Automatic 3D CAD models reconstruction from 2D orthographic drawings

arXiv 2023 35.8 method

TLDR

A two-stage method reconstructs 3D CAD models from 2D orthographic drawings using pattern-matching and loop detection, achieving 99.59% F-score.

Reasoning

The paper presents a clear, well-validated approach with strong quantitative results on a public dataset. However, it lacks discussion of limitations or comparison to modern learning-based methods, and the abstract does not detail the pattern-matching algorithm's robustness to complex drawings.

Read-first score

Read-first score 35.8, weighted from topical fit, citation, graph, method, reproducibility, and recency signals. Original total remains 42.

Methodology quality 25%

Screens visible abstract and analysis fields for experiment, dataset, baseline, metric, and limitation evidence. markers=dataset,evaluation,result

Recency 8%

65.1

Uses a gentle age decay so recent papers surface without erasing older foundations. 2023

Topical relevance 42%

26.2

Uses existing LLM keyword relevance scores normalized to 0-100. AI for CAD,computer-aided design,neural CAD,generative CAD,parametric CAD,B-Rep,boundary representation,constructive solid geometry,CSG,sketch extrusion,CAD generation,CAD reconstruction,text-to-CAD,image-to-CAD,point cloud to CAD,CAD program

Reproducibility 25%

Screens links and visible text for paper, code, dataset, artifact, and repository signals. pdf=False; code=False; dataset=False; markers=dataset

Field roles

Methodology anchor

Rank sensitivity

Stability: volatile; rank range: 79.

Keyword Scores

CAD reconstruction

computer-aided design

image-to-CAD

B-Rep

boundary representation

AI for CAD

CAD generation

generative CAD

parametric CAD

CAD program

neural CAD

constructive solid geometry

CSG

sketch extrusion

text-to-CAD

point cloud to CAD

Deep Analysis

Innovations

Pattern-matching algorithm for reconstructing a network of 3D edges from 2D edge features in multiple orthographic views
Loop detection and clustering algorithm to identify faces from the 3D wireframe graph
Two-stage pipeline that trims and stitches faces to produce a watertight 3D CAD model

Methodology

The method uses a two-stage process: first, a pattern-matching algorithm reconstructs 3D edges by matching 2D edge features from orthographic views; second, a loop detection algorithm identifies possible face loops from the 3D wireframe, a clustering algorithm recognizes faces, and faces are trimmed and stitched to form a watertight model. Evaluation was performed on a public dataset of several thousand 3D objects using F-score.

Key Results

The approach achieved 99.59% well-reconstructed models in F-score on a public dataset of several thousand 3D objects.