Contour-contrast enhancement based on retinal layer processing

Abstract

The Human Visual System (HVS) model based image quality metrics, correlates strongly with the evaluations of image quality as well as with human observer performance in visual recognition process. Physiological modeling of retina plays a vital role in the development of high performance image processing methods to produce better visual perception. The retina is a complex neural structure, which does not only detect an incoming light signal, but also proceeds to many complex signal transformations. The retina is a laminated structure consisting of alternating layers of cell bodies and cell processes. The main aim of this work is to develop a model of photoreceptor and outer plexiform layer considering the properties of compression and spatiotemporal filtering in the processing of visual information. In this work spatial frequency value is evaluated using Discrete cosine Transform (DCT) technique thereby enhancing the contrast visibility in dark area and maintaining the same in bright area using photoreceptor layer of retina and a model to enhance contour contrast is developed using outer plexiform layer of retina.