Application of Colour Image Processing Techniques on Identification of Retinal Hard Exudates

Retinal exudates are a characteristic feature of many retinal diseases such as Diabetic Retinopathy. In this paper we address the development of a method to evaluate and quantitatively diagnose these random yellow patches in colour retinal images automatically. After an initial colour normalisation and contrast enhancement pre-processing step, colour retinal image pixels are classified to exudate and non-exudate classes. Gaussian quadratic, K nearest neighbour and Gaussian mixture model classifiers are examined within the pixel-level exudate recognition framework. A Gaussian mixture model-based classifier provided the best classification performance with 89.2% sensitivity and 81.0% predictivity in terms of pixel-level accuracy.

Keywords:

Diabetic retinopathy , exudate identification , retinal image