A5086753365- Remote-Sensing Image Classification
- Laser-induced spectroscopy and plasma
- Infrared Target Detection Methodologies
- Spectroscopy and Chemometric Analyses
- Laser Design and Applications
- Fault Detection and Control Systems
- Risk Perception and Management
- Laser Material Processing Techniques
- Advanced Measurement and Detection Methods
- Semiconductor Lasers and Optical Devices
- Laser-Matter Interactions and Applications
- Advanced Image and Video Retrieval Techniques
- Climate variability and models
- Target Tracking and Data Fusion in Sensor Networks
- Precipitation Measurement and Analysis
- Image Processing Techniques and Applications
- Optical Polarization and Ellipsometry
- Advanced Combustion Engine Technologies
- Advanced Optical Imaging Technologies
- Robotic Path Planning Algorithms
- Optical Coherence Tomography Applications
- Social Acceptance of Renewable Energy
- Optical measurement and interference techniques
- Optical and Acousto-Optic Technologies
- Human-Automation Interaction and Safety
University of Liverpool
2014-2021
Laboratoire de Mathématiques
2005-2020
Département Mathématiques et Informatique Appliquées
2005-2011
Université des Antilles
2005
Institut Fresnel
1999-2002
École Normale Supérieure - PSL
1998-2000
<para xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> This letter presents the experimental results obtained for an automatic predetection of small ships (about 5 <formula formulatype="inline"><tex Notation="TeX">$\times$</tex></formula> pixels) in high-resolution optical satellite images. Our images are panchromatic SPOT images, whose resolution is m per pixel. detection method based on Bayesian decision theory and does not need any preprocessing....
The advent of sophisticated robotics and AI technology makes sending humans into hazardous distant environments to carry out inspections increasingly avoidable. Being able send a robot, rather than human, nuclear facility or deep space is very appealing. However, building these robotic systems just the start we still need range verification validation tasks ensure that be deployed are as safe reliable possible. Based on our experience across three research innovation hubs within UK’s “Robots...
This paper reports on a technique demonstrating for the first time successful multi-point laser-induced spark generation, which is variable in three dimensions and derived from single laser beam. Previous work ignition of internal combustion engines found that simultaneously igniting more than one location resulted stable faster – key potential advantage over conventional ignition. However, previous approaches could only generate secondary foci at fixed locations. The reported here an...
Autonomous systems are increasingly being used in safety-and mission-critical domains, including aviation, manufacturing, healthcare and the automotive industry. Systems for such domains often verified with respect to essential requirements set by a regulator, as part of process called certification. In principle, autonomous can be deployed if they certified use. However, certification is especially challenging condition both system its environment will surely change, limiting effective use...
To expand our knowledge of the climate in Lesser Antilles, we attempted to identify spatio-temporal configurations daily weather. We noticed certain pitfalls that can lead poor results when using clustering algorithms and have proposed some steps towards solution. These advancements might prove interesting for informatics, as well many applications cluster physical fields. illustrated with a dataset cumulative rainfall from NASA's Tropical Rainfall Measuring Mission period 2000 2014. First,...
Public perceptions of Robotics and Artificial Intelligence (RAI) are important in the acceptance, uptake, government regulation research funding this technology. Recent has shown that public's understanding RAI can be negative or inaccurate. We believe effective public engagement help ensure opinion is better informed. In paper, we describe our first iteration a high throughput in-person activity. use light touch quiz-format survey instrument to integrate in-the-wild participation into...
We propose a technique to increase the robustness of snake-based segmentation method originally introduced track shape target with random white Gaussian intensity upon background. Because these statistical conditions are not always fulfilled optronic images, we describe two improvements that field application this approach. first show regularized whitening preprocessing allows one apply original successfully for correlated texture then introduce simple multiscale approach increases against...
In future inspections of offshore assets utilizing robots, robots will not only be expected to collate new data from their payload instruments, but they also interact with the infrastructure being inspected, undertake remedial tasks and engage embedded monitoring systems asset. This increasing level interaction deployment frequency robot requires an understanding how we can embed safe trusted operational architectures within robots. Currently, constrained semi-autonomous inspections, using...
We address the problem of localizing small targets with random gray levels that appear in background clutter. consider recently proposed maximum-likelihood ratio test (MLRT) algorithm, which scans observed scene an estimation window local statistics are estimated. In presence a spatially homogeneous background, we show if is few times larger than target itself, MLRT quasi-equivalent to optimal (ML) uses whole for estimating statistics. The thus constitutes efficient alternative ML algorithm...
Laser Ignition (LI) is a new, innovative technology which overcomes several limitations of conventional spark ignition; it presents many potential advantages for engine dynamic control, especially gasoline direct injection engine. This paper highlights new approach, using principle component analysis the pressure curve, to dynamically control multiple pulse laser ignition (MPLI). In results, experimental data how MPLI improves performance and demonstrates effectiveness in lean combustion....
A drive to reduce costs, carbon emissions, and the number of required personnel in offshore energy industry has led proposals for increased use autonomous/robotic systems many maintenance tasks. There are questions over how such missions can be shown safe. corollary exists manned aviation world helicopter–ship operations where a test pilot attempts operate from ship under range wind conditions provides subjective feedback on level difficulty encountered. This defines ship–helicopter...
The benefits of firing the second laser pulse controlled by predicting performance a cycle from cylinder pressure earlier in are investigated using methods that uses curve matching and neural networks.
This paper highlights a new innovative approach to create and control multiple pulse laser ignition applied on GDI engine. The results present the dual effectiveness in super lean condition it indicates has significantly improved engine stability.
The concept of a statistical filter for objects that comprise several regions is introduced. process optimal in the presence nonoverlapping noise target and may perform independently variations mean value every region. basic performance described, comparison with other types processing made.
We generated diffractive multi-beams for laser ignition using a spatial light modulator (SLM). Breakdown sparks with arbitrary geometrical location were created. The multi-beam patterns delivered to engine tests without clipping.
The task of detection objects composed several regions by means statistical filters is analyzed. target assumed to have different unknown mean values in each its regions. based on likelihood estimation, after performing an estimation the actual configuration region. A simplified filter that reduces computational complexity also proposed. performance analyzed theoretically and tested computer experiments.