Beyond human vision

Research within this theme looks at building a generation of computer vision systems with capabilities that exceed those of human vision. This involves studying the techniques and tools for computer vision, image processing and analysis, and pattern recognition. Our work is diverse and includes developing new algorithms and vision hardware.

To find out more about research in this area, contact Professor Richard Wilson or read our recent publications.


Theme lead: Professor Richard Wilson

Dr Adrian Bors

Dr Siamak Fayyaz

Professor Edwin Hancock

Dr Suresh Manandhar

Dr Nick Pears

Dr William Smith


Automated Inspection of Railway Assets

Contact: Professor Richard Wilson

Funder: Innovate UK/Omnicom

Visual inspection of tracks and switches is crucial to the safety and efficiency of the UK rail network. Omnicom Engineering runs vision systems mounted on the front and underneath inspection trains, which generate hundreds of miles and many terabytes of data each day. Working with Omnicom, the project uses state-of-the-art artificial intelligence techniques to automatically identify and classify faults on the rail network.


Automating Visual Inspection of Solar Photovoltaics Using Drones

Contact: Professor Richard Wilson

Funder: EPSRC

The push towards renewable energy has led to a huge increase in solar farms, both in the UK and all over the world. These farms need regular maintenance to identify failed solar cells and connectors, which reduce efficiency. This project is a collaboration with Drones on Demand who use drones mounted with infrared cameras to quickly survey the solar farms without the need to walk round on foot. Computer vision techniques are used to identify and locate the solar panels in the farm, and to automatically locate and identify faults in the cells and connectors.


Construction of a Plenoptic Imaging System for Seeing Through Turbulence

Contact: Professor Richard Wilson

Funder: DSTL

This project saw the design and construction of a light-field camera system aimed at measuring turbulence in the atmosphere and reconstructing distortion-free imagery by compensating for the effects. This involved building a high-frame-rate plenoptic system and developing the reconstruction algorithms.


Human Activity Recognition from Video Sequences

Contact: Dr Adrian Bors

Funder: DSTL

The first part of the project involved the development of an observational video analysis system that would learn the activities taking place in a given space based on their movements’ directions and intensities. This would allow for a dictionary of activities to be formed that any new activity could be compared with to identify it as a known activity or anomalous.

During the second part of the project, a classification system for human group interaction activities was developed. Human group activities display complex movement, where people either replicate movement or interact softly and violently with each other. The software being developed was applied for assessing video data showing various activities taking place outdoors by groups of four to six people.