# Directional TSDF: Modeling Surface Orientation for Coherent Meshes

## TL;DR

This paper introduces the directional TSDF, a new 3D reconstruction method that separately models opposite surfaces to improve mesh accuracy, especially for thin structures, outperforming existing TSDF algorithms.

## Contribution

The paper proposes the directional TSDF, a novel representation that enhances surface modeling by storing opposite surfaces separately and modifies the marching cubes algorithm for better mesh coherence.

## Key findings

- Outperforms state-of-the-art TSDF methods in mesh accuracy
- Improves reconstruction of thin structures
- Uses surface gradient-based ray casting for better data fusion

## Abstract

Real-time 3D reconstruction from RGB-D sensor data plays an important role in many robotic applications, such as object modeling and mapping. The popular method of fusing depth information into a truncated signed distance function (TSDF) and applying the marching cubes algorithm for mesh extraction has severe issues with thin structures: not only does it lead to loss of accuracy, but it can generate completely wrong surfaces. To address this, we propose the directional TSDF - a novel representation that stores opposite surfaces separate from each other. The marching cubes algorithm is modified accordingly to retrieve a coherent mesh representation. We further increase the accuracy by using surface gradient-based ray casting for fusing new measurements. We show that our method outperforms state-of-the-art TSDF reconstruction algorithms in mesh accuracy.