CONTROL OF SEAT BELTS OF VEHICLE DRIVERS WHILE DRIVING WITH AN UNMANNED AERIAL VEHICLE WITH ARTIFICIAL INTELLIGENCE

Today, with the rapid development of technology, the areas of use of artificial intelligence technologies are also rapidly increasing. Artificial intelligence applications are frequently used in many fields such as education, engineering and health. One of the important areas of use of artificial intelligence systems is mechatronic engineering. Artificial intelligence methods are frequently used especially in robotics and unmanned aerial vehicle applications. In the study, an artificial intelligence model developed to detect seat belt use by drivers using unmanned aerial vehicles is introduced. Seat belts play an important role in reducing injuries and deaths in traffic accidents, but current examination methods are time-consuming and limited. In this study, image processing techniques were used to determine whether drivers are wearing seat belts. For this purpose, a dataset consisting of in-car images taken under different driving conditions was created and Gaussian filters were applied to these images to remove noise and interference. Convolutional neural network architecture was used for model training and the results were compared with common models such as ResNet-18 and AlexNet. The model developed as a result of training has an accuracy rate of 94.55%. Test results showed that the developed special convolutional neural network model is superior to other models in terms of accuracy and performance. The study revealed that artificial intelligence and image processing techniques can increase traffic safety by monitoring seat belt use more effectively.