Presents a new formulation of moving object segmentation and motion estimation to quantify the potential gain in using object-oriented motion compensation in image sequence coding. Motivated by Rissanen's (1983) minimum description length (MDL) principle that allows unified treatment of coding model complexity and parameter values, an objective function is constructed in terms of an ideal coding length function for each motion-compensated frame. Based on this new objective function, a procedure to segment and estimate moving objects in a scene is proposed. Each object in the scene is represented by its boundaries, motion parameters, and motion-compensated prediction error (MCPE). A number of experimental comparisons between block-oriented and object oriented coding schemes suggest that significant potential coding gain using the new MDL-based criterion is possible.<>
Application of the minimum description length principle to object-oriented video image compression
Proceedings of 1st International Conference on Image Processing ; 2 ; 400-404 vol.2
01.01.1994
535777 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Application of the Minimum Description Length Principle to Object-Oriented Video Image Compression
British Library Conference Proceedings | 1994
|Principle of representational minimum description length in image analysis and pattern recognition
British Library Online Contents | 2012
|Minimum description length principle in the field of image analysis and pattern recognition
British Library Online Contents | 2011
|Minimum Noiseless Description Length (MNDL) Thresholding
IEEE | 2007
|Minimum description length method for facet matching
SPIE | 1998
|