A New Local Descriptor Based on Strings for Face Recognition
Algorithm 4
Classification of the feature vectors.
Input: Feature vector of a new face image
Output: The class of the image
ā
Step 1: Repeat the steps in Algorithm 2 for a test image.
ā
Step 2 (NNC): Measure the similarity between the feature vector representing the new face image and all feature vectors in the database using HD. The class (the person) is located by the vector which has the lower distance with the reference vector.