Eyes Location Method Based On The Adaboost Algorithm And Block Integral Projection

In this paper, an effective method of eye location based on the AdaBoost algorithm and block integral projection is proposed. First the eye classifier based on the AdaBoost algorithm is used to quickly detect the eye candidate region in face region image. Then according to the position and proportion of the eyes, the eye candidate region is divided into three sub-regions, the eyes positions can be precisely located by using gray integral projection in the left-eye and right-eye sub-regions. The experimental results show that the block integral projection can prevent the emergence of pseudo feature points because of face tilt, and the influences of the spectacle border and spectacle frame can be avoided. The proposed method has better real-time performance and locating accuracy, it has practical value.