Image and video understanding

Organisation

Cours/TD/TP/projet/examen : 33/0/12/0

Acquis d'apprentissage

At the end of this teaching unit, students will be able

- To specify & use core techniques for solving basic low-level image & video analysis problems, including denoising & enhancement, feature extraction, segmentation, motion analysis & tracking.

- To design & conduct an experimental performance study of an image understanding technique using the industry-standard Matlab Ò platform.

- To combine these methodological individual building blocks into an image / video understanding pipeline dedicated to a target application.

- To draft a concise report summarizing the methodological / technological / application-related state-of-the-art and emerging trends of an image / video understanding topical issue.

Compétences CDIO

• 1.3 - Advanced engineering fundamental knowledge, methods and tools
• 2.1 - Analytical reasoning and problem solving
• 2.2 - Experimentation, investigation and knowledge discovery
• 3.3.1 - Communications in English
• 4.7.3 - Thinking Creatively and Imagining Possibilites (which builds on and expands Creative Thinking 2.4.3)

Prérequis

Aucun

Mots-clés

Image modeling; visual feature extraction; image denoising, enhancement and restoration; image segmentation; motion estimation; video object tracking

Contenu

> Digital imaging products, systems and services: current industrial issues and technological challenges in image & video processing/understanding.

> Computational vision paradigms: low/mid/high-level vision, visual features, perceptual principles, mathematical image models.

> Still image analysis
- Digital imaging basics
• Image sampling (pixel grids, neighborhood systems, digital connectivity), quantization, color.
• Statistical image properties: histogram, local statistics
• Frequential image content: spatial/frequential resolution, Shannon theorem, image spectrum; local spectrum, Gabor filtering
• Changing image resolution: interpolation, multigrid representations
• Texture
- Local image geometry: edge, corner and characteristic line detection
- Binary and grey-level image morphology
- Image denoising, enhancement and restoration: morphological filtering, PDE filtering, NL-means
- Image segmentation: active contours, level set methods, region competition, Markov Random Fields

> Video analysis
- Motion estimation: dominant motion, optical flow
- Spatio-temporal segmentation, object tracking

Evaluation

The assessment pattern is practice-oriented and involves 3 components: continuous evaluation via homework (HW), labs (L), and a two-student group micro-project (E). The final grade is a weighted average of individual component grades. The 2nd session consists of a study with an oral defense (O).
- 1st session = Weighted Average (HW, L, E)) (S1)
- 2nd session = O (S2)
- Final grade = Max (SE1, SE2)

Formule de l'évaluation

L’évaluation du module, orientée vers la pratique expérimentale et les applications, repose sur 3 composantes : un contrôle continu sous forme de travail personnel hors présentiel (CC), des bureaux d’études (BE), et un micro-projet final en binôme (CF). La 2ème Session consistera en une étude avec soutenance orale (O).
- 1re session = Moyenne Pondérée (CC, BE, CF)) (S1)
- 2e session = (S2)
Note finale = Max (SE1, SE2)

Bibliographie

Supports de cours : Documentation remise par les intervenants
Bibliographie :
- A. Bovik (Ed.). Handbook of Image & Video Processing. Academic Press, 2000
- L.G. Shapiro and J-C. Stockman. Computer Vision. Prentice Hall, 2001
- E.R. Davies. Machine Vision: Theory, Algorithms, Practicalities. Academic Press, 1997
- R. Jain, R. Kasturi and B.G. Schunck. Machine Vision. McGraw-Hill, 1995