Face recognition has been the focus of computer vision researchers for many years. Variations in scale, position, illumination, orientation and facial expression make it difficult to distinguish the intrinsic differences between two different faces while ignoring differences caused by the environment. In this thesis, a face recognition system under varying pose is proposed. As a result of this resaerch, a number of face recognition algorithm were developed. These include: First, we use a HSI color information of face for facial feature finder. Segmented regions corresponding to the eyes and mouth in the face images usually exhibit regular properties in a HSI color space. So, the new facial feature finding algorithm is introduced that is much faster and more efficient than the other``s methods. Second, the template-based recognizer proposed by Brunelli is applied to recognize frontal face and show almostly the same result as his. Since the template based method is succesful on frontal or nearly frontal view, if we can generate frontal face from varying pose, the recognition performance of varying pose system will be considerable. From this motivation, the second proposed system is to build a face recognizer that works under varying pose.