The drone, also known as an unmanned aerial vehicle (UAV), is rapidly advanced in automation and autonomy with the use of computer vision technology. The design and deployment of a small, vision-based face tracking mini drone that is suitable for internal settings are presented in this study. The ESP32-CAM module, which acts as a camera and processing unit, is the central component of the system. It uses a Haar cascade classifier to detect faces in real time without the need for GPS chips or external processing. After processing the observed face coordinates, the respective flight command is generated with a positional offset (left, center, or right). F3 Evo flight controller receives these commands and uses brushless motors and ESC modules to modify the speed of the drone. The system is designed for stable indoor navigation and runs on a lightweight Li-Po battery. The suggested method does not require offboard processing, which reduces delay and system complexity, and it does not rely on GPS, making it perfect for enclosed areas where GPS signals are weak or unavailable. The accurate identification of a human face in real time and ability to monitor a human face in real time, as well as response with correct directional adjustments, is confirmed by experimental verification. This system is expandable for many applications, such as robotics, internal monitoring, and gesture-based interfaces, and its inexpensive and modular design. This study opens the door for more tests in vision-based navigation and indoor automation, showing that intelligent autonomous behavior in small drones can be successfully achieved using low-power embedded platforms.