Submit to Sensors Review for Sensors Propose a Special Issue

Journal Browser

► Journal Browser

Machine Learning, Signal, and/or Image Processing Methods to Enhance Environmental Sensors

Special Issue Editors
Special Issue Information
Published Papers

A special issue of Sensors (ISSN 1424-8220). This special issue belongs to the section "Intelligent Sensors".

Deadline for manuscript submissions: closed (10 December 2023) | Viewed by 9292

Share This Special Issue

Special Issue Editors

Dr. Matthieu Puigt

E-Mail Website
Guest Editor

Laboratoire d'Informatique, Signal et Image de la Côte d'Opale (LISIC), Université du Littoral Côte d'Opale (ULCO), F-62228 Calais CEDEX, France
Interests: signal processing; machine learning; low-rank approximations; matrix factorization; source separation; source localization; sensor calibration

Dr. George Tzagkarakis

E-Mail Website
Guest Editor

Institute of Computer Science (ICS), Foundation for Research and Technology Hellas (FORTH), GR-70013 Heraklion, Crete, Greece
Interests: compressive sensing and sparse representations; non-Gaussian heavy-tailed models; distributed signal processing in sensor networks; multimodal machine learning

Dr. Gilles Delmaire

E-Mail
Guest Editor

Laboratoire d'Informatique, Signal et Image de la Côte d'Opale (LISIC), Université du Littoral Côte d'Opale (ULCO), F-62228 Calais CEDEX, France
Interests: deconvolution; source apportionment; hyperspectral imaging; robust methods in signal processing

Prof. Dr. Gilles Roussel

E-Mail
Guest Editor

Laboratoire d'Informatique, Signal et Image de la Côte d'Opale (LISIC), Université du Littoral Côte d'Opale (ULCO), F-62228 Calais CEDEX, France
Interests: physics-driven data Processing; data assimilation; optimal sensor placement; latent variable estimation; source separation; low-rank approximation, hyperspectral imaging, machine learning

Special Issue Information

Dear Colleagues,

Over the past decades, environmental sensors knew tremendous developments, e.g., for the sake of miniaturization, or to lower the energy consumption. Such sensors may provide 1-D (e.g., gas sensor readings) to n-D (e.g., time series of hyperspectral images) data and may be deployed in many configurations. They allowed breakthroughs in, e.g., air or water quality monitoring, high precision agriculture, bio-acoustics, remote sensing. However, they also provide some specific issues for which modern machine learning, signal or image processing techniques were proposed. These approaches are based on statistics (e.g., sparse or low-rank approximation, latent variable analysis, deconvolution, supervised or unsupervised learning, deep learning) or reinforcement learning for example. They can be deployed in a centralized or distributed way, working in real time or as a batch processing, etc. They allow to tackle issues such as pollutant source detection and localization, optimal sensor placement, in-situ sensor calibration, heterogeneous sensor processing, (hyperspectral) camera demosaicing/stitching/super-resolution, (nonlinear) source separation for electronic tongues or noses, (heterogeneous) sensor fusion, etc.

This special issue will focus on the latest advances in machine learning and signal & image processing techniques for environmental sensors. Prospective authors are invited to submit original high-quality manuscripts on topics including (but not limited to):

offline or online, centralized or distributed learning/processing of (streams of) environmental data;
sensor fault detection and compensation;
optimal sensor placement;
pollution source detection, localization, separation, or classification;
hyperspectral image demosaicing, unmixing, super-resolution, sharpening, clustering;
sensor co-design

Dr. Matthieu Puigt
Dr. George Tzagkarakis
Dr. Gilles Delmaire
Prof. Dr. Gilles Roussel
Guest Editors

Manuscript Submission Information

Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All submissions that pass pre-check are peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 100 words) can be sent to the Editorial Office for announcement on this website.

Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-blind peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. Sensors is an international peer-reviewed open access semimonthly journal published by MDPI.

Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 2600 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.

Published Papers (5 papers)

Download All Papers

Research

22 pages, 11456 KiB

Open AccessArticle

Kalman-Based Scene Flow Estimation for Point Cloud Densification and 3D Object Detection in Dynamic Scenes

by Junzhe Ding, Jin Zhang, Luqin Ye and Cheng Wu

Sensors 2024, 24(3), 916; https://doi.org/10.3390/s24030916 - 31 Jan 2024

Viewed by 775

Abstract

Point cloud densification is essential for understanding the 3D environment. It provides crucial structural and semantic information for downstream tasks such as 3D object detection and tracking. However, existing registration-based methods struggle with dynamic targets due to the incompleteness and deformation of point clouds. To address this challenge, we propose a Kalman-based scene flow estimation method for point cloud densification and 3D object detection in dynamic scenes. Our method effectively tackles the issue of localization errors in scene flow estimation and enhances the accuracy and precision of shape completion. Specifically, we introduce a Kalman filter to correct the dynamic target’s position while estimating long sequence scene flow. This approach helps eliminate the cumulative localization error during the scene flow estimation process. Extended experiments on the KITTI 3D tracking dataset demonstrate that our method significantly improves the performance of LiDAR-only detectors, achieving superior results compared to the baselines. Full article

(This article belongs to the Special Issue Machine Learning, Signal, and/or Image Processing Methods to Enhance Environmental Sensors)

► Show Figures

Figure 1

17 pages, 6263 KiB

Open AccessArticle

Application of Machine Learning for Calibrating Gas Sensors for Methane Emissions Monitoring

by Ballard Andrews, Aditi Chakrabarti, Mathieu Dauphin and Andrew Speck

Sensors 2023, 23(24), 9898; https://doi.org/10.3390/s23249898 - 18 Dec 2023

Viewed by 1176

Abstract

Methane leaks are a significant component of greenhouse gas emissions and a global problem for the oil and gas industry. Emissions occur from a wide variety of sites with no discernable patterns, requiring methodologies to frequently monitor these releases throughout the entire production chain. To cost-effectively monitor widely dispersed well pads, we developed a methane point instrument to be deployed at facilities and connected to a cloud-based interpretation platform that provides real-time continuous monitoring in all weather conditions. The methane sensor is calibrated with machine learning methods of Gaussian process regression and the results are compared with artificial neural networks. A machine learning approach incorporates environmental effects into the sensor response and achieves the accuracies required for methane emissions monitoring with a small number of parameters. The sensors achieve an accuracy of 1 part per million methane (ppm) and can detect leaks at rates of less than 0.6 kg/h. Full article

(This article belongs to the Special Issue Machine Learning, Signal, and/or Image Processing Methods to Enhance Environmental Sensors)

► Show Figures

Figure 1

19 pages, 3700 KiB

Open AccessArticle

Embedded Temporal Convolutional Networks for Essential Climate Variables Forecasting

by Maria Myrto Villia, Grigorios Tsagkatakis, Mahta Moghaddam and Panagiotis Tsakalides

Sensors 2022, 22(5), 1851; https://doi.org/10.3390/s22051851 - 26 Feb 2022

Viewed by 1805

Abstract

Forecasting the values of essential climate variables like land surface temperature and soil moisture can play a paramount role in understanding and predicting the impact of climate change. This work concerns the development of a deep learning model for analyzing and predicting spatial time series, considering both satellite derived and model-based data assimilation processes. To that end, we propose the Embedded Temporal Convolutional Network (E-TCN) architecture, which integrates three different networks, namely an encoder network, a temporal convolutional network, and a decoder network. The model accepts as input satellite or assimilation model derived values, such as land surface temperature and soil moisture, with monthly periodicity, going back more than fifteen years. We use our model and compare its results with the state-of-the-art model for spatiotemporal data, the ConvLSTM model. To quantify performance, we explore different cases of spatial resolution, spatial region extension, number of training examples and prediction windows, among others. The proposed approach achieves better performance in terms of prediction accuracy, while using a smaller number of parameters compared to the ConvLSTM model. Although we focus on two specific environmental variables, the method can be readily applied to other variables of interest. Full article

(This article belongs to the Special Issue Machine Learning, Signal, and/or Image Processing Methods to Enhance Environmental Sensors)

► Show Figures

Figure 1

20 pages, 506 KiB

Open AccessArticle

Optimizing the Energy Efficiency of Unreliable Memories for Quantized Kalman Filtering

by Jonathan Kern, Elsa Dupraz, Abdeldjalil Aïssa-El-Bey, Lav R. Varshney and François Leduc-Primeau

Sensors 2022, 22(3), 853; https://doi.org/10.3390/s22030853 - 23 Jan 2022

Cited by 1 | Viewed by 1992

Abstract

This paper presents a quantized Kalman filter implemented using unreliable memories. We consider that both the quantization and the unreliable memories introduce errors in the computations, and we develop an error propagation model that takes into account these two sources of errors. In addition to providing updated Kalman filter equations, the proposed error model accurately predicts the covariance of the estimation error and gives a relation between the performance of the filter and its energy consumption, depending on the noise level in the memories. Then, since memories are responsible for a large part of the energy consumption of embedded systems, optimization methods are introduced to minimize the memory energy consumption under the desired estimation performance of the filter. The first method computes the optimal energy levels allocated to each memory bank individually, and the second one optimizes the energy allocation per groups of memory banks. Simulations show a close match between the theoretical analysis and experimental results. Furthermore, they demonstrate an important reduction in energy consumption of more than 50%. Full article

(This article belongs to the Special Issue Machine Learning, Signal, and/or Image Processing Methods to Enhance Environmental Sensors)

► Show Figures

Figure 1

27 pages, 31560 KiB

Open AccessArticle

Semi-Automatic Spectral Image Stitching for a Compact Hybrid Linescan Hyperspectral Camera towards Near Field Remote Monitoring of Potato Crop Leaves

by Pierre Chatelain, Gilles Delmaire, Ahed Alboody, Matthieu Puigt and Gilles Roussel

Sensors 2021, 21(22), 7616; https://doi.org/10.3390/s21227616 - 16 Nov 2021

Cited by 2 | Viewed by 2060

Abstract

The miniaturization of hyperspectral cameras has opened a new path to capture spectral information. One such camera, called the hybrid linescan camera, requires accurate control of its movement. Contrary to classical linescan cameras, where one line is available for every band in one shot, the latter asks for multiple shots to fill a line with multiple bands. Unfortunately, the reconstruction is corrupted by a parallax effect, which affects each band differently. In this article, we propose a two-step procedure, which first reconstructs an approximate datacube in two different ways, and second, performs a corrective warping on each band based on a multiple homography framework. The second step combines different stitching methods to perform this reconstruction. A complete synthetic and experimental comparison is performed by using geometric indicators of reference points. It appears throughout the course of our experimentation that misalignment is significantly reduced but remains non-negligible at the potato leaf scale. Full article

(This article belongs to the Special Issue Machine Learning, Signal, and/or Image Processing Methods to Enhance Environmental Sensors)

► Show Figures

Figure 1

Journal Menu

Journal Browser

Machine Learning, Signal, and/or Image Processing Methods to Enhance Environmental Sensors

Share This Special Issue

Special Issue Editors

Special Issue Information

Published Papers (5 papers)

Research

Further Information

Guidelines

MDPI Initiatives

Follow MDPI