In this paper, we would like to present the development of intelligent humanoid robot system. Our previous research have designed humanoid robot which could receive command and do speech recognition based on Bioloid GP robot and Raspberry Pi 2 as controller. In here, the robot system is extended to be able to detect and recognize human face. Based on OpenCV and Python, Viola-Jones method is used for face detection and PCA (Principal Component Analysis) for face recognition. Viola-Jones method provides a fast and robust framework for extracting and detecting face features. While, PCA or frequently called as eigenfaces is one of the algorithms to recognize face by reducing the image dimension using its eigen values. Generally, face recognition is influenced by many variables like light intensity, poses, and expression. By controlling these variables, the result of face recognition accuracy is about 93% (28 correct recognition of 30 experiments). Based on the recognized face, our humanoid robot gives the response action.