Model Synthesis: A General Procedural Modeling Algorithm.

Area of Science:

• Computer Graphics
• Computational Geometry
• Artificial Intelligence

Background:

• Procedural modeling is essential for generating complex 3D assets.
• Existing methods often lack flexibility or require significant user input.
• Automating the creation of detailed 3D models for diverse applications remains a challenge.

Purpose of the Study:

• To present a novel method for procedurally modeling general complex 3D shapes.
• To enable automatic generation of intricate models, including buildings and urban datasets.
• To ensure generated models satisfy user-defined dimensional, geometric, and algebraic constraints.

Main Methods:

• Developing an algorithm for automatic 3D model generation based on user inputs and constraints.
• Implementing techniques to handle complex shapes and optimize performance.
• Comparing the proposed model synthesis approach with existing procedural modeling techniques.

Main Results:

• The method successfully generates complex 3D models of buildings, man-made structures, and urban datasets.
• Models generated exhibit natural appearances by satisfying user-defined constraints.
• Efficient techniques were developed for handling complex shapes, demonstrating strong performance across various model types.

Conclusions:

• The presented procedural modeling method offers an efficient and flexible approach to generating complex 3D shapes.
• User-defined constraints effectively guide the generation process, capturing user intent for naturalistic results.
• The approach shows a strong connection to context-sensitive grammars, offering potential for further theoretical development.