Article ID Journal Published Year Pages File Type
536348 Pattern Recognition Letters 2005 6 Pages PDF
Abstract

A novel approach of removing eyeglasses from frontal facial images is proposed. The region occluded by eyeglasses is firstly detected; a natural looking eyeglassless facial image is then synthesized by recursive error compensation of PCA reconstruction. The synthesized images have neither trace of the eyeglasses, nor the reflection and shade caused by the eyeglasses. Experimental results show that our method can effectively synthesize eyeglassless facial images.

Related Topics
Physical Sciences and Engineering Computer Science Computer Vision and Pattern Recognition
Authors
, ,