Transition Detection Using Hilbert Transform and Texture Features

Abstract

In this paper, we propose a new method for detecting shot boundaries in video sequences by performing Hilbert transform and extracting feature vectors from Gray Level Co-occurrence Matrix (GLCM). The proposed method is capable of detecting both abrupt and gradual transitions such as dissolves, fades and wipes in the video sequences. The derived features on processing through Kernel k-means clustering procedure results in efficient detection of abrupt and gradual transitions, as tested on TRECVid video test set containing various types of shot transition with illumination effects, object and camera movement in the shots. The results show that the proposed method yields better result compared to that of the existing transition detection methods.