Visual content analysis

 

Organization:

  • Teaching Load / Total Load: 45/90
  • Lectures/Exercices/Labs/Final Exam 1: 36/0/9/0

 

Objectives:

  • To master the core techniques for low-level image & video analysis as a preliminary step to interpretation and content-based access.
  • To understand related technological challenges and gain insight into emerging application issues.
  • To turn into practice computer vision applications (e.g. human motion analysis, object detection, scene activity monitoring...) by means of image & video analysis, exploiting the industry-standard Matlab platform capabilities.

Reference to CDIO Syllabus:

1.3 Advanced engineering fundamental knowledge, methods and tools 2.1 Analytical reasoning and problem solving
2.2 Experimentation, investigation and knowledge discovery
3.2.3 Written communication

4.7.1 Thinking creatively and imagining possibilities

Keywords: Visual feature extraction; denoising, enhancement & restoration; segmentation & grouping; motion estimation & tracking; shape analysis.

Prerequisites: None

Course outline:

  • Digital imaging products, vision (sub)systems and visual media-based services: current industrial issues and technological challenges of image & video processing and understanding
  • Image & video analysis: paradigms and models
  • Computational vision paradigms: hierarchical processing, low/mid/high-level
  • vision, visual features, Gestalt principles
  • Image & video models: functional, stochastic, statistical, algebraic
  • Still image analysis
  • Characterizing and exploiting global image properties: histogram techniques,

frequency filtering:

  • Extracting image local geometry: edge and corner detection
  • Binary and grey-level mathematical morphology
  • Inverse problems in image analysis: deterministic and stochastic regularization

page6image920

- Deterministic image segmentation: variational methods and graph cuts
 Active contours and level set methods
 The Mumford-Shah model, deterministic & statistical region competition,

multi-feature variational segmentation - Scale-space and PDE image filtering
- Bayesian methods, Markov Random Fields - Texture modeling and analysis

- Video analysis
- Motion measurement and optical flow estimation
- Spatio-temporal segmentation and object tracking

Assessment:

The assessment pattern involves 3 components: continuous evaluation via homework on selected topics (”coursework”) (CW), lab assignments (L), and a two-student group written final exam (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 (CW, L, E)) (S1)

  • 2nd session = O (S2)

  • Final grade = Max (SE1, SE2)

Learning materials and literature:

 
Learning materials: Documentation provided by lecturers
 
Literature:

  • 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

Person in charge: Dr. Nicolas ROUGON (nicolas.rougon@telecom-sudparis.eu)

 
Lecturers from Télécom SudParis:
  • Dr. Nicolas ROUGON
  • Dr. Catalin FETITA