Computer Vision Methods and Applications

Short Description

The Computer Vision Methods and Applications (CVMA) course is intended to equip students with the necessary theoretical and practical understanding of image processing and computer vision techniques to enable them to meet the challenges of building advanced image-based applications. Examples of potential vision-based applications include: image understanding in mobile devices (cameras, phones, tablet computers etc.), robot vision systems, autonomous vehicle guidance and road monitoring, driver attention monitoring, image database query systems, creative media production tools, interactive gaming, augmented reality and visual biometrics, forensic image analysis, security and surveillance, and medical imaging. The course will focus on the application of recent advances in Computer Vision techniques that underpin a wide variety of systems and products based on methods such as: face detection, object recognition, tracking, segmentation and 3D imaging.

Requirements of Entry

No previous experience in computer vision is required, however a basic understanding of mathematics concepts, for example: matrices, vectors, elementary calculus and basic probability would be helpful (as provided by Math1RS or Math1RT), but not essential.

Assessment

Examination 100%.
The laboratory work will be assessed in a compulsory question set in the examination.
Main Assessment In: April/May

Course Aims

1. To provide a theoretical and practical understanding of 2D and 3D visual perception based on current image analysis techniques and currently available vision software libraries.
2. To equip the student with the ability to tackle the practical aspects of developing algorithms for vision-based applications as listed above (section 13). Therefore, CVMA will provide the student with the basic tools to undertake Level 4 and Masters projects that require vision to be applied within in these related disciplines.
3. To prepare the student for a career in Industry as a Computer Vision specialist in areas such as Research & Development, Technical Marketing and Intellectual Property Management; or for an Academic career, e.g. PhD research or Research Assistantship.
Intended Learning Outcomes of Course

By the end of the course students will be able to:

1. Analyse critically computer vision algorithms and applications based on knowledge of image representation, image formation and basic processing techniques;
2. Implement feature extraction and object recognition algorithms;
3. Critically evaluate the basic geometric concepts in 3D computer vision and employed in recovery of 3D
surfaces from stereo-pair images, or motion fields from image sequences;
4. Demonstrate the ability to apply the rudiments of information theory and basic image compression
techniques to the design of image coding/decoding algorithms;
5. Demonstrate competence in the use of the MATLAB programming language for vision-based
applications prototyping. Demonstrate competence in the application of the key current image analysis
libraries, for example, supporting the Microsoft Kinect sensor for 3D imaging and human interaction
interpretation.