Iris liveness detection for mobile devices

摘要

An approach for an iris liveness detection is provided. A plurality of image pairs is acquired using one or more image sensors of a mobile device. A particular image pair is selected from the plurality of image pairs, and a hyperspectral image is generated for the particular image pair. Based on, at least in part, the hyperspectral image, a particular feature vector for the eye-iris region depicted in the particular image pair is generated, and one or more trained model feature vectors generated for facial features of a particular user of the device are retrieved. Based on, at least in part, the particular feature vector and the one or more trained model feature vectors, a distance metric is determined and compared with a threshold. If the distance metric exceeds the threshold, then a first message indicating that the plurality of image pairs fails to depict the particular user is generated. It is also determined whether at least one characteristic, of one or more characteristics determined for NIR images, changes from image-to-image by at least a second threshold. If so, then a second message is generated to indicate that the plurality of image pairs depicts the particular user of a mobile device. The second message may also indicate that an authentication of an owner to the mobile device was successful. Otherwise, a third message is generated to indicate that a presentation attack on the mobile device is in progress.

主权项

1. A method comprising:
acquiring a plurality of image pairs using one or more image sensors of a mobile device; selecting a particular image pair, from the plurality of image pairs, that depicts an eye-iris region in-focus; generating a hyperspectral image by fusing images included in the particular image pair based on, at least in part, the particular image pair; generating a particular feature vector for the eye-iris region depicted in the particular image pair based on, at least in part, the hyperspectral image; retrieving one or more trained model feature vectors generated based on depictions of one or more facial features of a particular user of the mobile device; determining a distance metric based on, at least in part, the particular feature vector and the one or more trained model feature vectors; determining whether the distance metric exceeds a first threshold; and generating a first message indicating that the plurality of image pairs fails to depict the particular user of the mobile device in response to determining that the distance metric exceeds the first threshold.