Computer Vision: What It Is and Everyday Examples [Easy Guide]

Computer vision sounds like something straight out of a science fiction film, and yet it surrounds us in the most ordinary parts of our day. You benefit from it every time you record an Instagram video with a filter that smooths away your under-eye circles, or when you leave a car park without inserting a ticket because the barrier lifts as you approach.

What is computer vision?

It is a subcategory of artificial intelligence focused on developing techniques that help machines understand what they see and act accordingly. It works in a way that is a lot like the human body: a camera acts as the eye, software as the brain and a mechanism as the arm. That is why, when your car approaches a car park barrier, a camera reads the licence plate, the software checks that you have paid, and the barrier lifts.

Computer vision took its first steps back in the 1950s, but it was in the 1970s that it really started to help us, getting a computer to "read" a text and turn it into spoken narration. Until recently it worked in fairly limited ways; it has been in the last decade that the big boom has happened, with advances in optics that have raised identification accuracy from 50% to 99%, to the point of surpassing humans at detecting and classifying objects.

How does computer vision work?

It is a machine learning discipline that is already widely adopted in business, because it reduces workloads and increases control over tasks. In manufacturing, factories use it to spot production defects in real time. Picture a winery: between the filling and the sealing of each bottle, a computer vision system is installed. Each bottle trips a sensor as it leaves the filler, a light switches on, a camera takes a photo, and software analyses the image to decide whether the fill level is correct; if it is, the process continues and the bottle is corked; if not, the bottle is automatically diverted and the operator is alerted. An extra set of eyes making sure everything goes right, eyes that can be chained together across the entire manufacturing process.

Current examples of computer vision

It is already part of our reality in countless applications: the facial-recognition filters on Instagram, Snapchat or TikTok; the digitisation of restaurant menus with QR codes; the scanning of labels in retail to check stock; the "hawk-eye" that tracks the path of the ball in tennis or football; apps like Google Lens that turn handwritten notes into editable documents; or the sorting of food by size, shape and colour in the food industry.

Future applications

Over the coming years, computer vision will keep improving our lives: in autonomous driving, interpreting information more deeply (if a ball rolls in front of the car, it will know it should brake because a child is probably right behind it); in smart cities, tracking available parking spaces in real time; and in tourism, where pairing it with augmented reality will let you see a building's history overlaid simply by looking at it through your smartphone.

The advantages of computer vision

Applied to production systems, its three main advantages are faster and simpler processes, monotonous tasks run at a higher pace and operators focus on what adds the most value, better products, a well-trained system is almost infallible, shortens timelines and raises quality, and lower costs, because there is no money to spend fixing errors that simply never happen.

Computer vision is already letting many companies stop worrying about their processes and focus on what truly matters: growing. At Neurafy we take care of everything, from 0 to 100: consulting, design, build and rollout. Discover the power of having an infallible extra set of eyes.