Classification of photorefraction images using neural networks

The early evaluation of the visual status of human infants is of critical importance. Among the objective methods of refraction photorefractive techniques are specifically designed for screening young children. Over the years we have set-up a number of systems with different grades of complexity and automation. One problem that we have to deal with in any approach is the interpretation and classification of the acquired photorefraction images. Previously we used conventional image processing operators and Fourier techniques. In this paper we report on the use of neural networks for automated classification of photorefraction images