影像區塊分割 (region segmentation) 是影像應用的一個重要之前處理步驟，目的在於分割完整區塊以簡化在分析影像時候需要處理之資料量及運算量，並且提升正確性。為了將影像內相似性質且相鄰的像素分割成同一區塊，常使用的方法有區塊成長法及影像灰階群聚法。而當影像區塊間部份邊界不明顯及區塊內灰階不一致，上述方法容易在這些區塊偵測錯誤的輪廓。因此我們使用邊緣偵測法先取得區塊間灰階差異較大的像素位置作為初始邊緣線段。再以初始邊緣的斷點為起點作邊緣延伸將區塊分割，最後完成封閉輪廓。本研究提出以粒子濾波器來延伸斷邊使影像形成封閉輸廓之方法。
我們所提出的粒子濾波器延伸斷邊成封閉輪廓的方法，共分成四個部份：影像平滑化、邊緣偵測、粒子濾波邊緣延伸、相鄰區塊合併。首先，以雙向濾波器 (bilateral filter) 去除影像雜訊並保留重要的邊緣資訊。接著，再由影像的梯度量與方向 (gradient magnitude and direction) 偵測邊緣。然後，偵測邊緣的斷點位置，並以此斷點為起點，做粒子濾波器邊緣延伸以形成封閉輪廓。最後，考慮相鄰區塊的灰階差異，決定是否進行合併以減少過度分割。
我們以紅外線熱感影像 (infrared thermography, IRT)、核磁共振影像 (magnetic resonance imaging, MRI) 、X 光斷層成像 (computed tomography, CT)、及自然影像為實驗影像。在實驗中，將本研究分割方法與 mean shift 及 k-mean 分割方法做比較。計算分割結果中各區塊內之色差平方和 (squared color error) 來評估影像分割結果。本論文使用此準則評估影像分割結果。藉由此準則，可以取得分割區塊內部的灰階變化量。較好的區塊分割，區塊內部的灰階差異小則平均色差平方和小。以此準則評估結果，本論文方法的分割結果優於 mean shift 及 k-mean 的分割結果。

Image segmentation is an important processing step in image analysis and computer vision. The goal of segmentation is to simplify and change the representation of an image into something that is more meaningful and easier to analyze. Similar regions of an image are regions containing common characteristics. Region growing method and gray clustering method are used to segment an image. However these methods are hard to segment image that have low contrast.
In this paper, we propose contour detection by linking broken edges using particle filter. The method includes four parts: image smoothing, edge detection, edge extension, region merge. First, we use the bilateral filter to reduce image noise and to preserve edge. Second, we use gradient magnitude and direction to detect edge. Third, we use the particle filter to link broken edges in order to obtain a closed contour. Finally, the region merging method is used to merge regions with similar gray level.
In the experiment, we use infrared thermography, magnetic resonance imaging, computed tomography, and natural image as experimental images. The segmentation method in this study is comparison to mean shift segmentation and k-mean segmentation. We calculate the squared color error of each region to evaluate the image segmentation results. Experimental results of the proposed method are better than mean shift and k-mean segmentation results in this evaluate criterion.

[1] Achanta, R., A. Shaji, K. Smith, A. Lucchi, P. Fua, and S. Süsstrunk, "SLIC superpixels compared to state-of-the-art superpixel methods," IEEE Trans. on Pattern Analysis and Machine Intelligence, vol.34, no.11, pp.2274-2281, 2012.
[2] Arulampalam, M. S., S. Maskell, N. Gordon, and T. Clapp, "A tutorial on particle filters for online nonlinear/non-Gaussian Bayesian tracking," IEEE Trans. on Signal Processing, vol.50, no.2, pp.174-188, 2002.
[3] Bardera, A., J. Rigau, I. Boada, M. Feixas, and M. Sbert, "Image segmentation using information bottleneck method," IEEE Trans. on Image Processing, vol.18, no.7, pp.1601-1612, 2009.
[4] Batenburg, K. J. and J. Sijbers, "Adaptive thresholding of tomograms by projection distance minimization," Pattern Recognition, vol.42, no.10, pp.2297-2305, 2009.
[5] Beucher, S., "Watersheds of functions and picture segmentation," in Proc. IEEE Int. Conf. on Acoustics, Speech, and Signal Processing . Paris, France, May.3-5, 1982, pp.1928-1931.
[6] Borsotti, M., P. Campadelli, and R. Schettini, "Quantitative evaluation of color image segmentation results," Pattern Recognition Letters, vol.19, no.8, pp.741-747, 1998.
[7] Bresenham, J. E., "Algorithm for computer control of a digital plotter," IBM Systems Journal, vol.4, no.1, pp.25-30, 1965.
[8] Canny, J., "A computational approach to edge detection," IEEE Trans. on Pattern Analysis and Machine Intelligence, vol.8, no.6, pp.679-698, 1986.
[9] Chou, C.-H. and Y.-C. Li, "Perceptually tuned subband image coder based on the measure of just-noticeable-distortion profile," IEEE Trans. on Circuits and Systems for Video Technology, vol.5, no.6, pp.467-476, 1995.
[10] Chung, R. H. Y., N. H. C. Yung, and P. Y. S. Cheung, "An efficient parameterless quadrilateral-based image segmentation method," IEEE Trans. on Pattern Analysis and Machine Intelligence, vol.27, no.9, pp.1446-1458, 2005.
[11] Colorni, A., M. Dorigo, and V. Maniezzo, "Distributed optimization by ant colonies," in Proc. of the first European conf. artificial life, Paris, France, Dec.11-13, 1991, pp.134-142.
[12] Comaniciu, D. and P. Meer, "Robust analysis of feature spaces: color image segmentation," in Proc. IEEE Conf. on Computer Vision and Pattern Recognition, San Juan, CA, Jun.17-19 1997, pp.750-755.
[13] Comaniciu, D. and P. Meer, "Mean shift: a robust approach toward feature space analysis," IEEE Trans. on Pattern Analysis and Machine Intelligence, vol.24, no.5, pp.603-619, 2002.
[14] Cuevas, E., D. Zaldivar, and M. Pérez-Cisneros, "Seeking multi-thresholds for image segmentation with learning automata," Machine Vision and Applications, vol.22, no.5, pp.805-818, 2011.
[15] Ghita, O. and P. F. Whelan, "Computational approach for edge linking," Journal of Electronic Imaging, vol.11, no.4, pp.479-485, 2002.
[16] Gordon, N. J., D. J. Salmond, and A. F. M. Smith, "Novel approach to nonlinear/non-Gaussian Bayesian state estimation," IEE Proc. F Radar and Signal Processing, vol.140, no.2, pp.107-113, 1993.
[17] Jianqing, L. and Y. H. Yang, "Multiresolution color image segmentation," IEEE Trans. on Pattern Analysis and Machine Intelligence, vol.16, no.7, pp.689-700, 1994.
[18] Jung, C. R. and J. Scharcanski, "Robust watershed segmentation using wavelets," Image and Vision Computing, vol.23, no.7, pp.661-669, 2005.
[19] Jung, C. R., "Combining wavelets and watersheds for robust multiscale image segmentation," Image and Vision Computing, vol.25, no.1, pp.24-33, 2007.
[20] Kass, M., A. Witkin, and D. Terzopoulos, "Snakes: Active contour models," Int. Journal of Computer Vision, vol.1, no.4, pp.321-331, 1988.
[21] Liu, J. S., "Metropolized independent sampling with comparisons to rejection sampling and importance sampling," Statistics and Computing, vol.6, no.2, pp.113-119, 1996.
[22] Lu, D. S. and C. C. Chen, "Edge detection improvement by ant colony optimization," Pattern Recognition Letters, vol.29, no.4, pp.416-425, 2008.
[23] Ma, W. Y. and B. S. Manjunath, "Edge flow: a framework of boundary detection and image segmentation," in Proc. IEEE Conf. on Computer Vision and Pattern Recognition, San Juan, CA, Jun.17-19 1997, pp.744-749.
[24] Metropolis, N. and S. Ulam, "The monte carlo method," Journal of the American statistical association, vol.44, no.247, pp.335-341, 1949.
[25] Ng, H. P., S. H. Ong, K. W. C. Foong, P. S. Goh, and W. L. Nowinski, "Medical image segmentation using k-means clustering and improved watershed algorithm," in Proc. 7th IEEE Southwest Symp. on Image Analysis and Interpretation, Denver, CO, March.26-28, 2006, pp.61-65.
[26] Shih, M.-Y. and D.-C. Tseng, "A wavelet-based multiresolution edge detection and tracking," Image and Vision Computing, vol.23, no.4, pp.441-451, 2005.
[27] Tomasi, C. and R. Manduchi, "Bilateral filtering for gray and color images," in Proc. Sixth Int. Conf. Computer Vision, Bombay, India, Jan.4-7 1998, pp.839-846.
[28] Topal, C. and C. Akinlar, "Edge drawing: a combined real-time edge and segment detector," Journal of Visual Communication and Image Representation, vol.23, no.6, pp.862-872, 2012.
[29] Wan, T., N. Canagarajah, and A. Achim, "Segmentation-driven image fusion based on alpha-stable modeling of wavelet coefficients," IEEE Trans. on Multimedia, vol.11, no.4, pp.624-633, 2009.
[30] Wang, W. and C. R. Chung, "Image segmentation with complementary use of edge and region information," Int. Journal of Image and Graphics, vol.11, no.4, pp.549-570, 2011.
[31] Wu, Z. and R. Leahy, "An optimal graph theoretic approach to data clustering: theory and its application to image segmentation," IEEE Trans. on Pattern Analysis and Machine Intelligence, vol.15, no.11, pp.1101-1113, 1993.
[32] Zhu, G., S. Zhang, X. Chen, and C. Wang, "Novel gradient vector flow-based balloon force for active contours," Journal of Electronic Imaging, vol.18, no.2, pp.02300701-02300708, 2009.