Fundus Images Segmentation by Unsupervised Classification

In this paper, we present an original unsupervised segmentation scheme which splits a grey level image into different sets of connected pixels whose grey levels are homogeneous. This approach is based on an analysis of a triangular table denoted ”Normalized connectivity degrees pyramid”. This method is used in order to detect cytomegalovirus retinitis lesions by fundus image analysis. First, we determine the number of pixels classes and their cores. The core of each class Cj is represented by an interval of grey levels[minCj ;maxCj ]. For classification purpose, the pixels whose grey level belongs to such an interval are labelled to the corresponding class. The other pixels are assigned by comparison of their conditional probability to belong to the different classes.

[1]  Ludovic Macaire,et al.  Automated visual inspection of galvanized and painted metallic strips , 1993, 1993 CompEuro Proceedings Computers in Design, Manufacturing, and Production.

[2]  Jzau-Sheng Lin,et al.  The application of competitive Hopfield neural network to medical image segmentation , 1996, IEEE Trans. Medical Imaging.

[3]  Jeffrey E. Boyd,et al.  Automated diagnosis and image understanding with object extraction, object classification, and inferencing in retinal images , 1996, Proceedings of 3rd IEEE International Conference on Image Processing.

[4]  Sankar K. Pal,et al.  A review on image segmentation techniques , 1993, Pattern Recognit..

[5]  Raimondo Schettini,et al.  A segmentation algorithm for color images , 1993, Pattern Recognit. Lett..

[6]  Supun Samarasekera,et al.  Multiple sclerosis lesion quantification using fuzzy-connectedness principles , 1997, IEEE Transactions on Medical Imaging.

[7]  Michele Nappi,et al.  Different methods to segment biomedical images , 1997, Pattern Recognit. Lett..

[8]  Alain Trémeau,et al.  A region growing and merging algorithm to color segmentation , 1997, Pattern Recognit..