A NEW GAUSSIAN DISTRIBUTION BASED ADAPTIVE SIMULATION ALGORITHM DEVELOPED FOR ARTIFICIAL SIGHT SYSTEMS


Ozden M., KARAGÖZ İ.

JOURNAL OF THE FACULTY OF ENGINEERING AND ARCHITECTURE OF GAZI UNIVERSITY, cilt.26, sa.3, ss.691-700, 2011 (SCI-Expanded) identifier identifier

Özet

Simulation of the visual prosthesis system is a very important step in the development of visual prosthesis systems which aim to sight restoration for totally blind persons. Although some limited works have been published in the literature, there is still a need for prosthetic vision simulation algorithms for better understanding of the prosthetic vision systems that present a very low resolution phosphene image to the patients. In this study, a new visual prosthesis simulation algorithm that transforms the color image sequences into "phosphene images" using Gaussian distribution as an adaptive approach was developed. The main processing steps of the algorithm were described in detail. To obtain quantitative results about the performance of the algorithm, mean squared error (MSE) values between segmented versions of the original and the phosphene images were compared. In addition, the comparisons were performed with the algorithms that use normal distributed spots and standard Gaussian distributed spots in literature. In these comparisons as well as MSE criteria, the preservation of some important visual cues on the images (e.g. corners, region numbers, and edges) is taken into account. Obtained results are commented and presented as a bar graphics and contribution of the study for prosthetic vision systems is discussed.