There is an increasing need to extract key information automatically from v
ideo for the purposes of indexing, fast retrieval and scene analysis. To su
pport this vision, reliable scene change detection algorithms must be devel
oped. This paper describes a unified approach for scene change detection in
uncompressed and MPEG-2 compressed video sequences using statistical prope
rties of each image. An efficient algorithm is proposed to estimate statist
ical features in compressed video without full frame decompression and used
these features with the uncompressed domain algorithms to identify scene c
hanges in compressed video. Proposed scheme aims at detecting abrupt transi
tions and gradual transitions in both uncompressed and MPEG-2 compressed vi
deo using a single framework. Results on video of various content types are
reported and validated. Furthermore, results show that for uncompressed vi
deo the accuracy of the detected transition region is above 98% and above 9
5% for MPEG-2 compressed video.