Abstarct: Video games today have a special place among entertainments. Along with the benefits listed for video games, overuse can also cause a variety of problems. In the meantime, body-controlled games can be a good alternative to other video games because they stimulate people. In this thesis, we intend to develop a video game for Android devices so that we can control the personality of the game by identifying the user's hand gestures and recognizing his face. One of the problems with this, was the lower processing power of mobile phones than computers. For this reason, we have used real-time methods to identify hand gestures and face detection. In this thesis we used cascading classifiers with extraction of Haar-like features and local binary pattern and various ideas were used to apply them to the most ideal case. By recognizing their faces, hands, and direction, we control the game's character. Processing algorithms are applied to the images captured from the front camera of the mobile phone and we use the OpenCVForUnity library. Unity software is used to design the video game and the programming language is C #. At the end, we developed the game for both Windows and mobile operating systems, and the accuracy for both operating systems was 86.4 percent and in the Windows and Android versions we achieved frxame per second rate 11 and 8, respectively.