跳到主要內容

臺灣博碩士論文加值系統

(44.200.27.215) 您好!臺灣時間:2024/04/15 05:15
字體大小: 字級放大   字級縮小   預設字形  
回查詢結果 :::

詳目顯示

我願授權國圖
: 
twitterline
研究生:林祐銘
研究生(外文):Yu-Ming Lin
論文名稱:結合色彩與動態特徵分析法於擷取與分類F1影片中
論文名稱(外文):F1 Video Segmentation and Scene Classification based on Color and Motion Information
指導教授:張敏寬
指導教授(外文):Min-Kuan Chang
學位類別:碩士
校院名稱:國立中興大學
系所名稱:電機工程學系所
學門:工程學門
學類:電資工程學類
論文種類:學術論文
論文出版年:2006
畢業學年度:94
語文別:英文
論文頁數:52
中文關鍵詞:場景變動偵測影像切割動態向量顏色直方圖鏡頭變動偵測關鍵畫面場景分類
外文關鍵詞:Shot change detectionvideo segmentationmotion vectorcolor histogramcamera motionkeyframescene classification
相關次數:
  • 被引用被引用:0
  • 點閱點閱:164
  • 評分評分:
  • 下載下載:0
  • 收藏至我的研究室書目清單書目收藏:0
隨著資訊科技的快速發展,影像數位化已普及於日常生活,例如多媒體文件、數位圖書館…等。如何將龐大的影像資料,依據影像內容的相關性,做出適當的影像資料分類,以期達到有效的管理並且方便使用者搜尋,這將是未來一項重要的發展。
本論文依據顏色直方圖、色彩空間、影像切割、影像的動態向量與動態補償,將F1賽車的影片,依照鏡頭的不同將之切割成許多影像片段,並且依據鏡頭的特性來做分析。希望達到將鏡頭分類的目的,以方便使用者搜尋需要的鏡頭內容。
一場F1比賽下來,就算同一個鏡頭所拍攝到的內容,也不可能完全一樣,而且還必須考慮到比賽中間出現的廣告,與一些特殊情況,例如:收訊的中斷。所以,如何在變化很多的鏡頭內容中,做出相同鏡頭的擷取、判斷與歸類,將是這篇論文的重點。這篇論文中,最主要用來判斷與歸類鏡頭的方法,是利用鏡頭內容裡面的色彩與動態向量的資訊,在加上F1賽車鏡頭的特性,來對鏡頭加以分析,做出最適當的鏡頭分類。因此,將可達到我們想到的分類效果。
In this paper, we proposed specific shot change detection for F1 videos. Video temporal segmentation is a very important step to index videos. Due to the characteristics of F1 video such as fast scene change, close-up view, and etc., the traditional shot change detection cannot successfully detect the shot boundary. Motion activity and color distribution are employed to analyze every shot and select a stable frame as keyframe to represent the shot. As we know, a lot of semantic information is contained in shot and motion activity. Through our proposed scheme, we can identify and classify the shots from the exact camera they were taken. Also, we can distinguish different genres of scenes from F1 videos such as stable scene, overlook scene, in-car camera scene, pit scene, yellow flag, and car trace scene. According to the experiential results, our proposed method can efficiently index F1 videos. Key applications of the proposed method include F1 videos scene classification, F1 videos event extraction and F1 videos highlight extraction.
Contents
1 Introduction 1
2 Robust feature extraction 5
2.1 Motion activity . . . . . . . . . . . . . . . . . 5
2.1.1 Full search algorithm . . . . . . . . . . . 6
2.1.2 2D-log search algorithm . . . . . . . . . . 6
2.1.3 Three steps search algorithm . . . . . .. . 7
2.2 Color space and segmentation . . . . . . . .. . . 9
2.2.1 RGB . . . . . . . . . . . . . . . . . . . . 9
2.2.2 HSI . . . . . . . . . . . . . . .. .. . . . 9
2.2.3 YUV . . . . . . . . . . . . . . . . . . . . 10
2.2.4 Color quantization . . . . . . .. . . . . . 11
2.2.5 K-mean clustering algorithm . . . . . . . . 11
2.3 Camera motion detection . . . . . . . . . . . . . 13
2.4 Keyframe selection . . . . . . . . . . . . . . . 15
2.4.1 Keyframe . . . . . . . . . . . . . . . . . 15
2.4.2 Keyframe selection rules . . . . . . . .. . 16
3 Speci‾c shot change detection for F1 video 18
3.1 Abrupt shot change . . . . . . . . . . . . . . . 18
3.2 Gradual shot change . . . . . . . . . . . . . . . 19
3.3 Basic inter-frame di®erence algorithm . . . . .. 22
3.3.1 Pixel-based algorithm . . . . . . . . . . . 22
3.3.2 Histogram-based algorithm . . . . . . . . . 22
3.4 Proposed algorithm . . . . . . . . . . . . . . . 24
4 Semantic scene classi‾cation for F1 video 31
4.1 Lap detection . . . . . . . . . . . . . . . . . . 31
4.1.1 Region growing . . . . . . . . . . .. . . . 32
4.1.2 Numeral string segmentation . . . . . . . . 33
4.2 Lap shot classi‾cation . . . . . . . . . . . . . 35
4.2.1 Keyframe information . . . . . . . . . . . 35
4.2.2 Edge information extraction . . . . . . . . 37
4.2.3 Pan and tilt motion activity . . . . .. . . 39
4.2.4 Difference of color histogram information .. 40
4.3 Genres of camera scene . . . . . . . . . . . . . 41
5 Experiment results 44
5.1 Performance parameters . . . . . . . . . . . . . 44
5.2 The statistics of the experiment results . . . . 45
5.2.1 Result of shot change detection . . . . . . 45
5.2.2 Result of lap shot classi‾cation . . . .. . 47
5.2.3 Result of camera scene classi‾cation . . . 48
6 Conclusion and future work 49
6.1 Conclusion . . . . . . . . . . . . . . . . . . . 49
6.2 Future work . . . . . . . . . . . . . . . . . . . 50
Bibliography
[1] H. M. Hang, Y.M. Chou, and S. C. Cheng, "Motion Estimation for Video Coding Standards", Journal of VLSI Signal Processing,17, pp. 113-136, 1997.
[2] A.Shari‾nejad and H.Mehrpour, "A fast full search block matching algorithm using three window search based on the statistical analysis of the motion vectors", IEEE Int. Conf. Communications,2002. Vol. 1, 28 April-2 May 2002, pp. 104 - 108.
[3] Rafael C. Gonzalez, Richard E.Woods, "Digital image processing".
[4] W. Wolf, "Key Frame Selection by Motion Analysis", in Proc.IEEE Int. Conf., Acoustics, Speech, and Signal, 1996, Vol. 2, 7-10 May. 1996. pp. 1228 - 1231.
[5] C. L. Huang and B. Y. Liao, "A robust scene-change detection method for video segmentation", IEEE Trans. Circuits and Systems for Video Technology, Vol. 11, Issue 12, Dec. 2001, pp. 1281- 1288.
[6] Q. Liu, L. Yang, "Twi-di®erence algorithm for video''s abrupt shot change detection", in Proc. Int. Conf., Neural Networks and Signal Processing, Vol. 2, 14-17 Dec. 2003, pp. 1177 - 1180.
[7] S.-R. Gong, Y.J. Fan, "Video abrupt shot change detection based on relation of the partial interframe differences", in Proc.Int.Conf., Machine Learning and Cybernetics, 2005, Vol. 9, 18-21 Aug. 2005,pp.5255 - 5260.
[8] C. W. Su, Liao H.-Y.M., H. R. Tyan, K. C. Fan, L.H. Chen, "A motion-tolerant dissolve detection algorithm", IEEE Trans., Multimedia, Vol. 7, Issue 6, Dec. 2005, pp. 1106 - 1113.
[9] W. Hua, M. Han, Y. Gong, "Baseball scene classi‾cation using multimedia features", in Proc. IEEE Int. Conf., Multimedia and Expo, 2002, Vol. 1, 26-29 Aug. 2002, pp. 821 - 824.
[10] J.Garcia-Consuegra, G. Cisneros, E. Navarro, "A sequential ECHO algorithm based on the integration of clustering and region growing techniques", in Proc. IEEE Int., Geoscience and Remote Sensing Symposium, Vol. 2, 24-28 July 2000, pp. 648 - 650.
QRCODE
 
 
 
 
 
                                                                                                                                                                                                                                                                                                                                                                                                               
第一頁 上一頁 下一頁 最後一頁 top