Abstarct: In video surveillance, the vision angle of face with respect to camera, called angular occlusion (also referred as head pose) will limit system’s ability in face recognition. In this thesis, a method for angular occlusion elimination in face images is proposed, which is baxsed on image morphing. The proposed method models a head-on face from a batch of images with different head pose, which belongs to a person. By simple techniques such as spatial interpolation of input images pixels with translation functions, linear interpolation and averaging from pixels intensity, the proposed method be able to model frontal face. To improvement modeling result, the proposed method apply on divergent face images (face images with symmetry poses) and then by low rank decomposition, modeled faces are aligned with respect to each other. Radial basis function neural network is considered for translation function. No need for depth information, calibrated images and head pose data i.e. (yaw,pitch,roll) that usually used in 3D modelling are some advantages of proposed method. The algorithm’s performance on PRIMA dataset is investigated. Experiment results show that the proposed method, as the face-modelling approach using face images with different head poses, will model frontal face image with proper accuracy.