
Computer Vision
A computer-vision application is a software that is able to look at images and provide business logic based on what it sees. For example, you might have heard of Deep Learning, which is a type of machine-learning that is particularly well suited to classify features and objects in images. To teach a model how to distinguish cats from dogs is no challenge, not anymore now that we can use Deep Learning. How about recognizing people in images? Not a challenge, we can even detect who they are based on their face. We may even make pretty good guesses about their mood, based on their facial expression.
The changing of bodies into light, and light into bodies, is very conformable to the course of Nature, which seems delighted with transmutations. Isaac Newton
Sometimes the business logic in a computer-vision application is about bringing the image content from some camera to the end-user. It may be as aid for visually impaired students in a class room or from an operator controlling a robot. Imagine if that robot is controlled by an embedded computer with cameras plugged into it. The images it captures are to be transmitted over some network to be received on a topside application to be presented to the operator, maybe even automatically highlighting interesting visual features. To build something like that you need expertise in native software development, cloud-computing and computer-vision. We have that.
Technologies
Here is a list of some of the technologies we use for computer vision:
Types of application
We have experience in building computer-vision software that leverages feature-classification and -extraction from images, used in both drone-applications and customer recognition and payment applications.
We have experience in building applications that captures data from cameras, sends images to another location over a network where it is analysed and consumed by end-users.
We’re working with customers in the health-industry to build applications that manages and analyses image-data from medical devices.