Here we have the Courses offered by UPB CEOSpaceTEch

The use of a coded aperture in imaging systems has resolved problems like the tradeoff between the angular resolution and SNR of telescopes, reduced the acquisition time of multispectral images to a single shot, improved the depth inference by analyzing blurred images and allowed the flattening of cameras by eliminating the lenses, to enumerate only a few of achievements. The image obtained with a coded aperture has to be processed in order to become intelligible. In that respect, the Coded Aperture Imaging belongs to the larger domain of Computational Imaging. The course presents several imaging systems with coded aperture, by focusing on image formation, coded aperture design, and reconstruction from the sensed image.


1.     Basics of imaging systems

2.     Radiation imaging

3.     Lensless coded aperture imaging 

4.     Single Pixel Imaging

5.     Coded Aperture Snapshot Spectral Imaging (CASSI)

6.     Depth from Defocus (DfD) 

7.     Emergent techniques

The deluge of Erath Observation (EO) images counting hundreds of Terabytes per day needs to be converted into meaningful information, largely impacting the socio-economic-environmental triangle.  Multispectral and microwave EO sensors are unceasingly streaming millions of samples per second, which must be analysed to extract semantics or physical parameters for understanding Earth spatio-temporal patterns and phenomena. The course presents AI solutions to leverage the synergy of EO sensors records for understanding and prediction of Earth processes. 


1. Sensory and semantic gaps

2. Hierarchical models

3. Explainable AI

4. Physics-aware AI

5. Self-learning AI

6. Virtual sensors

7. Satellite Image Time Series

Big Data Analytics, specifically for Earth Observation (EO) imagery, besides to the prevalent 3Vs, is postulating additional challenges emerging from its very particular nature: data sources are sensors and instruments as multispectral or Synthetic Aperture Radar (SAR), information is spatio-temporal, meaning is quantitative as physical parameters and qualitative as semantic descriptors, understanding is contextual in synergy with multi-sensor, in-situ, geoinformation and other sources of information. Therefore, the goal of the lecture is the presentation of specific leading edge concepts, methods and algorithms for information content exploration and intelligence extraction from Big Data provided by EO sensors and other related sources. The lecture offers a cross-disciplinary view of methods in signal processing, machine learning, deep learning, visualization and data mining also addressing the meaning extraction and semantic representations. Looking to the near future technologies, basic elements of quantum information processing will be also presented.


1. Introduction to EO

2. Data modeling and description 

3. Visual Data Mining

4. Big EO Data Mining

5. Benchmarking