This study investigates the performance of LiDAR sensors and camera-based systems in foggy indoor environments to ensure robust navigation for Autonomous Mobile Robots (AMR). AMR requires accurate self-localization and environmental perception, typically achieved through Simultaneous Localization and Mapping (SLAM). In indoor environments, LiDAR and cameras are commonly used, but their performance can be significantly affected by environmental factors such as fog. While LiDAR behavior in foggy conditions has been widely studied, research on camera-based recognition, especially using AR markers, remains limited. In this work, experiments were conducted in a controlled indoor fog chamber to evaluate the sensing capabilities of both LiDAR and camera systems under different visibility conditions. The results show that when visibility falls below a certain threshold, LiDAR point clouds become arc-shaped and fail to capture environmental structures. A method was proposed to model the arc shape as a function of time and assess its standard deviation, determining a threshold below which LiDAR becomes unreliable. Conversely, camera-based systems demonstrated improved robustness in foggy environments. By analyzing a black and white board, the extinction coefficient was derived from contrast and distance, enabling rough visibility estimation. Furthermore, AR marker recognition was enhanced by contrast adjustment, improving performance under low-visibility conditions.