This paper presents an integrated approach to autonomous drone navigation and positional stability using a combination of ArUco markers and QR codes. The proposed system includes four key functionalities: self-location determination through the analysis of visual markers, navigation to subsequent positions using encoded instructions, development of a probabilistic error model for movement between locations, and the integration of AI-based object recognition capabilities. The system uses computer vision techniques for localization and stabilization, utilizing clustering methods to mitigate noise and improve accuracy. Furthermore, the probabilistic model evaluates the positional uncertainty when navigating routes with sparse marker distribution, while QR codes encode detailed waypoint information to facilitate autonomous decision-making. Additionally, GPT-4 Vision is utilized for object identification tasks to enhance environmental analysis capabilities. The findings highlight the potential of combining visual markers, mathematical models, and AI to achieve reliable and efficient drone navigation, though challenges such as environmental instability and real-time processing constraints remain.