臺灣博碩士論文加值系統

隨著影像擷取裝置逐漸普及化，大量的照片被產生出來。然而，並非所有的照片都有好的
品質。在此論文中，我們提出了兩個影像偵測的方法，分別是模糊偵測與曝光偵測。
模糊偵測利用支援向量機器(support vector machines)估算一張影像的模糊程度，同時
判斷此影像是全域性的模糊抑或是區域性的。對於全域性的模糊影像，我們計算其點光源
擴散函數(point spread function) 並將其分類成相機搖晃或是失焦(out of focus)所造成的模
糊。對於區域性的模糊影像，我們利用影像分割的方法找出模糊的區塊，並利用點光源擴
散函數的計算將其分類成景深(depth of field)或是移動物體我產生的模糊。
和模糊問題一樣，曝光也是一個造成影像品質下降的常見的理由。結合了感興趣區
域(region of interest)偵測技術，我們可以判斷一張影像的前景/背景是否曝光過度/不足。
本論文所使用方法的優點是所有過程皆是自動進行，因此使用者可以簡單的找到他們
真正想要的影像。

Photos are massively produced while digital image capturing devices are becoming popular,
however, not every photo has good quality. In this thesis, two image detectors are proposed:
blur detector and exposure detector.
Blur detector uses support vector machines to estimate the blur extent of an image
and distinguishes the difference between locally blurred image and globally blurred image.
For globally blurred image, we estimate the point spread function (PSF) and the image is
classified to camera shake or out of focus. For locally blurred image, we find the blurred
regions using segmentation method, and the PSF estimation on the blurred region can sort
out the image with depth of field or moving object.
Exposure problem, as well as blur, is another familiar cause of defective images. Combining
with (ROI) estimation technique, we can tell if the background/foreground of the image
is over/under exposured.
The advantage of our framework is that the processes are automatic, so the users can
easily find the images they want by these hints.
v

1 Introduction 3
1.1 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.2 Defective Photo Classification . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.3 System Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.4 Organization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2 Related Work 7
2.1 Blur Estimation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.2 PSF Estimation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.3 Exposure Calibration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
3 Blur Extent Estimation 11
3.1 Image Gradient Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
3.2 Support Vector Machines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
3.2.1 Optimal Separating Hyperplane . . . . . . . . . . . . . . . . . . . . . 15
3.2.2 Soft Margin Classifier . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
3.2.3 Non-linear SVM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
3.2.4 Probability Estimates . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
3.3 Blur Extent Estimation using SVM . . . . . . . . . . . . . . . . . . . . . . . 18
3.3.1 Training Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
3.3.2 Prediction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
3.4 Grid search . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
3.5 Experiment and result . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
3.5.1 Blur Extent Estimation . . . . . . . . . . . . . . . . . . . . . . . . . 19
3.5.2 Grid Search . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
4 Camera Shake and Out of Focus Classification 25
4.1 Camera Shake and Out of Focus . . . . . . . . . . . . . . . . . . . . . . . . . 25
4.2 Point Spread Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
4.3 PSF Estimation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
4.4 PSF Classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
4.5 Experiment and Result . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
5 Locally Blurred Image Detection 35
5.1 Depth of Field and Moving Object . . . . . . . . . . . . . . . . . . . . . . . 35
5.2 Segment-based Blur Detection . . . . . . . . . . . . . . . . . . . . . . . . . . 36
5.2.1 Graph Based Image Segmentation . . . . . . . . . . . . . . . . . . . . 36
5.2.2 Blur Estimation on Small Regions . . . . . . . . . . . . . . . . . . . . 37
5.2.3 Locally Blurred Image Classification . . . . . . . . . . . . . . . . . . 38
5.3 Experiment and Result . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
6 Exposure Detection 43
6.1 Overexposed Image and Underexposed Image . . . . . . . . . . . . . . . . . 43
6.2 Histogram Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
6.3 Region of Interest and Exposure Detection . . . . . . . . . . . . . . . . . . . 44
6.4 Experiment and Result . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
7 Defective Photo Classification 47
7.1 Photo Classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
7.2 Limitation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
8 Conclusion and Future Work 49

[1] S. Battiato, A. Bosco, A. Castorina, and G. Messina. Automatic image enhancement
by content dependent exposure correction. EURASIP Journal on Applied Signal Processing,
2004(12):1849–1860, 2004. doi:10.1155/S1110865704404107.
[2] Moshe Ben-Ezra and Shree K. Nayar. Motion-based motion deblurring. IEEE Trans.
Pattern Anal. Mach. Intell., 26(6):689–698, 2004.
[3] Bernhard E. Boser, Isabelle M. Guyon, and Vladimir N. Vapnik. A training algorithm
for optimal margin classifiers. In COLT ’92: Proceedings of the fifth annual workshop
on Computational learning theory, pages 144–152, New York, NY, USA, 1992. ACM
Press.
[4] Chih-Chung Chang and Chih-Jen Lin. LIBSVM: a library for support vector machines,
2001.
[5] Corinna Cortes and Vladimir Vapnik. Support-vector networks. Mach. Learn.,
20(3):273–297, 1995.
[6] James H. Elder and Steven W. Zucker. Local scale control for edge detection and blur
estimation. In ECCV ’96: Proceedings of the 4th European Conference on Computer
Vision-Volume II, pages 57–69, London, UK, 1996. Springer-Verlag.
[7] Pedro F. Felzenszwalb and Daniel P. Huttenlocher. Efficient graph-based image segmentation.
Int. J. Comput. Vision, 59(2):167–181, 2004.
[8] R. Fergus, B. Singh, A. Hertzmann, S. T. Roweis, and W.T. Freeman. Removing
camera shake from a single photograph. ACM Transactions on Graphics, SIGGRAPH
2006 Conference Proceedings, Boston, MA, 25:787–794, 2006.
[9] L. Itti, C. Koch, and E. Niebur. A model of saliency-based visual attention for rapid
scene analysis. IEEE Transactions on Pattern Analysis and Machine Intelligence,
20(11):1254–1259, Nov 1998.
[10] J. W. Miskin and D. J. C. MacKay. Ensemble learning for blind image separation and
deconvolution. In M. Girolami, editor, Advances in Independent Component Analysis.
Springer, 2000.
[11] J. Platt. Probabilistic outputs for support vector machines and comparison to regularize
likelihood methods. In A.J. Smola, P. Bartlett, B. Schoelkopf, and D. Schuurmans,
editors, Advances in Large Margin Classifiers, pages 61–74, 2000.
[12] Ramesh Raskar, Amit Agrawal, and Jack Tumblin. Coded exposure photography: motion
deblurring using fluttered shutter. In SIGGRAPH ’06: ACM SIGGRAPH 2006
Papers, pages 795–804, New York, NY, USA, 2006. ACM Press.
[13] Stanley J. Reeves and Russell M. Mersereau. Blur identification by the method of
generalized cross-validation. IEEE Transactions on Image Processing, 1(3):301–311,
1992.
[14] F. Rooms, W. Philips, and J. Portilla. Parametric psf estimation via sparseness maximization
in the wavelet domain. In F. Truchetet and O. Laligand, editors, Proc. of
SPIE Wavelet Applications in Industrial Processing II, pages 26–33, 2004.
[15] Hanghang Tong, Mingjing Li., HongJiang Zhang, and Changshui Zhang. Blur detection
for digital images using wavelet transform. In ICME, pages 17–20, 2004.