Automatic Image Segmentation Using Active Contours with Univariate Marginal Distribution

Abstract

This paper presents a novel automatic image segmentation method based on the theory of active contour models and estimation of distribution algorithms. The proposed method uses the univariate marginal distribution model to infer statistical dependencies between the control points on different active contours. These contours have been generated through an alignment process of reference shape priors, in order to increase the exploration and exploitation capabilities regarding different interactive segmentation techniques. This proposed method is applied in the segmentation of the hollow core in microscopic images of photonic crystal fibers and it is also used to segment the human heart and ventricular areas from datasets of computed tomography and magnetic resonance images, respectively. Moreover, to evaluate the performance of the medical image segmentations compared to regions outlined by experts, a set of similarity measures has been adopted. The experimental results suggest that the proposed image segmentation method outperforms the traditional active contour model and the interactive Tseng method in terms of segmentation accuracy and stability.

Nondestructive inspection of aerospace composite laminate using thermal image processing

“…In the present experiment, Teflon pieces having a thickness of 0.1 mm and a size of 5 × 5 mm were embedded during the primary bonding of the laminate. The defects were embedded at 16 locations (layers 2, 4, 5,6,8,10,11,12,14,16,17,18,20,22,24,25) which corresponded to various depths in the laminate. Though it appears to be one sample, the defects were induced at different layers to create multiple scenarios of de-bonding.…”

“…In their work, several samples as varied as wood images, pear and potato images, stereogram of bent piece of paper and also a video sequence of human lip movements were investigated using active contours. Since then, active contours have been widely used in different areas such as the medical field [18][19][20][21] and urban planning [22][23][24] of which, studies in the medical field form the major chunk.…”

“…However, there are a few difficulties in the usage of active contours [18][19][20][21]. The first problem is the initialization of control points, which must be close to the object of interest; otherwise, convergence does not happen.…”

“…The second drawback is the propensity to stagnate in local minima giving an inaccurate convergence to the boundaries of the object. To address these shortcomings of the traditional active contour method (ACM), the univariate marginal distribution algorithm (UMDA) was adopted by [18]. Cruz-Aceves et al [19] proposed a new image segmentation method based on multiple active contours guided by differential evolution, called MACDE.…”

Anatomically Based Geometric Modelling Using Medical Image Data: Methods and Programs

“…Ramanujam Kayalvizh et al proposed a new intelligent arithmetic, that is, the Particle Swarm Optimization (PSO) [ 1 ]. I. Cruz-Aceves et al put forward an automatic image segmentation method based on active contour model theory and estimation of distribution algorithms [ 2 ]. Nihar Ranjan Nayak et al proposed improved clustering algorithm which is applicable to gray images by evaluating three different kinds of clustering algorithms [ 3 ].…”