the temporal dimension within the decomposition. Progressive encoding of video data represented by a 3D-subband decomposition was recently proposed as an extension of image coding techniques exploiting hierarchical dependencies between wavelet coefficients. In most of the previous image and video coding techniques, the compression is performed independently for luminance and chrominance coordinates. In this paper we propose a new coding technique for the chrominance coefficients, which not only delivers a bit-stream with a higher degree of embedding, but also takes advantage of the dependencies between luminance and chrominance components to provide an effective compression.