The Research On Video Fingerprint Algorithm Based On Salient Region

In this paper, a new method of video fingerprint extraction is proposed. First of all, to form a saliency map based on the modified FT (Frequency-Tuned) saliency test of each video frame, and the modified WLD (Weber local Descriptor) features are extracted from the spatial-temporal three-dimensional union in salient region. Then the video fingerprint is obtained by using it among the saliency map. The experimental results by using publicly available datasets indicate that the algorithm proposed in this paper can improve the efficiency of video fingerprint extraction, which has a positive effect on the balance keeping between the accuracy and practical performance of the video fingerprint algorithm.