In this research, a new automatic two-phase snake algorithm for segmenting multiple objects is proposed. The “automatic”comes from the distribution of grid active-points over the entire image so as to determine the initial snake contours at proper positions without manual interaction. Our algorithm is also capable of segmenting multiple or annular objects (more versatile than traditional snake algorithms). All processing is based on our proposed AGVF field which adopts space-varying weighting functions to provide a stable and uniform performance over a broad range of image SNRs. In the active-contours phase, a no-search strategy is adopted so as to increase processing speed and efficiency. The snake points move with the guidance of internal and external forces which are computed directly from the current snake contour status and AGVF vectors, respectively.
Hence, neither global nor local evaluation of the next movement is required.
One disadvantage of the GVF-based algorithm is that the GVF/AGVF fields must be obtained prior to snake deformation and it actually spends most CPU time, e.g., about 2 seconds on a Pentium III 1G Hz processor for a 256×256 image.
Fortunately, a multi-resolution GVF computing architecture has been recently proposed (Ntalianis, Doulamis, Doulamis, & Kollias, 2001) to speed up the computation significantly by nearly 40 times. Our two-phase AGVF snake processing can make use of a 2-level AGVF field to further improve the speed and efficiency.
The proposed algorithm will be versatile anywhere automatic detection or segmentation of multiple objects is required, e.g., in an image retrieval system and biomedical image processing and analysis.
REFERENCES
Chuang, C. H., & Lie, W. N. (2001). Automatic snake contours for the segmentation of multiple objects. Proceedings of IEEE International Symposium on Circuits and Systems II, Sydney, Australia, 389-392.
Davison, N. E., Eviatar, H., & Somorjai, R. L. (2000). Snakes simplified. Pattern Recognition, 33(10), 1651-1664.
Deng, J., & Tsui, H. T. (2002). A fast level set method for segmentation of low contrast noisy biomedical images. Pattern Recognition Letters, 23(1-3), 161-169.
Ji, L., & Yan, H. (2002a). Attractable snakes based on the greedy algorithm for contour extraction. Pattern Recognition, 35(4), 791-806.
Ji, L., & Yan, H. (2002b). Robust topology-adaptive snakes for image segmentation. Image and Vision Computing, 20, 147-164.
Kang, D. J. (1999). A fast and stable snake algorithm for medical images. Pattern Recognition Letters, 20(5), 507-512.
Kass, M., Witkin, A., & Terzopoulos, D. (1988). Snakes: active contour models.
International Journal of Computer Vision, 1(4), 321-331.
Lam, C. L., & Yuen, S. Y. (1998). An unbiased active contour algorithm for object tracking. Pattern Recognition Letters, 19(5-6), 491-498.
Lam, K. M. & Yan, H. (1994a). Fast greedy algorithm for active contours.
Electronics Letters, 30(1), 21-23.
Lam, K. M., & Yan, H. (1994b). Locating head boundary by snakes. Proceedings, International Symposium on Speech, Image Processing and Neural Networks, (ISSIPNN’94), Hong Kong, China, 17-20.
Lie, W. N., & Chuang, C. H. (2001). A fast and accurate snake model for object contour detection. Electronics Letters, 37(10), 624-626.
Looney, C. G. (1997). Pattern recognition using neural networks. New York, USA:
Oxford University Press.
Matsuda, Y., Sumi, Y., Kataoka, D., Ota, M., Yabuki, N., Fukui, Y., & Miki, S.
(2000). Proposal for a convergence criterion to the active net in two steps.
Proceedings of IEEE International Symposium on Circuits and Systems, Geneva, Switzerland, 313-316.
Mirhosseini, A. R., & Yan, H. (1997). An optimally fast greedy algorithm for active contours. Proceedings of IEEE International Symposium on Circuits and Systems, Hong Kong, China, 1189-1192.
Ntalianis, K. S., Doulamis, N. D., Doulamis, A. D., & Kollias, S. D. (2001).
Multiresolution gradient vector flow field: a fast implementation towards video object plane segmentation. Proceedings of IEEE International Conference on Multimedia and Expo, Tokyo, Japan, 1-4.
Osher, S., & Fedkiw, R. P. (2001). Level set methods: an overview and some recent results. Journal of Computational Physics, 169(2), 463-502.
Pardo, J. M., Cabello, D., & Heras, J. (1997). A snake for model-based segmentation of biomedical images. Pattern Recognition Letters, 18(14), 1529-1538.
Pardo, X. M., & Cabello, D. (2000). Biomedical active segmentation guided by edge saliency. Pattern Recognition Letters, 21(6-7), 559-572.
Park, J., & Keller, J. M. (2001). Snakes on the watershed. IEEE Transactions on Pattern Analysis and Machine Intelligence, 23(10), 1201-1205.
Precioso, F., & Barlaud, M. (2002). B-spline active contour with handling of topology changes for fast video segmentation. EURASIP Journal on Applied Signal Processing, 6, 555-560.
Sonka, M., Hlavac, V., & Boyle, R. (1999). Image processing, analysis, and machine vision. Pacific Grove, California, USA: Brooks/Cole, 228-289.
Velasco, F. A., & Marroquin, J. L. (2003). Growing snakes: active contours for complex topologies. Pattern Recognition, 36(2), 475-482.
Williams, D. J., & Shah, M. (1992). A fast algorithm for active contours and curvature estimation. CVGIP: Image Understanding, 55(1), 14-26.
Wong, Y. Y., Yuen, P. C., & Tong, C. S. (1998a). Contour length terminating criterion for snake model. Pattern Recognition, 31(5), 597-606.
Wong, Y. Y., Yuen, P. C., & Tong, C. S. (1998b). Segmented snake for contour detection. Pattern Recognition, 31(11), 1669-1679.
Xu, C., & Prince, J. L. (1998). Snakes, shapes, and gradient vector flow. IEEE Trans. Image Processing, 7(3), 359-369.
Yuen, P. C., Wong, Y. Y., & Tong, C. S. (1996). Contour detection using enhanced snakes algorithm. Electronics Letters, 32(3), 202-204.
Yuen, P. C., Feng, G. C., & Zhou, J. P. (1999). A contour detection method:
initialization and contour model. Pattern Recognition Letters, 20(2), 141-148.
Cheng-Hung Chuang received a B.S. degree in electrical engineering from Tatung Institute of Technology, Taiwan, in 1994, and M.S. and Ph.D. degrees in electrical engineering from National Chung Cheng University, Taiwan, in 1996 and 2003, respectively. From 2003 to 2007, he was a member of the Institute of Statistical Science, Academia Sinica, Taipei, Taiwan, as a postdoctoral fellow for human brain science research. In Feb. 2007, he entered the Department of Computer Science and Information Engineering, Asia University, Taiwan, as an assistant professor. His research interests include image/video processing, optical and biomedical signal processing, 3-D display systems, computer vision, and pattern recognition.
Wen-Nung Lie was born in Tainan, Taiwan, R. O.
C., in 1962. He received B.S., M.S., and Ph.D. degrees, all in electrical engineering, from National Tsing Hua University, Taiwan, R. O. C., in 1984, 1986, and 1990, respectively. From September 1990 to June 1996, he served as an assistant scientist in Chung Shan Institute of Science and Technology (CSIST), Taiwan, R. O. C., where he was devoted to the development of an infrared imaging system and target tracker for military applications.
In August 1996, he joined the Department of Electrical Engineering, National Chung Cheng University, Taiwan, R. O. C., as an associate professor. From July 2000 to Jan. 2001, he was a visiting scholar in the University of Washington, Seattle, WA, USA. His research interests include image/video compression, networked video transmission, audio/image/video watermarking, multimedia content analysis, standard-compliant multimedia encryption, infrared image processing, 3-D stereo virtual reality by image-based rendering, and industrial inspection by computer vision techniques.