Nearly everyone is surrounded by synthesized virtual environments daily, often without realizing it. Commercial graphics in street posters, web banners, and television clips frequently contain artificial content and special effects designed to make messages memorable. Current rendering algorithms produce synthetic images that are nearly indistinguishable from real-world photographs, allowing high-quality renderings to be completed in reasonable time even on standard hardware. Many practical applications, such as film productions, utilize digital replicas of real-world locations to incorporate artificial elements, enable complex camera movements, and simulate large-scale effects, which would be costly or impractical without virtual scenes. However, a significant challenge in creating these environments is content generation, which is time-consuming and often requires many artists, leading to high costs and organizational complexity. This thesis focuses on generating and rendering cities with common features like buildings and bridges. It aims to develop supportive tools that leverage procedural modeling techniques with user-friendly interfaces for intuitive creative input. The work is divided into three parts: analyzing procedural modeling techniques, introducing interactive extensions for non-expert users, and presenting an efficient rendering pipeline that generates high-quality city models without the need for extensive data
