Active contours for specific target detection

Abstract

A novel region-based active contour model for specific target segmentation is proposed. This model has a fitting term in the form of a quadratic parabola, which identifies targets based on their averages and variations. The proposed method is able to detect a specific target as edge-based active contours, but is more robust with respect to noise and fuzzy edge. Experiments on synthetic and real images demonstrate the effectiveness of the proposed method.