Iris authentication device, iris authentication method, and recording medium

An authentication method comprising:

• by at least one processor,
• acquiring a first image by capturing an authentication target, the authentication target being moving in a predetermined direction;
• acquiring a second image by capturing a left eye or a right eye of the authentication target;
• determining whether the eye included in the second image is of the left eye or the right eye of the authentication target using the first image;
• calculating a first verification score by comparing characteristic information on an iris of the eye in the second image with one or more sets of characteristic information about irises of eyes in a plurality of registrants stored in a storage, each of the plurality of registrants being an object to be authenticated;
• detecting an overlap between a region of the first image including the second image and a predetermined region of the first image, the predetermined region being a region where the left eye or the right eye of the authentication target is estimated to be captured using a machine learning engine, the machine learning engine receiving the first image and outputting information on the region of the first image where the left eye or the right eye of the authentication target is estimated to be captured;
• calculating a first weighted verification score by weighting the first verification score with a result of detection about the overlap;
• authenticating the authentication target using the first weighted verification score;
• and
• outputting an authentication result.

A non-transitory storage medium storing a computer program for causing a computer to implement:

• acquiring a first image by capturing an authentication target, the authentication target being moving in a predetermined direction;
• acquiring a second image by capturing a left eye or a right eye of the authentication target;
• determining whether the eye included in the second image is of the left eye or the right eye of the authentication target using the first image;
• calculating a first verification score by comparing characteristic information on an iris of the eye in the second image with one or more sets of characteristic information about irises of eyes in a plurality of registrants stored in a storage, each of the plurality of registrants being an object to be authenticated;
• detecting an overlap between a region of the first image including the second image and a predetermined region of the first image, the predetermined region being a region where the left eye or the right eye of the authentication target is estimated to be captured using a machine learning engine, the machine learning engine receiving the first image and outputting information on the region of the first image where the left eye or the right eye of the authentication target is estimated to be captured;
• calculating a first weighted verification score by weighting the first verification score with a result of detection about the overlap;
• authenticating the authentication target using the first weighted verification score;
• and
• outputting an authentication result.