Advancements in technology today have enabled the emergence of new systems aimed at increasing worker safety in the field of occupational health and safety. However, even with all these safety measures in place, it must not be forgotten that a worker's fatigue plays a critical role in safety. A tired worker may struggle to comply with safety protocols, regardless of how many are in place. Therefore, especially in industrial tasks that require attention and diligence, determining whether workers are fatigued is considered of vital importance. The study focused on detecting fatigue to ensure workers can sustain their work healthily in the long and short term. In this context, real-time video footage was utilized, and facial detection was performed using image processing techniques, mapping specific reference points on the face. Eye and mouth openness, along with the level of head tilt, were identified as signs of fatigue, and these parameters were evaluated against threshold values. Additionally, during fatigue detection, audio and visual questions related to occupational health and safety were asked to measure the worker's attention and knowledge. Thus, it was attempted to determine how prepared the worker was for certain work activities and equipment. Furthermore, individual fatigue reports were prepared using facial recognition in the proposed system. As a result of experimental studies, the performance of the proposed system on the subjective data set was calculated as accuracy 80%, precision 85%, recall 73% and F1 score 75%. On the YawDD dataset, accuracy was determined as 95.99%, precision as 96.83%, recall as 95.58% and F1 score as 95.59%s.