Shot Cut Detection Based on the Statistical Parameter Histogram with the Discrete Walsh Transform
Jianfeng Wang · 2010
In this paper, we propose an effective and fast shot cut detection algorithm directly in MPEG compressed domain. The proposed shot cut detection test the different between the current frame and next frame through the extracting the feature of each frame. When extract the features of frame, statistical parameters m1-σ from DWT coefficients without its inverse transform was computed, Except this, locating shot cuts is operated by comparison tests. In comparison with the latest research efforts in shot cut detection, our proposed algorithm achieves significant advantages including: (a): use the walsh transform to reduce the computation time for a given resolution or to increase the resolution without drastically increasing the computation time (b): there is no full decompression is needed; (c): adding correlation to judge the different between feature vectors; and (d): detection performance is competitive and fast.