The aim of this module is introduce the students to computer vision including basics of human visual system and image perception, acquisition and processing. In terms of processing, the module deals with low-level pixel-based transforms, linear, nonlinear and morphological filtering, Fourier analysis, multiscale representations, extraction of simple features and image descriptions. Furthermore, elementary grouping, segmentation and classification strategies will be discussed as well as quality and assessment methodologies for image processing algorithms. To put into practice the algorithms and techniques, the students will work on a concrete project along the course. The aim is to provide an applied knowledge of a broad variety of Computer Vision techniques applied to solve a real-world vision problem. The project goal is to detect specific objects in images using basic CV techniques such as linear and non-linear filtering segmentation, grouping, template matching, modeling, etc. The knowledge obtained can be used in a wide variety of applications, for instance, quality control, generic object detection, security applications, etc.
Project title: Museum Painting Retrieval
The aim of the project is to learn the basic concepts and techniques to build a simple query by example retrieval system for finding paintings in a museum image collection. The image retrieval will be performed using colour, texture, text information, key-points and local descriptors. Some other techniques will also be applied, such as morphological filters to detect and remove overlying text from images or filtering to remove the noise from the images. The resulting system can be applied to any small query-by-example problem
M1 Schedule – Academic Year 2021-2022 – Student Guide <here>