A5106124934- Remote-Sensing Image Classification
- Advanced Image Fusion Techniques
- Remote Sensing and Land Use
- Image and Signal Denoising Methods
- Remote Sensing in Agriculture
- Spectroscopy and Chemometric Analyses
- Advanced Image and Video Retrieval Techniques
- Geochemistry and Geologic Mapping
- Image Retrieval and Classification Techniques
- Sparse and Compressive Sensing Techniques
- Synthetic Aperture Radar (SAR) Applications and Techniques
- Remote Sensing and LiDAR Applications
- Image Enhancement Techniques
- Automated Road and Building Extraction
- Underwater Acoustics Research
- Advanced Neural Network Applications
- Advanced Image Processing Techniques
- Advanced SAR Imaging Techniques
- Video Surveillance and Tracking Methods
- Photoacoustic and Ultrasonic Imaging
- Medical Image Segmentation Techniques
- Infrared Target Detection Methodologies
- Advanced Chemical Sensor Technologies
- Face and Expression Recognition
- Seismic Imaging and Inversion Techniques
Université Grenoble Alpes
2016-2025
Institut polytechnique de Grenoble
2016-2025
Grenoble Images Parole Signal Automatique
2016-2025
Chinese Academy of Sciences
2020-2025
Laboratoire Jean Kuntzmann
2019-2025
Centre National de la Recherche Scientifique
2016-2025
Centre Inria de l'Université Grenoble Alpes
2019-2025
Aerospace Information Research Institute
2020-2025
Institut national de recherche en informatique et en automatique
2020-2025
Berlin Institute for the Foundations of Learning and Data
2024
Imaging spectrometers measure electromagnetic energy scattered in their instantaneous field view hundreds or thousands of spectral channels with higher resolution than multispectral cameras. are therefore often referred to as hyperspectral cameras (HSCs). Higher enables material identification via spectroscopic analysis, which facilitates countless applications that require identifying materials scenarios unsuitable for classical analysis. Due low spatial HSCs, microscopic mixing, and...
Hyperspectral remote sensing technology has advanced significantly in the past two decades. Current sensors onboard airborne and spaceborne platforms cover large areas of Earth surface with unprecedented spectral, spatial, temporal resolutions. These characteristics enable a myriad applications requiring fine identification materials or estimation physical parameters. Very often, these rely on sophisticated complex data analysis methods. The sources difficulties are, namely, high...
Pansharpening aims at fusing a multispectral and panchromatic image, featuring the result of processing with spectral resolution former spatial latter. In last decades, many algorithms addressing this task have been presented in literature. However, lack universally recognized evaluation criteria, available image data sets for benchmarking, standardized implementations makes thorough comparison different pansharpening techniques difficult to achieve. paper, authors attempt fill gap by...
Classification of hyperspectral data with high spatial resolution from urban areas is discussed. An approach has been proposed which based on using several principal components the and build morphological profiles. These profiles can be used all together in one extended profile. A shortcoming that it primarily designed for classification structures does not fully utilize spectral information data. Similarly, a pixel-wise solely content performed, but lacks structure features image. extension...
In January 2006, the Data Fusion Committee of IEEE Geoscience and Remote Sensing Society launched a public contest for pansharpening algorithms, which aimed to identify ones that perform best. Seven research groups worldwide participated in contest, testing eight algorithms following different philosophies [component substitution, multiresolution analysis (MRA), detail injection, etc.]. Several complete data sets from two sensors, namely, QuickBird simulated Pleiades, were delivered all...
The high number of spectral bands acquired by hyperspectral sensors increases the capability to distinguish physical materials and objects, presenting new challenges image analysis classification. This letter presents a novel method for accurate spectral-spatial classification images. proposed technique consists two steps. In first step, probabilistic support vector machine pixelwise is applied. second spatial contextual information used refining results obtained in step. achieved means...
Hyperspectral (HS) images are characterized by approximately contiguous spectral information, enabling the fine identification of materials capturing subtle discrepancies. Owing to their excellent locally contextual modeling ability, convolutional neural networks (CNNs) have been proven be a powerful feature extractor in HS image classification. However, CNNs fail mine and represent sequence attributes signatures well due limitations inherent network backbone. To solve this issue, we rethink...
Scene classification of remote sensing images has drawn great attention because its wide applications. In this paper, with the guidance human visual system (HVS), we explore mechanism and propose a novel end-to-end recurrent convolutional network (ARCNet) for scene classification. It can learn to focus selectively on some key regions or locations just process them at high-level features, thereby discarding noncritical information promoting performance. The contributions paper are threefold....
Hyperspectral images provide detailed spectral information through hundreds of (narrow) channels (also known as dimensionality or bands) with continuous that can accurately classify diverse materials interest. The increased such data makes it possible to significantly improve content but provides a challenge the conventional techniques (the so-called curse dimensionality) for accurate analysis hyperspectral images. Feature extraction, vibrant field research in community, evolved decades...
In this letter, a technique based on independent component analysis (ICA) and extended morphological attribute profiles (EAPs) is presented for the classification of hyperspectral images. The ICA maps data into subspace in which components are as possible. APs, extracted by using several attributes, applied to each image associated with an component, leading set EAPs. Two approaches including computed analysis. features processing then classified SVM. experiments carried out two images...
Learning-based infrared small object detection methods currently rely heavily on the classification backbone network. This tends to result in tiny loss and feature distinguishability limitations as network depth increases. Furthermore, objects images are frequently emerged bright dark, posing severe demands for obtaining precise contrast information. For this reason, we paper propose a simple effective "U-Net U-Net" framework, UIU-Net short, detect images. As name suggests, embeds U-Net into...
Morphological profiles (MPs) have been proposed in recent literature as aiding tools to achieve better results for classification of remotely sensed data. MPs are general built using features containing most the information content data, such components derived from principal component analysis (PCA). Recently, nonlinear PCA (NLPCA), performed by autoassociative neural network, has emerged a good unsupervised technique fit hyperspectral data into few components. The aim this letter is...
The pansharpening process has the purpose of building a high-resolution multispectral image by fusing low spatial resolution and panchromatic observations. A very credited method to pursue this goal relies upon injection details extracted from into an upsampled version low-resolution image. In letter, we compare two different methodologies motivate superiority contrast-based methods both physical consideration numerical tests carried out on remotely sensed data acquired IKONOS Quickbird sensors.
Hyperspectral remote sensing images (HSIs) usually have high spectral resolution and low spatial resolution. Conversely, multispectral (MSIs) resolutions. The problem of inferring which combine the resolutions HSIs MSIs, respectively, is a data fusion that has been focus recent active research due to increasing availability MSIs retrieved from same geographical area. We formulate this as minimization convex objective function containing two quadratic data-fitting terms an edge-preserving...
Hyperspectral imagery collected from airborne or satellite sources inevitably suffers spectral variability, making it difficult for unmixing to accurately estimate abundance maps. The classical model, the linear mixing model (LMM), generally fails handle this sticky issue effectively. To end, we propose a novel mixture called augmented LMM, address variability by applying data-driven learning strategy in inverse problems of hyperspectral unmixing. proposed approach models main (i.e., scaling...
Hyperspectral remote sensing is based on measuring the scattered and reflected electromagnetic signals from Earth’s surface emitted by Sun. The received radiance at sensor usually degraded atmospheric effects instrumental (sensor) noises which include thermal (Johnson) noise, quantization shot (photon) noise. Noise reduction often considered as a preprocessing step for hyperspectral imagery. In past decade, noise techniques have evolved substantially two dimensional bandwise to three ones,...
Hyperspectral imaging (HSI) has been extensively utilized in many real-life applications because it benefits from the detailed spectral information contained each pixel. Notably, complex characteristics, i.e., nonlinear relation among captured and corresponding object of HSI data, make accurate classification challenging for traditional methods. In last few years, deep learning (DL) substantiated as a powerful feature extractor that effectively addresses problems appeared number computer...
The foundation model has recently garnered significant attention due to its potential revolutionize the field of visual representation learning in a self-supervised manner. While most models are tailored effectively process RGB images for various tasks, there is noticeable gap research focused on spectral data, which offers valuable information scene understanding, especially remote sensing (RS) applications. To fill this gap, we created first time universal RS model, named SpectralGPT,...