Shot boundary Detection based on supervised locality preserving projections and KNN-SVM classifier
Yongliang Xiao, Limin Xia · 2010
We introduce a novel method to detect video shot boundary. The method includes two stages: video frame feature supervised extraction and video frame supervised classification. Firstly, we use a supervised locality preserving projections to extract video feature, which can enlarger the difference between two shots. Then we create two cascaded KNN-SVM classifier which combines the ideas of SVM and K nearest neighbor to classify video frame to abrupt transitions, gradual transitions or normal frames. Experimental results show the effectiveness of our method.