The study of algorithms for iris recognition is an area of continued growth. Although a variety of methods have been proposed, this paper focuses on the analysis and implementation of an algorithm which consists of the basic steps for iris recognition- including preprocessing, segmentation, and normalization, as well as features extraction and matching- on a low-cost, low-power ARM based board, BeagleBone Black (Rev C), using the open source computer vision library OpenCV. Experimental results show the feasibility of the device in the system implementation, provided that some algorithm stages are improved in order to reduce the processing time.