Labs exercises in NTUA (2018) for the Computer Vision course 8th semester
In this course, we studied several concepts of Computer Vision such as:
- Image Geometry and Creation of Images
- 2D and 3D Image analysis with linear filters and Fourier Analysis (mostly Gabor filters and Wavelts) or non-linear methods.
- Multiscalar Image Analysis
- Edge and Feature Detection
- Shape Analysis
- Texture Analysis and Modeling
- Segmentation
- Detection of 2D Optical Flow and 3D Movement
- 3D Shape estimation
- Object and Action Detection
The course also has a lab with the following assignments, which were done using Matlab:
Lab1:
Edge detection in noisy grayscale images (Part 1), interest points detection (Part 2) and image classification using the BOVW (bag of visual words) concept (Part 3). In Part 3, the training dataset is not provided due to its large size.
Lab2:
Tracking of human face using the skin colour. The training was
done by a
parameter fit to a 2D Gaussian function (Part 1). In the next part, we calculated the optical flow of the face along the different frames of the video and applied the Lucas-Kanade Algorithm in order to keep track of it (Part 2).
Contributors:
- manzar96
- alexkaf