Video Data Compression by Progressive Iterative Approximation

Author
Keywords
Abstract
In the present paper, the B-spline curve is used for reducing the entropy of video data. We consider the color or luminance variations of a spatial position in a series of frames as input data points in Euclidean space R or R3. The progressive and iterative approximation (PIA) method is a direct and intuitive way of generating curve series of high and higher fitting accuracy. The video data points are approximated using progressive and iterative approximation for least square (LSPIA) fitting. The Lossless video data compression is done through storing the B-spline curve control points (CPs) and the difference between fitted and original video data. The proposed method is applied to two classes of synthetically produced and naturally recorded video sequences and makes a reduction in the entropy of both. However, this reduction is higher for syntactically created than those naturally produced. The comparative analysis of experiments on a variety of video sequences suggests that the entropy of output video data is much less than that of input video data.
Year of Publication
9998
Journal
International Journal of Interactive Multimedia and Artificial Intelligence
Volume
In Press
Issue
In Press
Number
In Press
Number of Pages
1-7
Date Published
12/2020
ISSN Number
1989-1660
URL
https://www.ijimai.org/journal/sites/default/files/2020-12/ip2020_12_02_0.pdf
DOI
10.9781/ijimai.2020.12.002
Attachment