Skip to Main content Skip to Navigation

Monocular Visual-Inertial-Pressure Fusion for Underwater Localization and 3D Mapping.

Abstract : This thesis addresses the problem of real-time 3D localization and mapping in underwater environments. In the underwater archaeology field, Remotely Operated Vehicles (ROVs) are used to conduct deep-sea surveys and excavations. Providing both accurate localization and mapping information in real-time is crucial for manual or automated operation of the robots. While many localization solutions already exist for underwater robots, most of them rely on very accurate sensors, such as Doppler velocity logs or fiber optic gyroscopes, which are very expensive and may be too bulky for small ROVs. Acoustic positioning systems are also commonly used for underwater positioning, but they provide low frequency measurements, with limited accuracy. In this thesis, we study the use of low-cost sensors for accurate underwater localization. Our study investigates the use of a monocular camera, a pressure sensor and a low-cost MEMS-IMU as the only means of performing localization and mapping in the context of underwater archaeology. We have conducted an evaluation of different features tracking methods on images affected by typical disturbances met in an underwater context. From the results obtained with this evaluation, we have developed a monocular Visual SLAM (Simultaneous Localization and Mapping) method, robust to the specific disturbances of underwater environments. Then, we propose an extension of this method to tightly integrate the measurements of a pressure sensor and an IMU in the SLAM algorithm. The final method provides a very accurate localization and runs in real-time. In addition, an online dense 3D reconstruction module, compliant with a monocular setup, is also proposed. Two lightweight and compact prototypes of this system have been designed and used to record datasets that have been publicly released. Furthermore, these prototypes have been successfully used to test and validate the proposed localization and mapping algorithms in real-case scenarios.
Complete list of metadatas

Cited literature [245 references]  Display  Hide  Download
Contributor : Maxime Ferrera <>
Submitted on : Wednesday, March 4, 2020 - 5:27:41 PM
Last modification on : Friday, June 26, 2020 - 2:30:08 PM
Long-term archiving on: : Friday, June 5, 2020 - 3:26:20 PM


Files produced by the author(s)


  • HAL Id : tel-02462079, version 2


Maxime Ferrera. Monocular Visual-Inertial-Pressure Fusion for Underwater Localization and 3D Mapping.. Robotics [cs.RO]. Université de Montpellier, 2019. English. ⟨tel-02462079v2⟩



Record views


Files downloads