Abstract

The iris of the human eye has a very unique and different pattern on every human, so
it is possible to use it as a known biometric recognition basis
with iridology. Iridology is the method of reading the map on the eye to detect
several types of diseases using iris eye observation patterns. In this research
the author takes iris data using a digital camera, but the image of the iris
obtained still looks blurred so it requires processing to reduce
vagueness. The author designed the software to improve image quality iris images
eyes that have cholesterol symptoms. The method used is the extraction of the characteristics of the moment
invariant, assisted by K-Means Clustering algorithm for central distance calculation
clusters on the iris image. From the results of the iris image that has been tested can
grouped into normal iris and iris eyes of high cholesterol. In
this research accuracy level of research data is equal to 95%.