b1) Facial Features
Ex. 1. A method proposed by Leung et.al. [12]:
Facial features extraction and face detection steps:
- The original image is convolved with a set of Gaussian derivatives filters.
- The description of the local image brightness is made by means of a vector R(x,y) – spatial location.
- The candidate for ith facial feature are found by matching R(x,y) against a template vector of responses . We take into account: