Scaled monocular SLAM for walking people

The urban environments represent challenging areas for handheld device pose estimation (i.e., 3D position and 3D orientation) in large displacements. It is even more challenging with low-cost sensors and computational resources that are available in pedestrian mobile devices (i.e., monocular camera and Inertial Measurement Unit). To address these challenges, we propose a continuous pose estimation based on monocular Visual Odometry. To solve the scale ambiguity and suppress the scale drift, an adaptive pedestrian step lengths estimation is used for the displacements on the horizontal plane. To complete the estimation, a handheld equipment height model, with respect to the Digital Terrain Model contained in Geographical Information Systems, is used for the displacement on the vertical axis. In addition, an accurate pose estimation based on the recognition of known objects is punctually used to correct the pose estimate and reset the monocular Visual Odometry. To validate the benefit of our framework, experimental data have been collected on a 0.7 km pedestrian path in an urban environment for various people. Thus, the proposed solution allows to achieve a positioning error of 1.6–7.5% of the walked distance, and confirms the benefit of the use of an adaptive step length compared to the use of a fixed-step length.

show abstract

“…For example, Ref. [ 33 , 34 ] have the sensors attached to a helmet. This constraint is not valid in the context of handheld AR.…”

Section: Related Workmentioning

confidence: 99%

Solving Monocular Visual Odometry Scale Factor with Adaptive Step Length Estimates for Pedestrians Using Handheld Devices

Antigny

Uchiyama

Servières

et al. 2019

Sensors

View full text Add to dashboard Cite

show abstract

“…Part of this work was presented as conference papers in [15] and [16]. Now in this paper all the previous conference results obtained by experimentation with a catadioptric camera are integrated and improved.…”

Section: Introductionmentioning

confidence: 99%

True scaled 6 DoF egocentric localisation with monocular wearable systems

Gutiérrez-Gómez

Guerrero

2016

Image and Vision Computing

Self Cite

View full text Add to dashboard Cite

“…Our proposal in this field is an algorithm for scale estimation and scale drift avoidance when monocular SLAM is performed with a wearable camera by capturing the walking speed of the user, from the step frequency. Associated publications: [10], [2], [6] • Dense RGB-D SLAM: Recent algorithms for odometry estimation with RGB-D sensors, estimate the camera motion by performing pixelwise minimisation of the photometric and/or the geometric error. However, in many cases important properties of the error model of the depth sensor are ignored, which could affect the performance of odometry computation.…”

mentioning

confidence: 99%

Contributions to Real-time Metric Localisation with Wearable Vision Systems

Gutiérrez¹

2016

ELCVIA

View full text Add to dashboard Cite

Under the rapid development of electronics and computer science in the last years, cameras have become omnipresent nowadays, to such extent that almost everybody is able to carry one at all times embedded into their cellular phone. What makes cameras specially appealing for us is their ability to quickly capture a lot of information of the environment encoded in one image or video, allowing us to immortalize special moments in our life or share reliable visual information of the environment with other persons. However, while the task of extracting the information from an image may by trivial for us, in the case of computers complex algorithms with a high computational burden are required to transform a raw image into useful information. In this sense, the same rapid development in computer science that allowed the widespread of cameras has enabled also the possibility of real-time application of previously practically infeasible algorithms.Among the current fields of research in the computer vision community, this thesis is specially concerned in metric localisation and mapping algorithms. These algorithms are a key component in many practical applications such as robot navigation, augmented reality or reconstructing 3D models of the environment.The goal of this thesis * is to delve into visual localisation and mapping from vision, paying special attention to conventional and unconventional cameras which can be easily worn or handled by a human (Fig. 1). In this thesis I contribute in the following aspects of the visual odometry and SLAM (Simultaneous Localisation and Mapping) pipeline:• Generalised Monocular SLAM: State of the art visual SLAM algorithms are usually designed for conventional cameras, which can be modelled by a simple pin-hole camera model but have a narrow field of view. Vision systems consisting in a combination of a camera and a conic-shaped mirror, catadioptric cameras, offer a larger field of view, but in turn require of the use of a more complex projection model, which generalises the projection of central projection systems. Our proposal in this field is the adaptation to catadioptric cameras of a real time monocular SLAM system initially designed for conventional cameras.

show abstract

Scaled monocular SLAM for walking people

Cited by 7 publications

References 14 publications

Solving Monocular Visual Odometry Scale Factor with Adaptive Step Length Estimates for Pedestrians Using Handheld Devices

Solving Monocular Visual Odometry Scale Factor with Adaptive Step Length Estimates for Pedestrians Using Handheld Devices

True scaled 6 DoF egocentric localisation with monocular wearable systems

Contributions to Real-time Metric Localisation with Wearable Vision Systems

Contact Info

Product

Resources

About