CES8AH FROM DIGITAL PHOTOGRAPHY TO COMPUTATIONAL PHOTOGRAPHY

CES8AH

From digital photography to computational photography

Credits : 4 ECTS

Duration : 36 hours

Semester : S8

Person(s) in charge:

Frédéric SUR, Associate Professor,  frederic.sur@mines-nancy.univ-lorraine.fr

Keywords:

signal and image processing, algorithms, Fourier analysis, applied statistics, optimization

Prerequisites:
first-year courses in mathematics, statistics, computer science.

Goal:
Understand the mathematical models of digital photography and of image processing.

Program and contentes:

The goal of this elective course is to detail the whole image processing chain inside a digital camera, and the practical consequences of the technological constraints.

Topics of interest are:

• general introduction to photography
• fast Fourier transform
• sampling and quantization
• lossless compression, statistical theory of information
• lossy compression: JPEG
• color and demosaicing
• noise in digital images, and denoising algorithms
• image restoration, deconvolution
• panoramic image stitchig, HDR imaging

Hands-on-training with Matlab illustrates the theory.

Additional material is available on a dedicated web page.

Compétences : 

Niveaux

Description et verbes opérationnels

Connaître 

Comprendre 

Appliquer 

Analyser 

Synthétiser

Évaluer

Évaluations :

• Test écrit

• Contrôle continu

• Oral, soutenance

• Projet

• Rapport