Title :
An effective method for abrupt scene change detection
Author_Institution :
Xi´´an Aerotechnical Coll., Xi´´an, China
Abstract :
Abrupt scene change detection is one of important research fields of shot boundary detection. In this paper, a SVM classifier is trained to determine real cut positions. The experimental results demonstrate that the method based on SVM performs well, which has a good ability of generalization under the condition of little training samples.
Keywords :
pattern classification; support vector machines; video signal processing; SVM classifier; abrupt scene change detection; shot boundary detection; Feature extraction; Histograms; Image color analysis; Kernel; Support vector machines; Training; Video sequences; SVM; abrupt scene change; detection;
Conference_Titel :
Natural Computation (ICNC), 2010 Sixth International Conference on
Conference_Location :
Yantai, Shandong
Print_ISBN :
978-1-4244-5958-2
DOI :
10.1109/ICNC.2010.5583266
