Advanced Artificial Vision

Credits
4.5
Types
Elective
Requirements
This subject has not requirements , but it has got previous capacities
Department
CS;URV
Web
http://moodle.urv.cat/
This course aims at studying the fundamental techniques for image processing and advanced issues on machine vision related to the problems of automatic analysis and recognition of complex images. Practical applications will be developed on well-known machine vision software.

Weekly hours

Theory
1.5
Problems
0
Laboratory
1
Guided learning
0
Autonomous learning
5

Objectives

  1. To learn and practise the main algorithms and methods for image feature extaction.
    Related competences: CEA6, CEA14,
  2. To learn and understand the main concepts of image processing.
    Related competences: CEA6,
  3. To learn and practise the principal color and texture analysis methods.
    Related competences: CEA14, CEP5, CB7,
  4. To learn and practise the main image segmetation and classification techniques.
    Related competences: CEA14, CEP1, CEP5, CB7,
  5. To know some basics about stereoscopic vision and 3D models.
    Related competences: CEA14, CEP1, CEP5,
  6. To be able to analyze a real computer vision problem, and propose effective solutions.
    Related competences: CEP1, CEP5, CEP6, CT5, CB7,

Contents

  1. Chapter 1. Image Processing.
    Filtering and smoothing operations. Morphological techniques.
  2. Chapter 2. Feature Extraction.
    Lines and corners detection. Identification of basic geometrical structures.
  3. Chapter 3. Color and texture analysis.
    Color models, kinds of texture, texture feature extraction, geometrical methods.
  4. Chapter 4. Image Segmentation and Image Classification.
    Unsupervised segmentation based on regions and edges. Supervised classification, theoretical decision methods, statistical methods, neural networks.
  5. Chapter 5. Stereoscopic Vision.
    Camera calibration and camera systems, epipolar geometry, image rectification, search for correspondences, triangulation.
  6. Chapter 6. Perception and 3D Modeling.
    Range images generation, extraction of geometric elements, automatic scene generation, scene recognition, geometrical hashing.