跳到主要內容

臺灣博碩士論文加值系統

(34.204.198.73) 您好!臺灣時間:2024/07/21 15:41
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

: 
twitterline
研究生:李誠祥
研究生(外文):Cheng-Shiang Li
論文名稱:基於平均移動法與散度計算的前背景分離
論文名稱(外文):Figure-Ground Separation based on Mean Shift and Divergence Computation
指導教授:鮑興國鮑興國引用關係
指導教授(外文):Hsing-Kuo Pao
學位類別:碩士
校院名稱:國立臺灣科技大學
系所名稱:資訊工程系
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:2010
畢業學年度:98
語文別:英文
論文頁數:44
中文關鍵詞:平均移動法散度計算前背景分離
外文關鍵詞:mean shiftdivergence computationfigure-ground separation
相關次數:
  • 被引用被引用:0
  • 點閱點閱:326
  • 評分評分:
  • 下載下載:14
  • 收藏至我的研究室書目清單書目收藏:0
在人類視覺中,前背景分離是很個很重要的處理過程。該過程組合零散的視覺資訊組織並辨別影像中每個區域之間的層次關係。在本論文中,我們提出一個基於平均移動法與散度計算的新演算方式。該方式使用平均移動法估計各區域內的向量分佈並計算影像內各區域邊緣的散度。最後依照該散度值來判斷影像當中,各區域之間的前背景關係。此外,我們使用合成圖型,一般自然影像以及柏克萊前背景資料集等三種不同纇型的影像進行測試,並展示該演算法不但能應用在簡單的合成圖型,也能使用在自然影像中。
Figure-ground separation is an important process that separates the contours
between regions as figure and ground in human vision. This process organizes
the visual information and help people recognize the layer about objects in
the image.
In this paper, we proposed a new computing method with mean shift algo-
rithm and divergence computation. This approach bases on the convexity
and surroundedness cues and measures the vector distribution with mean
shift. Finally we determine the figure-ground labels with divergence com-
puted from vector distribution nearby the contours. In addition, we test our
method with synthesis image, nature image and Berkeley Figure/Ground
Dataset to show that this approach is not only effective on simple geometric
shape but also on the nature images.
1 Introduction 1
2 Related Works 3
2.1 Figure-Ground Separation . . . . . . . . . . . . . . . . . . . . 3
2.1.1 Size . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.1.2 Convexity . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.1.3 Lower Region . . . . . . . . . . . . . . . . . . . . . . . 8
2.1.4 Shapemes . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.2 Mean Shift . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.2.1 Kernel Density Estimation . . . . . . . . . . . . . . . . 11
2.2.2 Data Clustering . . . . . . . . . . . . . . . . . . . . . . 13
2.2.3 Image Filtering and Segmentation . . . . . . . . . . . . 14
2.3 Divergence Computation . . . . . . . . . . . . . . . . . . . . . 17
3 Mean Shift Figure-Ground Separation 19
3.1 Framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
3.1.1 Notations . . . . . . . . . . . . . . . . . . . . . . . . . 21
3.2 Image Segmentation and Contour Definition . . . . . . . . . . 23
3.3 Gradient Vector Extraction . . . . . . . . . . . . . . . . . . . 25
3.4 Mean Shift and Divergence Computation . . . . . . . . . . . . 26
3.4.1 Contour and Segmentation Sets . . . . . . . . . . . . . 26
3.4.2 Mean Shift . . . . . . . . . . . . . . . . . . . . . . . . 26
3.4.3 Divergence Computation . . . . . . . . . . . . . . . . . 27
3.5 Figure-Ground Labeling . . . . . . . . . . . . . . . . . . . . . 28
4 Experiment Results 30
4.1 Synthesis image . . . . . . . . . . . . . . . . . . . . . . . . . . 30
4.2 Nature Image . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
4.3 Berkeley Figure/Ground Dataset . . . . . . . . . . . . . . . . 34
5 Conclusion and Future Work 39
5.1 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
5.2 Future Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
[1] Alexander C. Berg and Jitendra Malik. Geometric blur for template
matching. Computer Vision and Pattern Recognition, IEEE Computer
Society Conference on, 1:607, 2001.
[2] David R. Martin Charless C. Fowlkes and Jitendra Malik. Local figure-
ground cues are valid for natural images. Journal of Vision, 7(8):1–9,
2007.
[3] Yizong Cheng. Mean shift, mode seeking, and clustering. IEEE Trans-
actions on Pattern Analysis and Machine Intelligence, 17(8):790–799,
August 1995.
[4] D. Comaniciu and P. Meer. Mean shift: a robust approach toward
feature space analysis. IEEE Transactions on Pattern Analysis and
Machine Intelligence, 24(5):603–619, May 2002.
[5] H.-K. Pao D. Geiger and N. Rubin. Salient and multiple illusory surfaces.
In Proc. IEEE Computer Society Conference on Computer Vision and
Pattern Recognition, pages 118–124, June 23–25, 1998.
[6] Andrew McCallum John Lafferty and Fernando Pereira. Conditional
random fields: Probabilistic models for segmenting and labeling se-
quence data. In International Conference on Machine Learning, pages
282–289. Morgan Kaufmann, 2001.
[7] G. Kanizsa and W. Gerbino. Convexity and symmetry in figure-ground
organization. Vision and artifact, pages 25–32, 1976.
[8] S. Lloyd. Least squares quantization in pcm. IEEE Transactions on
Information Theory, 28(2):129–137, 1982.
[9] Stephen E. Palmer. Vision Science: Photons to Phenomenology. The
MIT Press, 1 edition, May 1999.
[10] H.-K. Pao. A Continuous Model for Salient Shape Selection and Repre-
sentation. PhD thesis, New York University, 2001.
[11] H.-K. Pao, D. Geiger, and N. Rubin. Measuring convexity for fig-
ure/ground separation. In Proc. Seventh IEEE International Conference
on Computer Vision, volume 2, pages 948–955, September 20–27, 1999.
[12] M.A. Peterson. Object recognition processes can and do operate before
figure-ground organization. Current Directions in Psychological Science,
3:105–111, 1994.
[13] Xiaofeng Ren, Charless C. Fowlkes, and Jitendra Malik. Figure/ground
assignment in natural images. In Proc. 9th Europ. Conf. Comput. Vi-
sion, volume 2, pages 614–627. Springer, 2006.
[14] E. Rubin. Visuell wahrgenommene figuren: studien in psychologischer
analyse. Gyldendalske boghandel, 1921.
[15] C. Samson, L. Blanc-Feraud, G. Aubert, and J. Zerubia. A variational
model for image classification and restoration. IEEE Transactions on
Pattern Analysis and Machine Intelligence, 22(5):460–472, May 2000.
[16] Jianbo Shi and J. Malik. Normalized cuts and image segmenta-
tion. IEEE Transactions on Pattern Analysis and Machine Intelligence,
22(8):888–905, August 2000.
[17] Dale Varberg and Edwin J. Purcell. Calculus. Prentice Hall Interna-
tional, Inc., 7th edtion edition, 1997.
[18] S.P. Vecera, E.K. Vogel, and G.F. Woodman. Lower region: A new
cue for figure-ground assignment. Journal of Experimental Psychology:
General, 131(2):194–205, 2002.
QRCODE
 
 
 
 
 
                                                                                                                                                                                                                                                                                                                                                                                                               
第一頁 上一頁 下一頁 最後一頁 top