作者: Gildas Marin , Marc Le Renard
DOI:
关键词:
摘要: In the image processing method, a first two-dimensional (l(x, y)) of three-dimensional scene is obtained. Each pixel location associated with respective distance parameter (z(x, representing viewer's eye to an object appearing at this in image. A spreading (Φd(x, determined for each as function location, dioptric power (Pcorr) portion correction lens simulated placed along ray extending between and said image, maximum accommodation (Pacc) eye. second, blurred (l'(x, generated from using local convolution kernel varying across This can be made sparse, i.e. number non-zero points much smaller than blur region whose dimension given by parameter.