Image Processing and Analysis
DCU has a number of groups involved a wide range of computer vision research applications. This includes both static images and video analysis.
Deep Learning @ DCU
Artificial Intelligence (AI) technology and more specifically, Deep-Learning Neural Networks (DLNN) are transforming the smart video and image analytics markets. This is a strategic focus for DCU and we have a number of postdocs active in this space and a number of collaborative research projects with industry.
DLNN has already proven to significantly improve performance in many computer vision applications, for example - object detection, object tracking, face recognition, image classification, scene labelling and others.
Example projects demonstrate the breadth of expertise:
Smart Stadium for Smarter Living
DCU is leading a collaborative research project with Croke Park Stadium, Intel and Microsoft to develop Ireland’s first national test-bed for IoT (Internet of Things). The project combines Intel’s scalable sensor network, Microsoft’s cloud, Croke Park’s IT-infrastructure and DCU’s video analytics expertise to provide a complete end-to-end IoT pipeline – “sense-aggregate-analyse-feedback”. A core focus for DCU was research into advanced video analytics algorithms and a number of working demos were developed in real-life match day environments.
- A number of the computer vision research projects are medical. These include: Machine learning for sub-cellular images from electron microscopes
- Computer Aided Detection of ultrasound images for reproduction in cattle
- Computer analysis of MRI images for diagnoses
Other breakthrough image processing and analysis technologies include:
- Best in class crowd counting algorithms
- Optical character recognition for automatic logo detection