In this thesis, the Pan&Zoom system was developed, which enables controlling the display of content on the computer screen using hand gestures. The system is based on two modules: the MediaPipe framework is used to detect the key points of the palms and recognize hand gestures, and the Image-viewer manages the display of graphical content and the execution of pan and zoom actions. Two interaction modalities were designed and implemented for graphical content management: MH (Multi-hand zoom) and SH (Single-hand zoom). The MH modality allows panning with either hand, while both hands have to be used simultaneously for zooming. In contrast, in the SH modality, only the left hand is used for panning and the right hand for zooming. In order to compare the efficiency and interaction workload of the mentioned modalities, an HCI experiment was conducted with 30 test users that included corresponding tasks from three different complexity categories. Statistical analysis of the obtained results showed that the SH modality performs panning and zooming actions significantly more efficiently. It was also shown that the task's complexity significantly impacts the time of its execution. The above results were additionally supported by the analysis of standardized questionnaires, which showed that the MH modality leads to a significantly higher workload. The test users prefer the SH modality because it requires less physical effort, causes a lower level of frustration, and provides better efficiency.