Computer Vision

Computer Vision is a field of AI dedicated to understanding the visual world. Through the use of images, video and deep learning models, software/applications can be trained to recognise and classify objects. Once an object has been identified, it is possible to follow up with an automatic action or make a recommendation. Computer Vision will typically form part of a total solution in which other technologies such as Machine Learning or Advanced Analytics can also be used.

If you choose to have a solution designed by us, we will use a number of computer science and artificial intelligence techniques and tools to ensure the best possible solution to your needs. Our technical expertise includes areas such as Advanced Analytics, Machine Learning, Natural Language Processing (NLP) and Computer Vision, and we will often combine these techniques to achieve the most effective solution.

Computer Vision - examples of use

Image classification

Automatic image classification makes it easy for you to find out what an image contains.

This is useful if you need to sort many items based on image content. The process can involve everything from clustering data to converting images to text.

Fault detection

Automatic fault detection allows you to pick up on abnormalities.

It can be used on a production line, first to spot faults, and then to analyse data to discover the cause of the faults.

Automatic quality control can, for one, free up time spent on trivial tasks, and in some cases even be far more accurate than the human eye.

Object localisation + tracking in 3D space

By locating an object and following it around an area, e.g. people at a metro station or in a supermarket, it is possible to follow up with various analyses with a view to flow optimisation, optimal staffing of checkouts, etc.

Optical character recognition

Teaching a piece of software how to read makes it possible to automate simple processes where, for example, packages are marked with a specific code, or you want to be able to detect and read number plates or automatically extract key figures from various documents.

Object counting:

After identifying an object, it may be useful to be able to count it, e.g. in traffic solutions.

Once a car has been recognised, and the direction it is travelling in determined, it is also possible to tell how many free parking spaces there are in an area.

Graphic display of how we use Computer Vision and x-rays for quality control of fuel cells.

Visual quality control – something you spend time on?

If your company produces a large volume of products that have to be quality controlled, it is worth considering help from Computer Vision technology.

Computer Vision solutions consist of a camera solution and software models trained to know what a product should look like and what faults/defects look like. The solution can also collect data that can be worked on and that can be used to determine the causes of faults/defects.

Computer Vision Demo

Going to exhibitions, we often bring a small demo – and what can create more awareness about our competences and start a good dialogue than a Computer Vision demo with a twist.