Performance of statistical inference for block classification in reducing blocking artefact in video coded images
conference contribution
posted on 2004-01-01, 00:00authored byMorshed Chowdhury, S Ray, Thao Pham
This paper presents a performance study of four statistical test algorithms used to identify smooth image blocks in order to filter the reconstructed image of a video coded image. The four algorithms considered are the Coefficient of Variation (CV), Exponential Entropy of Pal and Pal (E), Shannon's (Logarithmic) Entropy (H), and Quadratic Entropy (Q). These statistical algorithms are employed to distinguish between smooth and textured blocks in a reconstructed image. The linear filtering is carried out on the smooth blocks of the image to reduce the blocking artefact. The rationale behind applying the filter on the smooth blocks only is that the blocking artefact is visually more prominent in the smooth region of an image rather than in the textured region.
History
Title of proceedings
CSITeA'04 : Third International Conference on Computer Sciences, Software Engineering, Information Technology, E-Business and Applications
Event
Conference on Computer Sciences, Software Engineering, Information Technology, E-Business and Applications (3rd: 2004 : Cairo, Egypt)
Pagination
1 - 4
Publisher
International Society for Computers and Their Applications (ISCA)
Location
Cairo, Egypt
Place of publication
Cairo, Egypt
Start date
2004-12-27
End date
2004-12-29
ISBN-13
9780974205915
ISBN-10
0974205915
Language
eng
Publication classification
E1 Full written paper - refereed; E Conference publication