Abstract :- Methods for segmenting face and for detecting facial landmark is presented. A face recognition which is automatic and the input is just the information of the depth of the face is the aim of development. The whole process of segmentation includes edge detection, region clustering, shape analysis is also included to extract the face region. The addition of detection of a surface curvature information and depth relief curve greatly help in finding the nose and eye land marks to evaluate the proposed system and its edges process for 3D recognition uses. The accuracy of the segmentation and landmark detection are analyzed. It has a better edge over the state of the arc works. The work of segmentation is a 3D recognition system where evaluated and to improvements obtained by application of landmark based technics where analyzed.