Over the past decades, there has been an ever-increasing saturation of traffic networks due to the growing number of road vehicles, and due to the available limited. To solve these problems, adaptive, (semi-) intelligent traffic control has been used widely for the last decades. These systems nevertheless, have some shortages, the most obvious one being that these systems use the presence of vehicles at the lanes immediately before reaching the intersections. The real queue size cannot be taken into consideration. In the present approach, the input values are supposed to come from cameras connected with image processing systems and directed microphones. We propose a new traffic signal control system with a hierarchical structure based on similarly Mamdani control, however, containing essentially novel elements and having more intelligent features. This new model and the connected algorithmic approach allow rather complex control strategies, but only a simple case study has been implemented. Compared with existing fuzzy traffic controls, the novel approach has more adaptability and flexibility, by having the potential to differentiate an arbitrary number of traffic directions and by increasing general safety by the additional emergency vehicle handling feature. In addition, the calculation with queues, and individual vehicles weighted with the waiting time makes the system more flexible than any existing intelligent model.