Image Sequence Coding Research Articles

A differential method for simultaneous estimation of rotation, change of scale and translation

The problem of 2-D motion estimation with application to image sequence coding is considered in this paper. A four-parameter motion model, capable of describing rotation, change of scale and translation simultaneously, is proposed. The four motion parameters are estimated directly from an image sequence without the need for establishing the correspondence between pixels or features. The method is a differential one applied in conjunction with a multi-resolution iteration scheme to enhance its measuring range and efficiency. Experimental results demonstrate that the motion parameters can be estimated accurately by the method with minimal computational effort.

Tree-structured scene adaptive coder

A new type of scene adaptive coder has been developed. It involves a quadtree mean decomposition of the motion-compensated frame-to-frame difference signal followed by a scalar quantization of the local means. As a fundamental property, the new coding algorithm treats the displacement estimation problem and the quadtree construction problem as a unit. The displacement vector and the related quadtree are jointly optimized in order to minimize the direct frame-to-frame update information rate (in bits), which turns up as a new and more adequate cost function in displacement estimation. This guarantees the highest possible data compression ratio at a given quality threshold. Excellent results have been obtained for coding of color image sequences at a rate of 64 kb/s. The quadtree concept entails a much lower computational complexity compared to the conventional motion-compensated transform coder while achieving a subjective image quality that is as good or better than that of the traditional transform-based counterpart. >

Radiographic image sequence coding using two-stage adaptive vector quantization

A two-stage adaptive vector quantization scheme for radiographic image sequence coding is introduced. Each frame in the sequence is first decomposed into a set of vectors, corresponding to nonoverlapping spatially contiguous block of pixels. A codebook is generated using a training set of vectors drawn from the sequence. Each vector is then encoded by the label representing the closest codeword of the codebook, and the label values in a frame label map memory at both ends of the communication channel. The changes occurring in the radiographic image sequences can be categorized into two types: those due to body motion and those due to the injected contrast dye material. In the second scheme proposed, encoding is performed in two stages. In the first stage, the labels of corresponding vectors from consecutive frames are compared and the frame label map memory is replenished (updated). This stage is sufficient to tack the changes caused by patient motions but not due to the injected contrast dye material. The resulting residual error vectors after the first stage coding are calculated for the latter changes and are further encoded by a second codebook, which is updated on a frame-to-frame basis.

Frame-adaptive vector quantization for image sequence coding

An adaptive technique for image sequence coding that is based on vector quantization is described. Each frame in the sequence is first decomposed into a set of vectors. A codebook is generated using the vectors of the first frame as the training sequence, and a label map is created by quantizing the vectors. The vectors of the second frame are then used to generate a new codebook, starting with the first codebook as seeds. The updated codebook is then transmitted. At the same time, the label map is replenished by coding the position and the new values of the labels that have changed from one frame to the other. The process is repeated for subsequent frames. Experimental results for a test sequences demonstrate that the technique can track the changes and maintain a nearly constant distortion over the entire sequence.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

A Low Bit-Rate Image Sequence Coder Combining a Progressive DPCM on Interleaved Rasters with a Hybrid DCT Technique

An image sequence coding scheme for transmission at 384 kbits/s and 64 kbits/s is presented. In comparison to known schemes, picture quality in still picture areas is improved without impairing the rendition of temporally changed areas. Bit-rate reduction is obtained by a conventional hybrid DCT coding technique in changed areas, and a predictive coding technique using a raster interleaving approach to update still picture areas. Preferentially treating the image signal in still areas with a not-block-oriented coder yields in these regions a reconstruction with full spatial resolution and free of any block effects. Experimental results are presented confirming that the quality in still picture areas is improved compared to that of known low bit-rate image sequence coding schemes, while the quality in changed areas is maintained.