Automatic guided hospital vehicle with intelligent visualsurveillance system

Automated guided vehicles (AGV); It is used in hospitals for purposes such as meeting patient needs faster, increasing patient care and efficiency and creating smart hospitals. In practice, it is ensured that AGV's travel without harming the people and objects in the hospital environment. Behavioral information of the people around is reported to the center in real-time with artificial intelligence and image processing methods. Possible problem; It is ensured that AGV can act according to situations where many people run at the same time, everyone falls and fights. In determining the behavior, a deep learning method trained according to the mathematical method and data obtained from the star skeleton models of volunteer subjects are used. All studies were simulated in the robot operating system (ROS) after mapping the experimental environment determined first. It was then tested with an experimental OGV in a real environment. As a result of the studies, it is seen that the deep learning method gives better results in behavior detection than the mathematical method. In addition, the thesis study has shown that it can be used in visual surveillance systems by offering a new field proposal for AGV's carrying materials in a hospital setting. Visual surveillance gains importance in environments where people are in groups such as the hospital environment. In this sense, hybrid AGV's are business facilitators for operators working in visual surveillance systems. Keywords: Robot operating system, deep learning, auto guided vehicle, behavior detection