Artificial vision systems
At this time, attempting to recreate the visual capabilities of human beings in computers is still a major challenge. However, a large toolkit of diverse techniques in machine vision has now been developed to solve specific kinds of problem. Many of these techniques are related to particular facets of human vision. For example, some techniques are able to segment different objects in an image. These techniques may then be combined with others to solve larger problems like recognising people present within a scene.
Computer vision in manufacturing
In automated manufacturing, machine vision techniques have become commonplace. They are used for automated visual inspection and quality control. In addition, visual information can be used to guide the adjustment of process control variables in order to improve the efficiency of manufacturing processes. Visual feedback has become important for robotic manipulators, providing them with information about the locations of objects and surfaces. This information can lead to significant increases in their flexibility and efficiency.
The spread of CCTV
The last few years have witnessed a tremendous expansion in CCTV surveillance systems in shops, town centres, and on public transport. They are used to search for particular persons, or to spot certain types of criminal behaviour. However, the problem is how to monitor large numbers of cameras simultaneously. There may be too many cameras for a small number of operators to monitor effectively.
One solution is the application of intelligent machine vision techniques to perform automated monitoring of CCTV images. Video analysis techniques have now been developed which can spot known persons from biometric characteristics such as gait, or spot particular kinds of event or behaviour such as a person running.
To contact us: please e-mail Dr Simon Stringer (CEO) at: