Based on the specific morphology of pupillary smooth muscles, a control model for pupillary light reflex was derived. A performance index was formulated, assuming that the pupil was controlled to react quickly and smoothly. The optimal solution derived by Euler-Lagrange equation was highly correlated with experimental data of pupil behavior. Moreover, the model was fitted to the changes of observed pupil diameters before and after mental work-load, for the purpose of quantitative estimation of visual fatigue. The derived parameters expressed the activities of the autonomic nervous system.
A filling in process is an interesting phenomena which surrounding visual information fills into a region. A region which is equiluminance and difference in other attribution to surroundings is easily filled in. This article reports a promotion of filling in process with a flickering neighbor. This suggests an interaction between color and texture channel and motion and space channel in visual process before filling in process is occurred.
This paper is an implementation version of our algorithm proposed in . The chip implemented solves first and second order regularization problems simultaneously which, in turn, enhances contrasts of images after smoothing. A 2μm standard CMOS technology was used with double metal and single poly. The computation is done by the dynamics and its execution time is within several micro seconds.
A Region-based Dichromatic (RBD) method for estimating illumination colons in an image is proposed. In this method, a property that each highlight on dielectric material has peculiar two reflection components is used. The procedure of this method is as follows; first, high luminance regions are extracted, second, they are divided into similar color regions, third, the regions which involve two reflection components are regarded as highlight regions, last, illumination color is estimated from the cross points of highlight lines in the chromaticity diagram. In conventional illumination color estimation method based on the Dichromatic theory, pixels whose colors are similar to illumination color affect the estimation result. The RBD method solved this serious problem.
In shape from shading, contours play an important role as mathematical boundary conditions. On the other hand, contours are difficult to extract from an image. We propose 3D shape estimation method without boundary conditions through the integration of edge and shading information. Assuming that intensity edge corresponds to the initial estimation of shape edge, we can approximately use some constraints of shape edge. This algorithm based on the forward and approximated inverse optics models is able to estimate 3D shape with small number of iterations.
Holography is most expected as a technology of the coming age for recording and repuroducing three dimensional (3-D) image. It has, however, a defect of needing coherent light. We, therefore, have developed a system which has advantages equal to those holography has, based on the principle of pinhole type integral photography. This system does not require coherent light and can make 3-D image zoom in, zoom out and move freely. Without any special glasses, 3-D image can be watched from anypoint with sensation in depth and focus of eyes can be controlled to any desired point. We manufactured a small working model to verify the principle and obtained a successful experimental result. This system has wide applications in science, education, art appreciation and amusement, etc..