ABSTRACT
Retinal images have the potential to facilitate the early detection of retinal pathologies. The clinically important features of a retinal image are the optic disc (OD), the macula, and the blood vessels. The ophthalmologists examine these features for signs of various eye-related diseases. The retinal image should be of sufcient quality for reliable diagnosis. In some applications like telemedicine, these images need to be processed for efcient storage and transmission. Any such processing causes distortion and degrades the quality of the image. But the important question is: will the processed image be clinically useful as the original? If the answer is yes, then it is required that little distortion occurs to the clinical information in the processed images. Then the processed image is of similar quality as that of the original. How does one evaluate the quality? The image quality is an attribute that has many denitions and interpretations depending on
CONTENTS
4.1 Image Quality Measures ..................................................................................................... 70 4.2 Wavelet Subband Analysis of Retinal Features ...............................................................72
