Research about Ocular Region Automatic Outline Fit
Hong Bao, Ming Li · 2015
In recent years, with the development of artificial intelligence, more experts and scholars have joined in the research about facial expression recognition.Algorithms of feature extraction and facial expression recognition are put forward in succession, and each has advantages and disadvantages.A research about feature extraction method and facial expression recognition recently put forward which is on the basis of Bezier curve found that representation of expression by geometrical characteristics of Bezier curve control points has the advantages of small characteristic dimension and rapid arithmetic speed, but the point selection needs to be marked manually, which can't meet the requirements of real time recognition of facial expression and can't ensure the consistency of several-time marks.Ocular region automatic outline fitting and features extraction method was put forward under such conditions, extracting convex hull area of eye outline by image processing techniques and Bezier curve fitting and control points obtaining by interpolation methods.Japanese female facial expression database (JAFFE).All images after several tests could gain favorable fitting effect and the data extracted was kept in consistency.