Télécom ParisTech

Image (IMA)

This program aims at providing the students with the necessary background, competences and skills for engineering and research positions in image processing, image interpretation, computer vision, 3D imaging, in various domains (biomedical imaging, computational photography, scene modeling and synthesis, remote sensing, biometry, defense, etc.).

The program includes the foundations of image processing as well as advanced courses on mathematical methods for images, computer vision and 3D reconstruction, AI inspired methods for image analysis, image classification and indexing, as well as the basis of video processing. The main application domains (medical imaging, aerial and satellite imaging, digital photography) are presented in the courses by researchers active in these fields, and industrial partners will describe other applications (biometry, industrial vision...).

The program provides strong and sound knowledge in the domain of image processing and image interpretation that will be useful both in industry and in research labs, with subsequent doctoral studies.

It can be associated with the IGR track (Human-Computer Interaction and 3D Computer Graphics) for topics related to virtual reality and computer graphics, as well as with the SIGMA track (Digital Signal Processing) or the SD track (Data Science), for learning and mining methods.

Students are expected to have previously acquired a solid background in applied mathematics, computer science, signal processing.

2nd year courses

IMA 2nd year program (192 hours)  Fall semester

Spring semester

Period 1 Period 2 Period 3 Period 4
Time slot A1 IMA201 Introduction to image processing IMA203 Variational and Bayesian methods / discrete optimization IMA205 3D computer vision and object recognition IMA207 Remote sensing / industrial applications
Time slot A2 IMA202 Multi-scale and morphological representations  IMA204 Biomedical imaging / Knowledge representation IMA206  Computational photography / patch-based methods IMA208 Multimedia / projects


Courses are associated with projects on the whole periods.

Fall semester, period 1

  • IMA 201 Introduction to image processing : 24 hrs
    This course covers the foundations of digital image processing and analysis: Acquisition pipeline and digital photography devices (optics, sampling, noise, radiometry, color). Restoration and image improvement (denoising, inverse problems and deconvolution, super-resolution, inpainting, etc.). Image analysis (differential operators, segmentation, local descriptors, texture, shape representation). This course is is a prerequisite for all courses proposed in the IMA track.

  • IMA 202 Multi-scale and morphological representations : 24 hrs
    Practical work in image processing: introduction, restoration, filtering, segmentation, descriptors
    Mathematical tools for image representations, at different scales: Multi-scale Gaussian and Laplacian representations. Wavelets. Discrete representations. Mathematical morphology.

Fall semester, period 2

  • IMA 203 Variational and Bayesian methods / discrete optimization : 24 hrs
    This course introduces variational and Bayesian methods for image filtering and segmentation. Particular attention will be paid to optimization methods involved in this context. Applications from the fields of digital photography, remote sensing and medical images will illustrate the interest of these approaches. Variational methods. Deformable models. Bayesian methods, Markov random fields. Discrete optimization using graph cuts.

  • IMA 204 Biomedical imaging / Knowledge representation : 24 hrs
    The first part of this course is dedicated to structural representation for image interpretation: Graph-based representations. Recognition and matching and artificial intelligence based approaches: Uncertainty modeling, Fusion, Higher level representations, Model- and knowledge-based  methods. The second part of the course offers an overview of medical imaging. It summarizes the main acquisition techniques (X-rays, nuclear, magnetic resonance, ultrasound imaging, tomographic reconstruction), and present applications in image processing, such as brain, cardiovascular, retinal, and biological imaging which give the opportunity to discuss the advanced methods in segmentation and modeling studied in the previous courses.

Spring semester, period 3

  • IMA 205 3D computer vision and object recognition : 24 hrs
    Ce cours présente l'ensemble de la chaîne permettant de passer de photographies d'une scène à sa représentation tri-dimensionnelle. Il y est également proposé une introduction à un autre domaine central de la vision par ordinateur, la reconnaissance d'objets. - Calibrage de caméras - Stéréovision - Nuages de points 3D - triangulation et modélisation géométrique - Descripteurs locaux, apprentissage, sacs de mots, reconnaissance de catégories d'objets et indexation.
  • IMA 206 Computational photography / patch-based methods : 24 hrs
    This course presents the complete pipeline from a photo to its three-dimensional representation. An introduction to object recognition, a core domain of computer vision, is also proposed. Camera calibration. Stereovision. 3D point cloud processing. Triangulation and geometric modeling. Local descriptors, learning, bags of words, object categorization and indexing.

Spring semester, period 4

  • IMA 207 Remote sensing / industrial applications : 24 hrs
    The first part of this course is about aerial and satellite imaging, from the physical principles to the applications in image processing: - Sensors - Optical imaging - Radar imaging - Multi- and hyper-spectral imaging - Applications. The second part will address a few industrial applications in image processing, with courses and conferences given by presenters from various origins, in particular from the industries of: - Biometry - Astrophysics - Image quality - Nondestructive control - Artificial retina.
  • IMA 208 Multimedia / projects : 24 hrs
    The first part of this course deals with video and multimedia, focusing on coding, compression, and analysis of dynamic scenes: - Movement estimation - Background suppression - Object tracking - Coding and compression - Multimedia applications. In this period, projects on all the topics covered by the courses of the IMA track will be completed.



3rd year options

The IMA education track includes a large number of options in the third year that can be selected by the students:

One option is located on the school premises

  •  Specialization IMAGE  at Telecom ParisTech, which includes courses chosen in the masters, in other tracks, or specific advanced courses, both on theory and on applications (from image improvement to image understanding, image mining, new imaging sensors...), with an insight on topics not covered during the 2nd year *large image databases, multimedia, image and art...).

Students may also apply to one of the following Masters of Science (M2) offered by the Paris-Saclay campus (UPSA) or other universities in Paris

  • (MVA) Mathematics, Vision and Learning (see Mathematics and applications  School)
  • (AIC) Machine Learning, Information and Content, (see Computer Science  School)
  • (ATSI) Control theory, signal and image processing  (see Electrical Engineering  School)
  • (IMA) Master in computer sciences, Image track (UPMC, Université Pierre et Marie Curie ) (see website IMA )
  • (BIM) Master in « Bioengineering »,  Bioimaging track, in English (Université ParisDescartes), (see website  BIM)