Dense reconstruction is a methodological approach within computer vision and related fields focused on generating highly detailed three-dimensional models of environments or objects from sensory data. It distinguishes itself by estimating geometric structure for a large proportion of the input data, typically resulting in dense 3D representations. As a research concept, it investigates robust and efficient techniques for recovering high-resolution geometric structure from sources such as multiple images, depth sensors, or point clouds under various conditions. Key characteristics include the generation of dense 3D outputs (e.g., dense point clouds, meshes, volumetric grids) that capture fine-grained surface details, contrasting with sparse methods that recover only prominent features. Its significance lies in enabling applications requiring detailed spatial understanding, such as autonomous navigation, virtual/augmented reality content creation, digital fabrication, and precise metrology.