Structure Information and Temporal Ordinal Measure Fused Video Copy Detection

Abstract

This paper propose a video copy detection method based on structure information and temporal ordinal measure. Firstly each frame is divided into a grid and calculate the difference of the mean of block brightness, then we treat these values as the structure information of each frame, whereas, along the time series these values are sorted in an ordinal ranking sequence, we put them as the temporal features; In the measurement of distance, a weighted matrix is used as the temporal ordinal measure the experiment results suggest that our proposed scheme could achieve better robustness for various content-preserving manipulations and also resist rotating, zooming, shifting, and other types of conventional geometric attacks very well.