Architectural Co-LOD Generation

TL;DR

Co-LOD is a new approach specifically designed for effective LOD management in architectural modeling that employs shape co-analysis to standardize geometric structures across multiple buildings, facilitating the progressive and consistent generation of LODs.

Abstract

Managing the level-of-detail (LOD) in architectural models is crucial yet challenging, particularly for effective representation and visualization of buildings. Traditional approaches often fail to deliver controllable detail alongside semantic consistency, especially when dealing with noisy and inconsistent inputs. We address these limitations with \emph{Co-LOD}, a new approach specifically designed for effective LOD management in architectural modeling. Co-LOD employs shape co-analysis to standardize geometric structures across multiple buildings, facilitating the progressive and consistent generation of LODs. This method allows for precise detailing in both individual models and model collections, ensuring semantic integrity. Extensive experiments demonstrate that Co-LOD effectively applies accurate LOD across a variety of architectural inputs, consistently delivering superior detail and quality in LOD representations.

Figures (24)

• Figure 1: Overview of Co-LOD workflow, which contains two main modules. The first module involves detecting primary planes and aggregating them to form structural segments from raw inputs. The second module quantitatively measures the similarity of these segments and clusters them to define LOD layers, a process akin to joint structural analysis. Together, these modules enable Co-LOD to effectively perform co-analysis for controlled LOD generation with semantic consistency.
• Figure 2: Visualization of primary planes and Bbox with voxel samples.
• Figure 3: An instance of structural segments aggregation. Plane-groups A, B, C, D, and E are concentrated based on related voxel regions. According to the overlap in $Hitlist$, interrelated planes (blue, cyan, and brown lines in the upper right corner) is detected to aggregate structures (A, B), then a structural segment (corresponds to the pale purple area) is generated.
• Figure 4: Illustration of segment-based similarity. With the D2 descriptor-based measurement and principal axes-based eigenvalue analysis, different segments can be quantitative compared, which considers shape and scale differences at the same time.
• Figure 5: Structural segment-based similarity map by $Dis_{set}$ (left) and $Sim_{co}$ (right). The latter one significantly distinguishes the difference between segments.