Image Processing And Analysis.docx

Image Processing and Analysis; Automatic Classification;

Analysis of Multitemporal Images; Regression and Estimation

Applications  Forestry  Agriculture  Damage Assessment  Urban area analysis  Building Detection  Risk assessment  Vegetation monitoring  Water quality  Snow and ice  Climatic changes Image Processing and Analysis 

Hyerarchical multilevel segmentation

Multiscale/multilevel feature extraction



Texture extraction

Image denoising

Automatic Classification  Statistical methods 

Machine learning (neural networks, support vector machine, ect)



Kernel methods



Semisupervised classification



Domain adaptation and active learning methods



Advanced multiscale feature extraction for VHR image classification



Feature selection and classification of hyperspectral images





Accuracy assessment in VHR classification maps



Classification of SAR signals

Analysis of Multitemporal Images 

Registration of multitemporal images



Modeling and mitigation of registration noise



Analysis and classification of multi- and hyper-temporal data



Change detection in medium and very high resolution SAR images



Change detection in medium and very high resolution Multispectral images



Change detection in medium and very high resolution Hyperspectral images

Regression and Estimation 

Model based regression methods



Regression based on machine learning techniques



Biophysical parameters estimation

RADARS Systems 

Synthetic Aperture Radar (SAR)

Ground penetrating radar and radar sounder

Applications 

Earth observation

Planet observation

Synthetic Aperture Radar (SAR) 

Analysis and modelling of HR and VHR SAR signals



Building detection and 3D reconstruction in VHR SAR images



Change detection (2D and 3D)



Automatic classification



Analysis of temporal series of HR and VHR SAR images.

Ground Penetrating Radar and Radar Sounder



Statistical analysis and characterization of radar sounder signals



Pre-processing techniques (clutter estimation and noise reduction, filtering, etc.)



Feature extraction and automatic classification of subsurface layers or patterns



Design of radar sounding instruments for planetary exploration

Radar sounder data of the subsurface of the North Pole of Mars acquired by the SHARAD instrument, and examples of possible elaborations: (top right) automatic detection of basal returns aimed at the estimation of ice thickness and basal topography; (bottom left) automatic detection of surface clutter returns on radargrams through clutter simulation and matching with real data; (bottom right) automatic detection and characterization of subsurface linear features aimed at the mapping of icy layers.

Radar sounder data of the subsurface of Byrd Glacier Antarctica acquired by the MCoRDS instrument, and examples of possible elaborations:

Examples of ice subsurface target classes backscattering and statistical analysis of the measured radar signal

Automatic detection of ice subsurface targets based on segmentation

Automatic classification of the ice subsurface based on feature extraction and machine learning

Pattern Recognition

Biomedical Signals and Image, Neuroscience, Industrial Visual Inspection, Content-Based Image Retrieval

Methods 

Supervised, semisupervised and unsupervised classification



Statistical methods for data analysis



Machine learning (neural networks, support vector machine, etc.)



Kernel-based methods and support vector machines



Domain adaptation and active learning algorithms



Multidimensional signal processing



2D and 3D image processing



Data fusion

Applications 

Remote sensing

Biomedical Signals and Images



Neuroscience

Industrial Visual Inspection



Others Biomedical Signals and Images  Analysis

of retina images for diseases detection and

mapping  Analisys

of MRI and fMRI images

 Analysis

of TAC images

 Analysis

of ECG and ECoG signals

 Analysis

of EEG signals

 Development

of Brain Computer Interface (BCI)

systems Neuroscience 

Analysis of fMRI signals



Analysis of EEG and MEG signals



Fusion between EEG, MEG and fMRI data



Pattern recognition for cognitive analysis

Laboratory of Functional Neuroimaging, CIMeC.

This activity is developed in cooperation with CIMeC – Centro interdipartimentale Mente/Cervello (Center for Mind/Brain Sciences), University of Trento. Industrial Visual Inspection 

Fig Quality Assessment

This activity is developed in cooperation with the Vision-Image Processing and Pattern Recognition Laboratory, Süleyman Demirel University, Isparta, Turkey. Content-Based Image Retrieval (CBIR) 

Image Feature Extraction for CBIR problems



Fast content-based Image Retrieval



Relevance Feedback Driven by Active Learning

Multisensor System

Data Fusion

Wireless Sensor Networks

Data Fusion 

Multi-resolution fusion (pansharpening)



Multi-sensor classification techniques



Integration of remote sensing, ancillary and in situ data



Architectures for multisensor data analysis



Change Detection in Multi-Temporal Multi-Sensor data

Systems for the integration of Remote Sensing and Wireless Sensor Networks



Remote Sensing-based framework for automatic and scalable Wireless Sensor Networks deployment planning in forests

LiDAR

Light Detection and Ranging 

Analysis of multireturn LiDAR data



Automatic classification of LiDAR data



Forest segmentation in the 3D LiDAR cloud space



Estimation of tree height, diameter and volume



Forest mapping using airborne LiDAR data



Building edge detection in the 3D LiDAR cloud space



Fusion between terrestrial and airborne LiDAR data



Fusion between airborne LiDAR data and hyperspectral/multispectral images