Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

MiCADangelo: Fine-Grained Reconstruction of Constrained CAD Models from 3D Scans

Ahmet Serdar Karadeniz, Dimitrios Mallis, Danila Rukhovich, Kseniya Cherenkova, Anis Kacem, Djamila Aouada

TL;DR

MiCADangelo introduces a human-inspired CAD reverse engineering pipeline that converts 3D scans into fully parametric CAD models by leveraging multi-plane cross-sections to identify key sketch planes, predicting constrained 2D sketches, and applying differentiable extrusion optimization. The method explicitly models CAD sketch constraints and integrates them into the reconstruction process, enabling editable design history and higher fidelity geometry than prior approaches. Across public benchmarks (DeepCAD and Fusion360), MiCADangelo achieves state-of-the-art accuracy on both 3D reconstruction metrics and 2D sketch fidelity, while demonstrating robustness to real-world scan artifacts. The work highlights the practical value of incorporating sketch constraints and cross-section-based reasoning for scalable, editable reverse engineering of complex CAD parts, with clear avenues for extending beyond extrusion-based operations.

Abstract

Computer-Aided Design (CAD) plays a foundational role in modern manufacturing and product development, often requiring designers to modify or build upon existing models. Converting 3D scans into parametric CAD representations--a process known as CAD reverse engineering--remains a significant challenge due to the high precision and structural complexity of CAD models. Existing deep learning-based approaches typically fall into two categories: bottom-up, geometry-driven methods, which often fail to produce fully parametric outputs, and top-down strategies, which tend to overlook fine-grained geometric details. Moreover, current methods neglect an essential aspect of CAD modeling: sketch-level constraints. In this work, we introduce a novel approach to CAD reverse engineering inspired by how human designers manually perform the task. Our method leverages multi-plane cross-sections to extract 2D patterns and capture fine parametric details more effectively. It enables the reconstruction of detailed and editable CAD models, outperforming state-of-the-art methods and, for the first time, incorporating sketch constraints directly into the reconstruction process.

MiCADangelo: Fine-Grained Reconstruction of Constrained CAD Models from 3D Scans

TL;DR

MiCADangelo introduces a human-inspired CAD reverse engineering pipeline that converts 3D scans into fully parametric CAD models by leveraging multi-plane cross-sections to identify key sketch planes, predicting constrained 2D sketches, and applying differentiable extrusion optimization. The method explicitly models CAD sketch constraints and integrates them into the reconstruction process, enabling editable design history and higher fidelity geometry than prior approaches. Across public benchmarks (DeepCAD and Fusion360), MiCADangelo achieves state-of-the-art accuracy on both 3D reconstruction metrics and 2D sketch fidelity, while demonstrating robustness to real-world scan artifacts. The work highlights the practical value of incorporating sketch constraints and cross-section-based reasoning for scalable, editable reverse engineering of complex CAD parts, with clear avenues for extending beyond extrusion-based operations.

Abstract

Computer-Aided Design (CAD) plays a foundational role in modern manufacturing and product development, often requiring designers to modify or build upon existing models. Converting 3D scans into parametric CAD representations--a process known as CAD reverse engineering--remains a significant challenge due to the high precision and structural complexity of CAD models. Existing deep learning-based approaches typically fall into two categories: bottom-up, geometry-driven methods, which often fail to produce fully parametric outputs, and top-down strategies, which tend to overlook fine-grained geometric details. Moreover, current methods neglect an essential aspect of CAD modeling: sketch-level constraints. In this work, we introduce a novel approach to CAD reverse engineering inspired by how human designers manually perform the task. Our method leverages multi-plane cross-sections to extract 2D patterns and capture fine parametric details more effectively. It enables the reconstruction of detailed and editable CAD models, outperforming state-of-the-art methods and, for the first time, incorporating sketch constraints directly into the reconstruction process.

Paper Structure

This paper contains 30 sections, 20 equations, 18 figures, 12 tables, 3 algorithms.

Table of Contents

  1. Introduction
  2. Related Work
  3. Preliminaries and Problem Statement
  4. Proposed Method
  5. Method Overview
  6. Sketch Plane Detection Network
  7. Constrained Sketch Parameterization Network
  8. Differentiable Extrusion Optimization
  9. Extrusion Length Optimization.
  10. Experiments
  11. Datasets.
  12. Metrics.
  13. Implementation Details.
  14. Comparison with state-of-the-art
  15. Impact of Constraints in 3D Reverse Engineering
  16. ...and 15 more sections

Figures (18)

  • Figure 1: MiCADangelo is a novel framework for CAD reverse-engineering that mimics human design workflows. It analyzes 3D scans via 2D cross-sections to detect sketch planes, predict constrained parametric sketches and optimize extrusions.
  • Figure 2: Overview of the method. MiCADangelo comprises three main components: Sketch Plane Detection, Sketch Parameterization, and Differentiable Extrusion. The generated constrained sketches, together with the optimized extrusion parameters, are assembled into the final parametric CAD model.
  • Figure 3: Qualitative comparison of of our method and that of khan2024cad on DeepCAD and Fusion360.
  • Figure 4: Qualitative comparison between our method and that of khan2024cad on complex models containing fine-grained geometric details.
  • Figure 4: Effect of contextual embeddings on plane detection performance.
  • ...and 13 more figures

Theorems & Definitions (9)

  • Definition 1: Sketch Primitive
  • Definition 2: Sketch Constraint
  • Definition 3: Constrained Sketch
  • Definition 4: Sketch Plane
  • Definition 5: 3D Mesh
  • Definition 6: Cross-section Slice
  • Definition 7: Closed Loop
  • Definition 8: Extrusion
  • Definition 9: CAD Model Representation