This paper addresses the issue of accurate lesion segmentation in retinal imagery, using level set methods anduda novel stopping mechanism - an elementary features scheme. Specifically, the curve propagation is guidedudby a gradient map built using a combination of histogram equalization and robust statistics. The stoppingudmechanism uses elementary features gathered as the curve deforms over time, and then using a lesionnessudmeasure, defined herein, ’looks back in time’ to find the point at which the curve best fits the real object.udWe compare the proposed method against five otherudsegmentation algorithms performed on 50 randomly selected images of exudates with a database of clinicianuddemarcated boundaries as ground truth.
展开▼