This article implements face tracking in the complicated environment, which combines the face detection with face tracking as well as image sampling preprocessing. The factors that affect face detection time and face tracking time are demonstrated and analyzed based on the collected data. The hardware architecture can make use of any type of camera to track with the target face and ensure the face being located in the middle of the visual field. The hardware configuration of the system is simple, low cost with high accuracy in terms of face tracking. In the future, we hope that tracking any person with any poses can come true. Therefore, it will be possible to take pictures without cameraman and capture great shot intelligently.