Title :
A method of shot detection based on color and edge features
Author :
Qu, Zhiyi ; Liu, Ying ; Ren, Liping ; Chen, Yong ; Zheng, Ruidong
Author_Institution :
Sch. of Inf. Sci. & Eng., Lanzhou Univ., Lanzhou, China
Abstract :
Aimed at that there is a missing detection in the video shot segmentation because of that the global color feature can´t reflect the changes in the image´s substance, a new method of shot detection is proposed using global color feature combined with the characteristic of local edge. In order to determine the border of the shot segmentation, the paper eliminated the flash interference with the gray value of the object´s edge, and constructed a gradient detector using edge features on the basis of calculating the inter-frame similarity according to eigenvalues of the global color. The experimental results show that the method is simple and effective, and has a high accuracy.
Keywords :
edge detection; eigenvalues and eigenfunctions; image colour analysis; image segmentation; object detection; video signal processing; color eigenvalue; color feature; edge feature; flash interference elimination; shot detection method; video shot segmentation; Brightness; Computer vision; Flowcharts; Gunshot detection systems; Histograms; Image edge detection; Image segmentation; Information science; Leak detection; Lenses; abrupt shot detection; color feature; edge feature; gradual shot detection; inter-frame similarity; shot segmentation;
Conference_Titel :
Web Society, 2009. SWS '09. 1st IEEE Symposium on
Conference_Location :
Lanzhou
Print_ISBN :
978-1-4244-4157-0
Electronic_ISBN :
978-1-4244-4158-7
DOI :
10.1109/SWS.2009.5271731