Distant Iris Recognition Using Deep Feature-based Classifications

For human identification, a promising biometric technique is using iris recognition from a face image acquired at a distance. Image preprocessing, feature extraction, and reliable classification techniques all are the components of this contemporary field of research. After separating eye images from an image of the entire face of a subject and also separating iris image from this eye by using segmentation, then extracting features from these iris images we want to utilize a convolutional neural network (CNN) model in this work. In general, convolutional neural network (CNN) models convolve images in different layers to extract effective features and then produce a probability output in the final layer by using the softmax function. In our proposed approach, we have used different classifications technique with CNN features. We use the publicly available CASIA.v4 distance image database. At last, we utilize several classification techniques on this database for human identification approaches, the experimental results demonstrate that our proposed approach obtains a satisfactory recognition result for the KELM classifier.

• Supervision