[11] Louis Giannetti, “Understanding Movies,” Prentice Hall, 1990.
[12] T. Hain and et al., “Segment Generation And Clustering In The Htk Broadcast News Transcription System,” in Proc. of 1998 Broadcast News Transcription and Understanding Workshop, pp. 133-137, 1998.
[13] ISO/IEC 11172-3:1993, “Information Technology — Coding of Moving Pictures and Associated Audio for Digital Storage Media at up to about 1.5 Mbit/s — Part 3: Audio.”
[14] Zhu Liu and Qian Huang, “Classification of Audio Events in Broadcast News,” in Proc. of IEEE Second Workshop on Multimedia Signal Processing, pp. 364-369, Dec. 1998.
[15] Zhu Liu, Jincheng Huang and Yao Wang, ” Classification TV Programs Based on Audio Information Using Hidden Markov Model,” in Proc. of IEEE Second Workshop on Multimedia Signal Processing, vol. , pp. 27-32, Dec. 1998.
[16] Zhu Liu and et al., “Audio Feature Extraction and Analysis for Scene Classification,” in Proc.
of IEEE First Workshop on Multimedia Signal Processing, pp. 343-348, June 1997.
[17] Beth Logan, “Mel Frequency Cepstral Coefficients for Music Modeling,” in Proc. of International Symposium on Music Information Retrieval, 2000.
[18] Lie Lu and et al., “Content Analysis for Audio Classification and Segmentation,” IEEE Transactions on Audio Classification and Segmentation, vol. 10, pp. 504-516, October 2002.
[19] J.P. Marques de Sá, ” Pattern Recognition Concepts, Methods and Applications,” Springer, 2001.
[20] MPEG Requirements Group, “Information technology - Multimedia Content Description Interface - Part 2:Description Definition Language,” ISO/IEC JTC1/SC29/WG11 N4002, Singapore, Mar. 2001.
[21] MPEG Requirements Group, “Information technology - Multimedia Content Description Interface - Part 4:Audio,” ISO/IEC CD 15938-4, Oct. 2000.
[22] MPEG Requirements Group, “Information technology - Multimedia Content Description Interface - Part 5:Multimedia Description Schemes,” ISO/IEC JTC1/SC29/WG11 N3966, Singapore, Mar. 2001.
[23] MPEG Requirements Group, “Overview of MPEG-7 Standard(version 8.0),” ISO/IEC JTC1/SC29/WG11 N4980, Singapore, July. 2002.
[24] A. Nagasaka and Y. Tanaka, “Automatic video indexing and fullvideo search for objects appearances,” Visual Database Systems II, E. Knuth and L. M. Wegner, Eds. New York:
Elsevier Science, pp.113–127, 1992.
[25] Y. Nakajima and et al., “A fast audio classification from MPEG coded data,” in Proc. of 1999 IEEE International Conference on Acoustics, Speech, and Signal Processing, vol. 6, pp.
3005-3008, March 1999.
[26] M.R. Naphade and et al., “Probabilistic multimedia objects (multijects): a novel approach to video indexing and retrieval in multimedia systems,” in Proc. of 1998 International Conference on Image Processing, vol. 3, pp. 536-540, Oct. 1998.
[27] N. Patel and I. Sethi, “Audio Characterization for Video Indexing,” in Proc. of SPIE Conf.
Storage Retrieval Still Image Video Databases, vol. 2670, pp. 373-384, 1996.
[28] Silvia Pfeiffer, Stephan Fischer and Wolfgang Effelsberg, “Automatic audio content analysis,”
in Proc. of the fourth ACM international conference on Multimedia, pp. 21-30, 1997.
[29] V. I. Pudovkin, “Film Technique, and Film Acting,” Grove Press, Jun 1970.
[30] J. Saunders, “Read-Time Discrimination of Broadcast Speech/Music,”in Proc. of IEEE ICASSP, pp. 993-996, 1996.
[31] E. Scheirer and M.Slaney, “Construction and Evaluation of a Robust Multifeature Speech/Music Discrimination,” in Proc. IEEE ICASSP, vol. 2, pp. 1331-1334, 1997.
[32] G.. Tzanetakis and P. Cook, “Multifeature Audio Segmentation for Browsing and Annotation,”
IEEE Workshop on Applications of Signal Processing to Audio and Acoustics, pp. 17-20, Oct.
[33] E. Wold and et al., “Content-based Classification, Search, and Retrieval of Audio,” IEEE Multimedia, vol. 3, pp. 27-36, Fall 1996.
[34] B. L. Yeo and B. Liu, “Rapid scene analysis on compressed videos,” IEEE Trans. Circuits Syst.
Video Technol., Vol. 5, No. 6, pp. 533-544, Dec. 1995.
[35] H. J. Zhang , A. Kankanhalli, and S. W. Smoliar, “Automatic partitioning of full-motion video,” Multimedia Systems, Vol.1, No.1, pp.10-28, June 1993.
[36] Tong Zhang and C. C. Jay Kuo, “Audio Content Analysis for Online Audiovisual Data Segmentation and Classification,” IEEE Transactions on Speech and Audio Processing, vol. 9, pp. 441 - 457, MAY 2001.
[37] 范世鎮、劉志俊, “利用特寫鏡頭偵測與主角辨識技術來自動建立電影摘要,” 第二屆數位
典藏技術研討會, 2003.
[38] 陳信修、劉志俊, “一種利用特寫鏡頭對數位電影資料進行自動化摘要合成之技術,” 第一
屆數位典藏技術研討會, 2002.
[39] 黃群菘、劉志俊, “MP3 數位音樂資料的自動化分類,” 第一屆數位典藏技術研討會, 2002
[40] 葉億真、劉志俊, “音效資料的內涵式分類及其在電影資料庫的應用,” 第二屆數位典藏技
術研討會, 2003.
[41] 劉志俊、傅佳源、王志浩、喻仲平, “一種利用物件形狀來進行 MPEG-4 鏡頭變化偵測之
技術,” 第一屆數位典藏技術研討會, 2002.
[42] 梅長齡, ”電影原理與製作,” 三民書局股份有限公司, 1978.
[43] 鄭煒平、劉志俊, “MPEG 電影音效自動分段系統,” 2004 數位生活與網際網路科技研討會,
2004.
[44] 鄭煒平、劉志俊, “網際網路電影資料庫之音效自動分段索引系統,” 網際網路應用與發展
學術研討會,Vol. 6, 2005.
附錄 A MPEG7 特徵値公式
表 9 頻率特徵 Frequency Features 公式表[40]
特徵值名稱 公式
AveFeq
[ ]
0),0 575(
28125 .
38 ≤ ≤
>
= ×
∑ ∑
ii M Count frameFeq line
f
frameFeq AveFeq
∑
f−=
1 0
AveBandwidth frameBW =
(
maxMDCT −minMDCT)
×38.28125 fframeBW th
AveBandwid
∑
f−=
1 0
表 10 頻譜特徵 Spectrum Features公式表[40]
特徵值名稱 公式
AveSpectralCentroid
[ ]
[ ]
,0≤ ≤575=
∑ ∑
ii M
i Cframe iM
f
=
∑
1 -f
0 Cframe lCentroid
AveSpectra
AveSpectralRollOff
∑
min[ ]
≥∑ [ ]
0
575
0
85 . 0
R
i M i
M
f lRollOff R
AveRpectra
∑
f−=
1
0 min
AveSpectralFlux F
[ ][ ]
k i = M[
k +1][ ]
i −M[ ][ ]
k i ,0≤ i≤ 575,0 ≤ k ≤ f −2[ ] [ ][ ]
1
2 0
=
∑
− − fi k F i
lFlux AveSpectra
f
AveFlux
[ ]
,0 575 576575
0 ≤ ≤
=
∑
Fluxi iAveFlux
AveNZFlux
∑
575Flux[ ]
i表 11 能量特徵 Energy Features公式表[40]
特徵值名稱 公式
AveRMS
( [ ] )
575 0
576 ,
575 0
2
≤
≤
=
∑
i i RMS M
( )
f AveRMS RMS
∑
f−=
1 0
AveFeatureVariance
[ ] [ ]
575 1
575 , iance 1
FeatureVar − − ≤ ≤
=
∑
M i M i i[ ] [ ]
575 Variance 1
AveFeature =
∑
M i −M i−AveNegPower
( [ ] )
575 0
576 , r 0
AveNegPowe
1
0 ≤ ≤
×
=
∑ ∑
− < if
i M Count f
AveIntensity
( [ ] )
575 0
, ty
AveIntensi
1 0
2
≤
≤
=
∑ ∑
− if i
f M
AvePower
( [ ] )
575 0
f , AvePower
1
0 ≤ ≤
=
∑ ∑
f− M i iAveLowEnergy
[ ]
,0 5753 . 0 ]
[ ≤ × ≤ ≤
=
∑ ∑
ii M
AvePower i
LE M
f AveES LE
∑
f−=
1 0
AveMidEnergy
[ ]
] 5.5 ,0 575[ 5
.
4 × ≤ ≤ × ≤ ≤
=
∑ ∑
ii M
AvePower i
M AvePower
ME
f AveES ME
∑
f−=
1 0
AveHigEnergy
[ ]
,0 5757 . 0 ]
[ ≥ × ≤ ≤
=
∑ ∑
ii M
AvePower i
HE M
f AveES HE
∑
f−=
1 0
AveEnergySequences
( [ ] [ ] )
575 1
, 575
7 . 0
1 ≥ × ≤ ≤
−
= Count M i −M i AvePower i ES
f AveES ES
∑
f−=
1 0
ALPE
( )
f
AveRMS RMS
Count
∑
f− < ×=
1
0 0.5
ALPE
AveSR
[ ]
,0 575max 05 . 0 ] [
575 0
≤
× ≤
= ≤
∑ ∑ ∑
ii M
MDCT i
SR M
f AveSR SR
∑
f−=
1 0
表 12 頻率能量特徵 Frequency-EnergyFeatures 公式表[40]
特徵值名稱 公式
AveMaxPowerFrequency
f f MDCT
∑
×=
1 -f
0 max 38.28125 rFrequency
AveMaxPowe AveLowFeqPower
[ ]
∑ ∑ [ ]
= 575
0 5 0
i M
i low M
f
=
∑
1 -f
0 low
ower AveLowFeqP AveMidLowFeqPower
[ ]
∑ ∑ [ ]
= 575
0 13 6
i M
i midlow M
f
=
∑
1 -f
0 midlow eqPower
AveMidLowF AveMidFeqPower
[ ]
∑ ∑ [ ]
= 575
0 26 14
i M
i mid M
f
=
∑
1 -f
0 mid
ower AveMidFeqP AveMidHigFeqPower
[ ]
∑ ∑ [ ]
= 575
53 27
i M
i midhig M