Multi-scale Analysis of Discrete Contours for Unsupervised Noise Detection

Blurred segments [2] were introduced in discrete geometry to address possible noise along discrete contours. The noise is not really detected but is rather canceled out by thickening digital straight segments. The thickness is tuned by a user and set globally for the contour, which requires both supervision and non-adaptive contour processing. To overcome this issue, we propose an original strategy to detect locally both the amount of noise and the meaningful scales of each point of a digital contour. Based on the asymptotic properties of maximal segments, it also detects curved and flat parts of the contour. From a given maximal observation scale, the proposed approach does not require any parameter tuning and is easy to implement. We demonstrate its effectiveness on several datasets. Its potential applications are numerous, ranging from geometric estimators to contour reconstruction.

[1]  J. Koenderink.  The structure of images , 2004, Biological Cybernetics.

[2]  François de Vieilleville,et al.  Maximal digital straight segments and convergence of discrete geometric estimators , 2005, SCIA.

[3]  Isabelle Debled-Rennesson,et al.  Optimal blurred segments decomposition of noisy shapes in linear time , 2006, Comput. Graph..

[4]  Jitendra Malik,et al.  Scale-Space and Edge Detection Using Anisotropic Diffusion , 1990, IEEE Trans. Pattern Anal. Mach. Intell..

[5]  Florent Brunet,et al.  Binomial Convolutions and Derivatives Estimation from Noisy Discretizations , 2008, DGCI.

[6]  Jacques-Olivier Lachaud,et al.  Espaces non-euclidiens et analyse d'image : modèles déformables riemanniens et discrets, topologie et géométrie discrète. (Non-Euclidean spaces and image analysis : Riemannian and discrete deformable models, discrete topology and geometry) , 2006 .

[7]  François de Vieilleville,et al.  Convex Digital Polygons, Maximal Digital Straight Segments and Convergence of Discrete Geometric Estimators , 2007, Journal of Mathematical Imaging and Vision.

[8]  Hong Jeong,et al.  Adaptive Determination of Filter Scales for Edge Detection , 1992, IEEE Trans. Pattern Anal. Mach. Intell..

[9]  Danny Barash,et al.  A Fundamental Relationship between Bilateral Filtering, Adaptive Smoothing, and the Nonlinear Diffusion Equation , 2002, IEEE Trans. Pattern Anal. Mach. Intell..

[10]  Thanh Phuong Nguyen,et al.  Curvature Estimation in Noisy Curves , 2007, CAIP.

[11]  Andrew P. Witkin,et al.  Scale-Space Filtering , 1983, IJCAI.

[12]  François de Vieilleville,et al.  Fast, accurate and convergent tangent estimation on digital contours , 2007, Image Vis. Comput..

[13]  L. Rudin,et al.  Nonlinear total variation based noise removal algorithms , 1992 .

[14]  Steven W. Zucker,et al.  Local Scale Control for Edge Detection and Blur Estimation , 1996, IEEE Trans. Pattern Anal. Mach. Intell..

[15]  Ke Chen,et al.  Adaptive smoothing via contextual and local discontinuities , 2005, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[16]  Jacques-Olivier Lachaud,et al.  Curvature estimation along noisy digital contours by approximate global optimization , 2009, Pattern Recognit..

[17]  Charles Kervrann.  An Adaptive Window Approach for Image Smoothing and Structures Preserving , 2004, ECCV.