Abstract
It is widely accepted that the design of morphological filters, which are optimal in some sense, is a difficult task. In this paper a novel method for optimal learning of morphological filtering parameters (Genetic training algorithm for morphological filters, GTAMF) is presented. GTAMF adopts new crossover and mutation operators called the curved cylinder crossover and master-slave mutation to achieve optimal filtering parameters in a global searching. Experimental results show that this method is practical, easy to extend, and markedly improves the performances of morphological filters. The operation of a morphological filter can be divided into two basic problems including morphological operation and structuring element (SE) selection. The rules for morphological operations are predefined so that the filter's properties depend merely on the selection of SE. By means of adaptive optimization training, structuring elements possess the shape and structural characteristics of image targets, and give specific information to SE. Morphological filters formed in this way become certainly intelligent and can provide good filtering results and robust adaptability to image targets with clutter background.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
References
Serra J. Image Analysis and Mathematical Morphology. Academic Press, London, 1988, pp. 217–245.
Maragos Pet al. Morphological system for multidimensional signal processing. InProc. IEEE, 1990, 78: 690–710.
Dougherty E R,et al. Digital Image Processing Methods. Marcel Dekker, New York, 1994, pp. 110–138.
Gong Wet al. Mathematical Morphology in Digital Space. Science Press, Beijing, 1997. (in Chinese)
Rudolph G. Convergence analysis of canonical genetic algorithms.IEEE Trans. Neural Networks, 1994, 5(1): 96–101.
Mahalb Uet al. Genetic algorithm for optical pattern recognition.Optics Letters, 1991, 16(9): 648–650.
Bhandarkar S M, Zhang H. Image segmentation using evolutionary computation.IEEE Trans. Evolutionary Computation, 1999, 3(1): 1–21.
Ehrhardt R. Morphological filter design with genetic algorithms.SPIE, 2300, 1994, pp. 2–12.
Huttunen Het al. Optimization of soft morphological filters by genetic algorithms.SPIE, 2300, 1994, pp. 13–24.
Chen J M. Rudiment for the generalized evolvement theory. Chinese Science Bulletin, 1999, 8(16): 1786–1792.
Srinivas Met al. Adaptive probabilities of crossover and mutation in genetic algorithms.IEEE Trans. Systems, Man and Cybernetics, 1994, 24(4): 656–667.
Zhang J S, Xu Z B, Liang Y. Integrated anneal genetic algorithm and its sufficient convergence precondition.Science in China (Series E), 1997, 27(2): 154–164.
Kohonen T. Self-Organization and Associative Memory, Springer-Verlag, Third Edition, 1989.
Vapnik V N. The Nature of Statistical Learning Theory. New York: Springer-Verlag, 1995.
Cortes C, Vapnic V. Support-vector networks.Machine Learning, 1995, 20(3): 273–297.
Wang R S. Image Understanding. NUDT Publishing House, 1995. (in Chinese)
Eren P Eet al. Robust, object-based high-resolution image reconstruction from low-resolution video.IEEE Trans. Image Processing, 1997, 6(10): 1446–1451.
Kim S P, Su W Y. Recursive high-resolution reconstruction of blurred multi-frame image.IEEE Trans. Image Processing, 1993, 2(4): 534–539.
Wang Y Jet al. Initialization research of the receptive field model for weights in BP learning networks.Progress in Natural Science, 1996, 6(3): 346–350.
Author information
Authors and Affiliations
Corresponding author
Additional information
This work is supported by the Shanghai Scientific Technology Development Foundation and Shanghai Post-Doctoral Research Foundation.
YU Nong was born in 1962. He received the Ph.D. degree from Institute of Electronic Science and Engineering, National University of Defense Technology, Changsha, China, in 2000. Now he is engaged in post-doctoral research work in Shanghai Institute of Technical Physics, Chinese Academy of Sciences. His research interests include avionics, computer vision, and automatic target recognition.
WU Hao was born in 1976. She received the B.S. degree in information engineering and M.S. degree in signal and information processing from Institute of Electronic Engineering at the National University of Defense Technology (NUDT), China, in 1997 and 2000 respectively. She is currently a Ph.D. candidate in the Institute of Electronic Engineering of NUDT, China. Her research interests include image processing, computer vision, pattern recognition, and computer graphics.
WU ChangYong was born in 1942. He is currently a professor and Ph.D. supervisor in Shanghai Institute of Technical Physics, Chinese Academy of Sciences. His research interests are infrared photoelectric technology and automatic target detection.
LI YuShu was born in 1963. She received the B.S. degree in medicine from Luzhou Medical College and M.S. degree in hydroelectric engineering from Huazhong University of Science and Technology, China, in 1985 and 2002 respectively. Her research interests include aeronautical maintenance engineering and educational psychology.
Rights and permissions
About this article
Cite this article
Yu, N., Wu, H., Wu, C. et al. Automatic target detection by optimal morphological filters. J. Comput. Sci. & Technol. 18, 29–40 (2003). https://doi.org/10.1007/BF02946648
Received:
Revised:
Issue Date:
DOI: https://doi.org/10.1007/BF02946648