- Medical Image Segmentation Techniques
- Image Retrieval and Classification Techniques
- Advanced Vision and Imaging
- Advanced Image Fusion Techniques
- Image and Object Detection Techniques
- Digital Image Processing Techniques
- Image and Signal Denoising Methods
- Color Science and Applications
- Image Enhancement Techniques
- Industrial Vision Systems and Defect Detection
- Point processes and geometric inequalities
- AI in cancer detection
- Advanced Numerical Analysis Techniques
- Neural Networks and Applications
- Computer Graphics and Visualization Techniques
- Advanced Image and Video Retrieval Techniques
- Surface Roughness and Optical Measurements
- Manufacturing Process and Optimization
- Image Processing and 3D Reconstruction
- Advanced Surface Polishing Techniques
- Laser Material Processing Techniques
- Radiomics and Machine Learning in Medical Imaging
- Mathematical Dynamics and Fractals
- Bacterial Identification and Susceptibility Testing
- Dental Radiography and Imaging
Centre National de la Recherche Scientifique
2004-2024
Laboratoire Hubert Curien
2014-2024
Université Jean Monnet
2004-2022
International Prevention Research Institute
2015-2020
International Performance Research Institute
2018
Université Claude Bernard Lyon 1
2010-2013
École d'Ingénieurs en Chimie et Sciences du Numérique
1995-2011
École Nationale d'Ingénieurs de Saint-Étienne
2011
Laboratoire Interuniversitaire des Systèmes Atmosphériques
1996-2007
Laboratoire Techniques, Territoires et Sociétés
1996-2007
Precise nanostructuration of surface and the subsequent upgrades in material properties is a strong outcome ultra fast laser irradiations. Material characteristics can be designed on mesoscopic scales, carrying new optical properties. We demonstrate this work, possibility achieving modifications using short pulses, via polarization dependent structures generation, that generate specific color patterns. These oriented nanostructures created metal surface, called ripples, are typically smaller...
SUMMARY Up to now, image processing and analysis techniques have borrowed their basic tools from functional analysis: Fourier filtering, differential integral calculus, so on. These tools, however, only realize efficiency when they are put into a well‐defined algebraic frame, most of the time vectorial nature. Unfortunately, class functions modelling ‘images’, commonly referred as ‘grey tone functions’ does not necessarily present this very type structure. We here an operation for ‘addition’...
Logarithmic images, such as images obtained by transmitted light or those produced the human visual system, differ from linear images. Their processing and analysis require consequently specific laws structures. The latter have been developed in concept of a logarithmic image (LIP) model (Jourlin & Pinoli, 1987, 1988; 1987a). This permits introduction well-justified contrast definition: physical point view, it is closely linked with mathematical set up an algebraic structure. applications...
Thermal images are widely used for various applications such as safety, surveillance, and Advanced Driver Assistance Systems (ADAS). However, these typically have low contrast, blurred aspect, resolution, making it difficult to detect distant small-sized objects. To address issues, this paper explores preprocessing algorithms improve the performance of already trained object detection networks. Specifically, mathematical morphology is favor small bright objects, while deblurring...
SUMMARY The digital processing of serial electron‐microscope sections containing laser‐induced topographical references allows a three‐dimensional (3‐D) reconstruction entire cells at depth resolution 40–60 nm by the use novel image analysis methods. images are directly processed video‐camera placed under electron microscope in TEM mode or counting device STEM mode. deformations associated with cutting embedded back‐calculated new computer algorithms developed for and treatment. They correct...
In this paper, we introduce two new morphological transforms for pattern matching in gray scale images. They rely on a profiling approach and are defined the context of mathematical morphology. The first transform allows to detect all occurrences single an image, which justifies name SOMP (single object using probing). It is shown have properties metric therefore returns measure similarity between search image reference pattern. Other relative noise computation time highlighted. second MOMP...
The Fourier transform is well suited to the study of stationary functions. Yet, it superseded by Wavelet for powerful characterizations function features such as singularities. On other hand, LIP (Logarithmic Image Processing) model a mathematical framework developed Jourlin and Pinoli, dedicated representation processing gray tones images called hereafter logarithmic images. This mathematically defined model, comprising Transform "of its own", provides an effective tool obtained transmitted...
Logarithmic Image Processing (LIP) Model consists of a mathematical and physical framework dedicated to image processing. In this paper, novel application LIP operators is presented: we explain how addition subtraction can be used simulate variation sensor sensitivity or changes in source intensity depending on the transmitted reflected imaging context. Such property generates an interesting result another operator: additive contrast independent brightness variations, due exposure time...
Summary This paper presents a global solution from acquisition to visualization for the three‐dimensional reconstruction of cell sections. Original techniques are proposed correct handling geometrical section distortions, and new interpretation based on logarithmic image processing (LIP) model is applied in order create normalized grey‐level sections where these missing. Finally, method generating mesh triangles describe envelope reconstructed proposed, as well mixing synthesis information....
In this paper, we propose a new algorithm for line pattern segmentation. It allows to recover structures of various widths and intensities removes other than structures, effects non-uniformity the background level noise. This is composed two stages : enhancement extraction. Firstly, carried out using matching method, relying on profiling approach defined with mathematical morphology operators. transform detect multiple objects varying size or shape simultaneously unique template. For...
Thermal sensors are underrepresented in the field of Advanced Driver Assistance Systems whereas their capabilities to acquire images independently weather or daytime can be very helpful achieve optimal pedestrian and vehicle detection.This underrepresentation is due small amount available public datasets.This lack training samples difficulties building such datasets a real hurdle development an object detector dedicated thermal images.Thanks YOLOv4 its detection performance, we show this...