For the low accuracy of face detection and tracking system on most Android platforms at this stage, an unsupervised face target detection and tracking system based on Android platform is proposed in this paper. Adopting the method based on particle filtering and background subtraction, this system can automatically detect and track the moving face targets in a video surveillance sequence without prior knowledge. VC++ and Open CV are used to carry out the specific implementation of the method, the establishment of development environment is introduced, and the method is transplanted to the Android platform successfully. The experimental results show that the face target can be detected and tracked by using this method successfully. Compared with other methods, the proposed method has higher realtime performance and robustness, and it has better tracking effects.