After that, significant key point c

After that, significant key point candidates are localized and refined by eradicating the low contrast points. Then key-point orientation assignment based on local image gradient is performed and in the last step, a descriptor is generated to compute the local image descriptor for each key point based on image gradient magnitude and orientation. Number of key-points generated for various samples of iris images varies from 95 to 250. First 20 key-points are kept for further processing in order to make comparisons. Each key-point is a descriptor vector of length 128. The key-point is uniquely identified by this feature vector. The obtained feature vector is invariant to rotation, scale variations and is immune to illumination changes.