Digital processing of signal and images is essential in many applications of optics. Optical communications, remote sensing, industrial control often require signal detection, parameter estimation, target identification.

As another example, in modern imaging systems,optics is designed together with image processing algorithms. Basic knowledge of image processing is thus essential to develop optical systems and quantify their performance.

This course is an introduction to signal and image processing for optics scientists. Half of it consists of « interactive » lectures where basic principles are explained and illustrated with exercises. The second half consists of laboratories where students develop signal and image processing algorithms using Matlab.

1. Basics of probability theory and random functions
• Random variables used in physics, random vectors, central limit theorem,

2. Introduction to estimation theory:
• Bias and variance of an estimator, Maximum likelihood, Cramer-Rao lower bound, matched filter
Application to distance and position estimation (radar, lidar, …).

3. Introduction to detection theory :
• Neyman-Pearson theory, likelihood ratio, nuisance parameters, generalized likelihood ratio
Application to radar, communications, edge detection in images

Requirements : Basic Fourier analysis, probability theory

Evaluation mechanism : Labwork reports (1/3) Project report (1/3) Written examination (1/3)

Last Modification : Saturday 15 September 2012