Représentation des images numériques. Éléments de perception visuelle. Géométrie des systèmes d'acquisition d'images. Amélioration d'images et restauration d'images. Segmentation et identification de primitives. Analyse des images à partir de leur contenu. Compression d'images et standard de compression.

STE 5023
562-5800 poste 6707
laganier@eecs.uottawa.ca
http://www.site.uottawa.ca/~laganier

Le cours se base sur le livre suivant:
OpenCV 3 Computer Vision Application Programming Cookbook,
Packt Publishing, third edition, 2017.

Les notes de cours se trouvent ici.

• Le cours abordera les thèmes suivants:
  • The human visual system .
  • Image acquisition.
  1. Playing with Images. Installing the OpenCV library. Loading, displaying, and saving images. Exploring the cv::Mat data structure. Defining regions of interest.
  2. Manipulating the Pixels. Accessing pixel values. Scanning an image with pointers. Scanning an image with iterators. Writing efficient image-scanning loops. Scanning an image with neighbor access. Performing simple image arithmetic. Remapping an image.
  3. Processing the colors of an image. Comparing colors using the Strategy design pattern. Segmenting an image with the GrabCut algorithm. Converting color representations. Representing colors with hue, saturation, and brightness.
  4. Counting the Pixels with Histograms. Computing the image histogram. Applying look-up tables to modify the image appearance. Equalizing the image histogram. Backprojecting a histogram to detect the specific image content. Using the mean shift algorithm to find an object. Retrieving similar images using the histogram comparison. Counting pixels with integral images.
  5. Transforming Images with Morphological Operations. Eroding and dilating images using morphological filters. Opening and closing images using morphological filters. Applying morphological operators on gray-level images. Segmenting images using watersheds. Extracting distinctive regions using MSER.
  6. Filtering the Images. Filtering images using low-pass filters. Downsampling images with filters. Filtering images using a median filter. Applying directional filters to detect edges. Computing the Laplacian of an image.
  7. Extracting Lines, Contours, and Components. Detecting image contours with the Canny operator. Detecting lines in images with the Hough transform. Fitting a line to a set of points. Extracting connected components. Computing shape descriptors.
  8. Detecting Interest Points. Detecting corners in an image. Detecting features quickly. Detecting scale-invariant features. Detecting FAST features at multiple scales.
  9. Describing and Matching Interest Points. Matching local templates. Describing and matching local intensity patterns. Matching keypoints with binary descriptors.
  10. Estimating Projective Relations in Images. Computing the fundamental matrix of an image pair Matching images using random sample consensus Computing a homography between two images Detecting a planar target in images
  11. Reconstructing 3D Scenes. Calibrating a camera Recovering camera pose Reconstructing a 3D scene from calibrated cameras Computing depth from stereo image
  12. Processing Video Sequences. Reading video sequences. Processing the video frames. Writing video sequences. Extracting the foreground objects in a video.
  13. Tracking Visual Motion. Tracing feature points in a video. Estimating the optical flow. Tracking an object in a video.
  14. Learning from Examples. Recognizing faces using nearest neighbors of local binary patterns. Finding objects and faces with a cascade of Haar features. Detecting objects and people with Support Vector Machines and histograms of oriented gradients.

25% [MS] Quiz
30% [EF] Examen final (et voici un petit examen de pratique).
45% [ND] Projet

La description du projet se trouve ici.

La présence au cours est obligatoire. Comme le stipulent l'ensemble des règlements scolaires, l'étudiant(e) qui ne se présente pas à au moins 80 % du cours ne pourra pas écrire l'examen final.

Toutes les composantes du cours (tels que les rapports de laboratoire, les devoirs, etc.) doivent être remplies sinon l'étudiant(e) pourrait recevoir la note de INC comme note finale (équivalente à un F).

Pour satisfaire aux exigences du cours, l'étudiant devra obtenir une note d'au moins 50% au total de ses examens

La fraude scolaire peut avoir des conséquences importantes