Adaptive Image Filtering and Segmentation Using Robust Estimation of Intensity
A structure-adaptive approach to robust statistical estimation of image intensity for adaptive filtering and segmentation of images is described in the context of a two-region structural image model. The proposed adaptive estimation procedure is based on the selection of best fitting structuring region relatively to a current point from all available multiple structuring regions by the maximum a posteriori probability principle. In application to image filtering, the described method allows to suppress noise and, at the same time, not damage the initial image including corner edges and image fine details. It provides also a robust binary segmentation of local objects of interest and their edges on noisy background.
KeywordsRobust Estimation Impulsive Noise Polynomial Regression Model Binary Segmentation Robust Parameter Estimation
- 2.V. Koivunen. “A robust nonlinear filter for image restoration”, IEEE Trans., Vol. IP-4, (1995), 569–578.Google Scholar
- 3.G. Danuser and M. Strieker. “Parametric model fitting: from inlier characterization to outlier detection”, IEEE Trans., Vol. PAMI-20, No. 2, (1998) 263–280.Google Scholar
- 5.C. V. Stewart. “MINPRAN: a new robust estimator for computer vision”, IEEE Trans. Vol. PAMI-17, No. 10, 1995) 925–938.Google Scholar
- 9.P. M. Djuric. “Asymptotic MAP criteria for model selection”, IEEE Trans., Vol. SP-46, No. 10, (1998) 2726–2735.Google Scholar
- 10.R. M. Palenichka and I. B. Ivasenko. “Fast and robust parameter estimation in the polynomial regression model of images”, SPIE Proc, Nonlinear Image Processing X, Vol. 3646, (1999) 28–37.Google Scholar