基于視頻編輯模型的視頻淡入、淡出和疊化的檢測(cè)
Fade and dissolve detection based on video editing model
-
摘要: 該文基于視頻編輯模型,推導(dǎo)出幾個(gè)能用于檢測(cè)鏡頭淡入、淡出和疊化的重要特征:方差、邊界強(qiáng)度、有效平均梯度和雙重色度差。通過比較,我們選取了方差序列和雙重色度差特征作為檢測(cè)依據(jù),首先利用方差序列檢測(cè)淡入、淡出和可能的疊化位置,以保證取得高的查全率,然后利用雙重色度差序列確認(rèn)疊化以提高檢測(cè)準(zhǔn)確率。由于檢測(cè)對(duì)DC序列進(jìn)行,算法計(jì)算時(shí)間復(fù)雜度比較低。
-
關(guān)鍵詞:
- 鏡頭檢測(cè); 淡入; 淡出; 疊化
Abstract: Based on video editing model, several important characteristics that can be used for detecting fade and dissolve, such as the variance, edge intensity, Effective Average gradient (EAG) and Double Chromatic Difference (DCD) of video sequence are derived. After comparison, the variances are used to achieve robust fade detection. It is approved that in terms of recall rate for dissolve detection, the variance, edge intensity and EAG have similar efficiency. So, the variances are used for selection of possible positions of dissolve transitions, which guarantees high recall rate, and the DCD are then used for further confirmation to improve precision rate. Since the detection is carried out on DC sequence, the computation complexity of this method is relatively low. -
J. Boreczky, L. Rowe, Comparison of video shot boundary detection techniques, SPIE, 1996, 2670.122-128.[2]H.J. Zhang, A. Kankanhalli, S. W. Smoliar, Automatic partitioning of full-motion video.[J]. ACM Multimedia Systems.1993,1:10-[3]B.L. Yeo, B. D. Liu, Rapid scene analysis on compressed video, IEEE Trans. on CASVT, 1995,5(6), 533-544.[4]R. Zabih, J. Miller, K. Mai, A feature-based algorithm for detecting and classifying production effects, Multimedia Systems, 1999, 7(2), 119-28.[5]H. Yu.[J].G. Bozdagi, S. Harrington, Feature-based video segmentation, ICIP97, Santa Barbara,USA.1997,:-[6]Y.J. Zhang, J. J. Gerbrands, Transition region determination based thresholding, Pattern Recognition Letters, 1991, 12(1), 13-23. -
計(jì)量
- 文章訪問數(shù): 2517
- HTML全文瀏覽量: 165
- PDF下載量: 433
- 被引次數(shù): 0
下載:
