Robust and efficient transform domain video sequence analysis: An approach from the generalized color ratio model

Indexing is an important aspect of video database management, Video indexing involves the analysis of video sequences, which is a computationally intensive process. However, effective management of digital video requires highly efficient, yet robust indexing techniques, Based on the properties of the generalized color ratio model, the paper presents techniques for robust and efficient video indexing, by analyzing the video sequence directly in the compressed domain, Illumination and motion invariant color ratio features, appropriate for block-based transform domain video indexing are presented, Robustness is achieved by use of(i) illumination and motion invariant color ratios; (ii) formally derived thresholds with predictable performance; and (iii) both DC and AC coefficients, Gains in efficiency accrue from: (i) operating only on a formally selected minimal set of blocks; (ii) performing all operations directly in the compressed domain; and (iii) avoiding computationally expensive motion compensation procedures. Some experimental results are presented which show the superiority of the proposed methods over traditional approaches. (C) 1997 Academic Press.