Performance Improvement of Automatic Facial Midline Detection by Chain–coded Merlin–Farber Hough Transform

概要

論文の詳細を見る
We propose a novel approach for detection ofthe facial midline from a frontal face image. The facial mid-line has several applications, for instance reducing computa-tional cost required for facial feature extraction (FFE) and post-operative assessment for cosmetic or dental surgery. The pro-posed method detects the facial midline of a frontal face from anedge image as the symmetry axis using the Merlin–Faber Houghtransformation. Experimental results on the FERET databaseindicate that the proposed algorithm can accurately detect mid-lines of face images with different scales and rotation.