Paper Details: Downloads: 947
Serial Number: P1152336725
Title: Iris Recognition using high order statistical features and SVM classifier
Authors: Rafika HARRABI
Abstract: This paper presents a new iris recognition method based on a modified Fuzzy C-Means (MFCM) and SVM algorithms. The proposed technique is a hybrid technique used to find out the identity of human iris image. It uses the pre-processing algorithms: Fuzzy c-means algorithm, and high order statistical features to localize the iris in the eye images and uses the Principal Component Analysis (PCA) with SVM algorithm for the classification process. In the proposed method, the recognition system operates in two modes: pre-processing and classification. The first stage attempts to isolate the iris from its surroundings, using high order statistical features and Fuzzy C-means algorithm. The second stage defines the person by classifying the segments obtained from the initial stage, where each iris is classified using SVM algorithm. This specific classifier must determine whether these new images belong to one of the 108 classes. The CASIA iris dataset is used to assess our proposed method. During the pre-processing stag
Keywords: Keywords—Statistical features; second order Statistical features; Fuzzy c-means; PCA; SVM.
Journal/Conference: International Journal of Graphics, Vision and Image Processing
Volume: 24
Issue: 1
Submission Date: 9/6/2023 12:00:00 AM
Review Date: 9/24/2023 12:00:00 AM
Publishing Date: 1/5/2024 12:00:00 AM
Article Downloads: 947
Download:

Facebook