IEEE Sensors 2021

Two Early Stage Researchers (ESR) of the project MENELAOS^NT represented ZESS at the IEEE SENSORS 2021 conference.

MENELAOS^NT is an EU H2020 Marie Skłodowska-Curie Innovative Training Network (ITN) coordinated by the University of Siegen, hosting 15 ESRs, and spanning over top research institutes in five European countries. The presence of two ESRs from Siegen at IEEE SENSORS 2021 further consolidates the privileged position of ZESS as a world-class player in the sensors’ community.

 

Faisal Ahmed (left) and Alvaro Lopez Paredes (right) will represent ZESS at IEEE SENSORS 2021.

The first research outputs of an ambitious project

Faisal Ahmed and Alvaro Lopez Paredes started their PhD studies at the Center of Sensor Systems (ZESS) of the University of Siegen in November 2020. They were selected within the Marie Sklodowska-Curie Innovative Training Network Project MENELAOSNT, an ambitious project which involves several universities, research institutions, and industrial partners from Romania, Spain, Germany, Switzerland and Turkey, and aims at applying Novel Technologies to realize multimodal-multi sensor data fusion to optimally combine the information delivered by different sensors (in-situ/remote, optical/non optical) on different scales, with different resolutions, and with different reliability.

Over the past few months, both were focused on the study of Time-of-Flight (ToF) systems and Compressive Sensing (CS) techniques from different perspectives. ToF systems allow sensing the world in 3D, while CS techniques allow for capturing the relevant information with minimal data flow, that is, in an efficient manner. As a result of their respective researches, they will take part in the IEEE SENSORS 2021 conference, the flagship conference of the IEEE SENSORS Council, to be held virtually in Sydney, Australia between 29th of October and 4th of November 2021. This is an important milestone for both of them in their respective research careers and represents not only a step forwards in their PhD studies but a recognition of the valuable work performed in the scientific context.

From lamps to sources of internet and 3D images

Faisal Ahmed received his MEng in Telecommunication Engineering from the Mehran University of Engineering and Technology (MUET), Pakistan in 2019. In 2020, he joined the ZESS as ESR-3 under the umbrella of Marie Skłodowska-Curie Innovative Training Network Project MENELAOSNT. Currently, he is working on "Pseudo-Passive Indoor 3D ToF sensing exploiting Visible Light Communication infrastructure". This is a novel ToF sensing scheme which replaces the active illumination unit present in ToF cameras with the inherent modulation present in Visible Light Communication (VLC) infrastructure, turning background indoor light into a useful optical signal for the ToF camera for thefirst time.
This is an interesting and captivating topic that enables multiple functionalities to be used simultaneously. ToF 3D-sensing has become a hot subject of research over the past few years, and this trend will be further accelerated by the application of VLC, which bundles together emerging services such as intelligent lightning, light-based communications, and 3D imaging.

 

The research project of Faisal Ahmed proposes using the VLC infrastructure as conceptual evidence for a passive ToF sensing configuration.

 

State-of-the-art ToF cameras suffer from high power consumption and may rise safety concerns due to the dedicated high-power illumination units. Faisal Ahmed's research focuses on the investigation of a novel passive 3D modality which will eventually relegate to oblivion the illumination units in existing ToF systems. This ambitious task has become feasible thanks to rapid advances in semiconductor technologies that have transformed lighting infrastructure. As a result, illumination devices are currently migrating from conventional lamps to light-emitting-diodes (LEDs). LEDs bring smarter capabilities into the lighting infrastructure and enable one of the most prominent emerging communication technologies for the indoor environments, known as VLC.
Within this framework, the lighting infrastructure does not only satisfy illumination purposes, but also provides high-speed communication. This yields the pervasive presence of modulated light signals in the indoor infrastructure. The research work of Faisal Ahmed takes this a step further by tapping VLC infrastructure as illumination source for Passive 3D ToF sensing.
This novel idea has a great potential in the field of 3D-sensing and, thus, is expected to awake industrial interest in the near future. The significant reduction of cost and power consumption allows the entry into new application fields, such as smart homes, office environments, industries and vehicles, where the VLC infrastructure and ToF cameras are valuable assets.
In his paper, Faisal Ahmed studies the feasibility of this new concept by using an off-the-shelf VLC module which follows the four-phases algorithm, a standard tool for phase retrieval in Continuous-Wave (CW) ToF systems. CW-ToF cameras make use of a light source which illuminates the scene in connection with an array of ToF pixels. This allows to recover the phase shifts from the backscattered optical signals that return from the scene. Hence, depth information can readily be obtained. Moreover, the VLC module provides two dominant frequencies arising from the underlying clock signal and coding scheme. Consequently, he adopts the CW-ToF pipeline at both dominant frequencies to carry out his simulations. The simulation results show that accurate depth estimation is attainable both, in short and long ranges. For further read, see:

F. Ahmed, M. Heredia Conde, and O. Loffeld, "Pseudo-Passive Indoor ToF Sensing exploiting Visible Light Communication Sources", 2021 IEEE SENSORS (to appear in)

3D-mapping the world in an effective manner

Alvaro Lopez Paredes received his MEng in Aerospace Engineering by the Technical University of Madrid (UPM) in 2012. In addition, he completed a Master's Degree in Numerical Simulation majoring in Solid Mechanics in 2016, and a Specialist's Degree in Numerical Simulation majoring in Computational Fluid Dynamics in 2020, both by the UPM in collaboration with ANSYS. He is the ESR-4 in the Marie Skłodowska-Curie Innovative Training Network Project MENELAOSNT and he will present the research paper titled "Effective very-wide-area ToF sensing" at IEEE SENSORS 2021. This study discusses the reconstruction of high-angular resolution and long-range 3D scenarios using 3D ToF imaging systems by making use of Compressive Sensing (CS) and sparsity-aware techniques.
The objects that surround us-vehicles, houses, trees, pedestrians-may be seen as discrete, sparse, points in the space and can be expressed by a few coefficients if they are represented in the appropriate basis, as a sinusoidal function can be represented by its frequency and amplitude coefficients in the Fourier domain. This property is called sparsity and is a fundamental concept in the 3D imaging theory. However, these discrete points can be located anywhere in the space, as the frequency can take any arbitrary positive value. This leads to the second and third fundamental concepts for recovering sparse signals: randomness and incoherence. The sensing scheme must efficiently scatter the spatial domain in order to capture these unknown locations. The randomness during the sensing processes avoids any possible bias on the scattering of the scene. The incoherence is a measurement of the degree of dissimilarity between the columns of the sensing matrices. This dissimilarity is good as the measurements arising from targets at different locations can be seen and explained in a unique way by the corresponding column of the matrix and this significantly facilitates the recovery of the signals by using greedy algorithms.

 

 

The research project of Álvaro López Paredes investigates novel ways of acquiring data with a ToF camera for efficiently sensing wide scenes exploiting its sparsity in 3D space.

The research focuses on pulsed ToF systems, which measure samples of the correlation between time-delayed pulses representing the response of the scene under study, and a number of predefined sequences. The assembly of these predefined sequences yields sensing matrices and the retrieval of the depth and intensity values may be seen as an under-determined optimization problem, such as the recovery of a sparse signal from few measurements. Specifically, his paper proposes a methodology capable of reconstructing an array of highly sparse signals based on a preliminary slicing of the spatial domain into several partitions and a posterior localization of the signals within them. This permits to neglect large empty areas and lower the size of the problem with the consequent decrease of retrieval errors and processing times. Then, the minimization problem is solved by applying a greedy algorithm. The main objective of this study is to develop a robust sensing scheme capable of working in nearly real time at a relatively low computational cost or, in other words, a 3D imaging system capable of seeing as far, wide, and fast as possible, taking advantage of the sparse nature of the signals under study in pulsed ToF imaging systems. The research investigates these theoretical concepts, studies the feasibility, and demonstrates the applicability of the sensing approach by numerical simulations on synthetic datasets with the ultimate scope of implementing such a scheme on simple illumination systems and state-of-the-art ToF sensors. On the hardware front, a rotary ToF system is proposed along with a drastic reduction of the exposure times in order to increase the lateral resolution and avoid the introduction of motion artifacts into the view. This paper represents the first step to build a fully-operational ToF camera prototype at the ZESS facilities, a challenging and breath-taking objective to be fulfilled over the next two years. For further read, see:

A. Lopez Paredes, M. Heredia Conde, and O. Loffeld. "Effective Very-wide-area 3D ToF Imaging" IEEE SENSORS 2021 (to appear in)

.-.-.-