Large-Scale Dense 3D Mapping Using Submaps Derived From Orthogonal Imaging Sonars

John McConnell,Ivana Collado-Gonzalez,Paul Szenher,Brendan Englot

2024-12-05

Abstract:3D situational awareness is critical for any autonomous system. However, when operating underwater, environmental conditions often dictate the use of acoustic sensors. These acoustic sensors are plagued by high noise and a lack of 3D information in sonar imagery, motivating the use of an orthogonal pair of imaging sonars to recover 3D perceptual data. Thus far, mapping systems in this area only use a subset of the available data at discrete timesteps and rely on object-level prior information in the environment to develop high-coverage 3D maps. Moreover, simple repeating objects must be present to build high-coverage maps. In this work, we propose a submap-based mapping system integrated with a simultaneous localization and mapping (SLAM) system to produce dense, 3D maps of complex unknown environments with varying densities of simple repeating objects. We compare this submapping approach to our previous works in this area, analyzing simple and highly complex environments, such as submerged aircraft. We analyze the tradeoffs between a submapping-based approach and our previous work leveraging simple repeating objects. We show where each method is well-motivated and where they fall short. Importantly, our proposed use of submapping achieves an advance in underwater situational awareness with wide aperture multi-beam imaging sonar, moving toward generalized large-scale dense 3D mapping capability for fully unknown complex environments.

Robotics

What problem does this paper attempt to address?

### What problems does this paper attempt to solve?

This paper aims to address the challenges of large - scale dense 3D mapping in underwater environments, especially for complex and unknown environments. Specifically, the authors attempt to recover 3D perception data by using a pair of orthogonal multi - beam imaging sonars and propose a submap - based mapping system to generate dense 3D maps.

#### Main problems and challenges:
1. **Lack of 3D information**: There is a lack of 3D information in a single sonar image, especially in cases of high noise and low resolution.
2. **Environmental condition limitations**: In underwater environments, optical sensors such as cameras and LiDAR are difficult to apply because they have high requirements for water quality, lighting, etc. Although sonar is highly adaptable, it has a low signal - to - noise ratio and low resolution.
3. **Dependence on simple and repetitive objects**: Existing mapping systems rely on the existence of simple and repetitive objects in the environment to construct high - coverage maps, which limits their application in complex environments.
4. **Large - scale dense 3D mapping**: How to achieve large - scale dense 3D mapping in completely unknown and complex environments, especially in the absence of simple and repetitive objects.

#### Solutions:
- **Orthogonal sonar pair**: Use a pair of orthogonal multi - beam imaging sonars to obtain more 3D information.
- **Submap method**: Introduce the submap method combined with the simultaneous localization and mapping (SLAM) system to generate dense 3D maps, which is suitable for complex and unknown environments.
- **No prior information required**: The new method does not depend on the existence of simple and repetitive objects in the environment, thereby improving the universality and applicability of the system.

Through these improvements, the method proposed in the paper can significantly enhance the situational awareness ability of underwater autonomous systems, especially in complex and unknown underwater environments.

### Formula summary

1. **Conversion from spherical coordinates to Cartesian coordinates**:
\[
\begin{pmatrix}
X \\
Y \\
Z
\end{pmatrix}
= R \begin{pmatrix}
\cos\phi \cos\theta \\
\cos\phi \sin\theta \\
\sin\phi
\end{pmatrix}
\]
where \(R\) is the distance, \(\theta\) is the azimuth angle, and \(\phi\) is the pitch angle.

2. **Fusion of measurement values**:
\[
z_{\text{Fused}} = \left( \frac{R_h + R_v}{2}, \theta(h), \phi(v) \right)^\top
\]
where \(z_h=(R_h, \theta, \gamma_h)^\top\) and \(z_v=(R_v, \phi, \gamma_v)^\top\) are the horizontal and vertical sonar observations respectively.

3. **Placing observations in a fixed frame**:
\[
M = \left\{ \hat{z}^{(I)} \mid \hat{z}^{(I)} = T \hat{z}_{\text{Fused}}, \forall \hat{z} \in b_z \right\}
\]

Through these formulas, the paper shows how to extract and fuse 3D information from sonar data and finally generate a dense 3D map.