Computer Vision

A machine vision or computer vision system is a system for measurement, inspection or surveillance, based on connecting a video camera to a computer.

Camera-based quality and production monitoring systems manage more and more complex tasks. With the advent of cheap and accurate semiconductor camera sensors and ever increasing computer power this kind of systems will be more common in every day life.

The most important task when designing a computer vision system is to ensure good quality images enhancing the effects to be analysed. SINTEF's competence covers illumination, physics, special optics, calibration and camera technology, which are crucial for good imaging.

In doing so SINTEF starts by analysing the measuring situation, to see if the physics of the problem, special optical solutions, particular illumination and imaging strategies can contribute to a better data capture.

When the images (or video) are acquired, objects in the images can be characterized by their shape, size, color, texture and porosity. Features like these can be used to distinguish objects of interest from the background or from each other. Read more about image processing and pattern recognition on their respective pages. 

SINTEF's competence covers algorithm development as well as real-time software and hardware solutions used when implementing the algorithms.

Application examples:

• Recycling technology. The green-wave slogan ”Waste is displaced resources” is certainly true. However, recycling works best where resource recovery can be made into profitable industry. The key to profitable recycling lies in reliable and efficient classification and separation technology.
• Measurements of e.g. quantity and size of products
• Inspection of products on-line or off-line
• Surveillance of e.g. traffic, people or fish
• Tracking

The weed seedlings are detected in an image of a cereal field.