Human visual system has a space-variant resolution nature. In the retinal receptive field, the resolution is not uniform but sampled finest in the central fovea and coarser in the peripheral. This variable resolution mapping function is born by the cerebral primary visual cortex V1. It has a clear visual field map of spatial information, and this spatial mapping structure is called Retinotopy. The forward mapping to visual cortex from retina is characterized with complex LPT (Log-PolarTransform) by Schwartz. The retinal receptive field image is reconstructed by inverse projection LPT-1 from V1. This reconstructed process is called F oveated I maging. Since the spatial information is concentrated in the center of the visual field, the Foveated Imaging is applied to image compression, pattern recognition, robot vision, and/or computer vision. The retinal receptive field image is suitable for material appearance expression with natural blurring due to peripheral vision.<br/> However, the complexity of the inverse transform LPT-1 was a bottleneck. This paper proposes a Double- Ring-structured novel Foveated Imaging method using positive and negative Gaussian blur masks without using the inverse transform LPT-1 of Schwartz theory and reports the evaluation of reproduction errors.
The electrical activity in a photoreceptor is initiated when photons are absorbed by photopigment molecules of the cell. When the receptor is exposed to a photon flux of a particular wavelength, the actual number of photons absorbed in a cell varies with Poisson fluctuation. This fluctuation introduces a spatial variation in absorption by cells and a temporal variation, with repeated exposure, in the number of cells absorbing each specific level of light energy. Here we characterize such variations and quantify the relationship between the spatial and temporal variations for an array of receptors exposed to an arbitrary light spectrum. The spatial variation in absorption by cone cells implies that visual stimulation produces a distribution of responses in cone excitation space. We show that the resulting excitations directly reproduce MacAdam's (1942) classic measurements of the variability of color matches. Our model applies to both a living and a non-living array of photosensitive elements. We carried out a performance evaluation by a CMOS sensor repeatedly exposed to uniformly illuminated color patches. Our findings suggest that spatial fluctuation in absorbed light energy by cells is invariant with respect to the total number of like-type cells from which the histogram is obtained. However, temporal variation with repeated exposure in the proportion of cells at a specific level of absorbed light energy decreases as the spatial variance and number of pixels increases. Our results also support the assumption that each cell absorbs light energy independently from the other cells. The proposed characterization is important for understanding the retinal factors that limit color detection and discrimination in the human visual system. It also has technical significance for color image enhancement in imaging by a digital sensor.