A new method of detecting human eyelids based on deformable templates

In this work a new method of automatically detecting human eyelids using deformable templates is proposed. Deformable templates are often used in eye´s feature detection although there are several drawbacks such as high computational complexity, unexpected shrinking of the templates, adjusting the weights of energy terms and failure of orientation adjustment. Based on the analysis on these problems, an improved template for eyelid edge detection with several new ideas is presented. The problems of unexpected shrinking of templates, adjusting the weights of energy terms and failure of orientation adjustment are overcome by using a new kind of potential field and accurate estimations of the template parameters. Experiments show that the eyelid template and the optimization method are effective even for complex changes in facial images, such as eye opening and closing, obvious irises movements and rotations of heads in small angles.