AN IMAGE PROCESSING APPROACH FOR THE EVALUATION OF POSTERIOR CAPSULE OPACIFICATION

摘要

This paper presents an image analysis approach for the medical application of assessing of Posterior Capsule Opacifica-tion (PCO) based on digital images of the eye's lens capsule. The approach is based on a simple yet effective image filtering technique that permits the extraction appropriate features from the filtered image. These features reflect to a reasonable extent the amount of PCO in the image. The originality of the approach is that it does not involve any segmentation, contrary to other approaches which are based on segmentation of the image. This process is quite problematic when the images presents irregular texture and spoiled with parasite lights as it is the case for the PCO images. Experimental results from real images confirmed the validity of our approach.