3D reconstruction for SLAM using multisensor fusion and block-based inpainting

A. Zelensky; V. Voronin; N. Gapon; E. Semenishchev; V. Egipko; I. Khamidullin

doi:10.1117/12.2625905

17 May 2022 3D reconstruction for SLAM using multisensor fusion and block-based inpainting

A. Zelensky, V. Voronin, N. Gapon, E. Semenishchev, V. Egipko, I. Khamidullin

Proceedings Volume 12138, Optics, Photonics and Digital Technologies for Imaging Applications VII; 121380X (2022) https://doi.org/10.1117/12.2625905
Event: SPIE Photonics Europe, 2022, Strasbourg, France

Conference Poster

Abstract

Simultaneous localization and mapping (SLAM) systems are useful for camera tracking, and 3-D reconstructions may be desired for many robotic tasks. There is a problem consisting of a decrease in the accuracy of planning the movement trajectory caused by incorrect sections on the depth map due to incorrect distance determination to objects. Such defects appear as a result of poor lighting, specular or fine-grained surfaces of objects. As a result, the effect of increasing the boundaries of objects (obstacles) appears, and the overlapping of objects makes it impossible to distinguish one object from another. In this paper, we propose a multisensor SLAM system capable of recovering a globally consistent 3-D structure. The proposed method mainly takes two steps. The first step is to fusion images from visible cameras and depth sensors based on the PLIP model (parameterized model of logarithmic image processing) close to the human visual system's perception. The second step is image reconstruction. This article presents an approach based on a modified exemplar block-based algorithm using the autoencoder-learned local image descriptor for image inpainting. For this purpose, we learn the descriptors using a convolutional autoencoder network. Then, a 3-D point cloud is generated by using the reconstructed data. Our system outperforms the state-of-the-art methods quantitatively in reconstruction accuracy on a benchmark for evaluating RGB-D SLAM systems.

Citation Download Citation

A. Zelensky, V. Voronin, N. Gapon, E. Semenishchev, V. Egipko, and I. Khamidullin "3D reconstruction for SLAM using multisensor fusion and block-based inpainting", Proc. SPIE 12138, Optics, Photonics and Digital Technologies for Imaging Applications VII, 121380X (17 May 2022); https://doi.org/10.1117/12.2625905

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available