With input from the team at advert tracking service Adclip, the media update team looks at image recognition technology powered by AI and how it can help brands identify when, and where, their imagery is used.

Image recognition technology has become more valuable, and necessary, as the use of visuals continue to increase in digital publications and social media. A SmartInsights article cited that every minute in 2016, 3.3 million Facebook posts were uploaded, 500 hours of video was uploaded to YouTube, and 65 972 photos were posted onto Instagram.

A few decades ago, sticking film photos into a album was a quick process. Now that images are cheaper to print and mostly displayed digitally, hundreds of thousands of them are being created every second.

The massive quantities of visual data being generated creates the need for automated tools to organise images, categorise them for quick access, and extract high-level information about the objects pictured in them.

To do so efficiently, electronic devices and computers have to be able recognise the items in the visual. This is where image recognition, and a number of other methods like pattern recognition and visual grouping, comes in.

What is image recognition?

Service providers like Adclip use what the public often refers to as image recognition. This technology identifies the logos of thousands of brands and businesses in newspapers and magazines. This information is used to notify clients where and when their adverts, or those of their competitors, have been published.

“The Adclip team uses image recognition to detect objects in images and compare these items to named images in our database,” explains Catherine Dabbs, a representative of Newsclip. The company offers ad tracking solutions through its Adclip service.

“In the process, image recognition technology detects signs, such as company or brand logos, with images, and determines the brand it represents.”

Decades before image recognition was used to identify highly detailed visuals like logos in images, the technology had more basic, yet revolutionary, applications – like barcode readers. This innovation was designed to transform an image of a barcode into a set of numbers, or basically: data.

Fast forward to today and image recognition, together with a number of other AI technologies, has evolved into computer vision. It allows computers to understand what they ‘see’, which is why it is being used in self-driving cars, augmented reality, robots, and various other applications.

How image recognition is being used in media intelligence

Image recognition and, more specifically, computer vision have been helping businesses do things more efficiently. One way it has been doing so is by helping brands keep track of their media advertising.

Image recognition has extensive application in the media intelligence sector. It can be used to recognise logos in publications and even detect signage in videos and television broadcasts.”

Dabbs notes that Adclip uses image recognition to find brand logos in publications. “This allows our clients to verify that their adverts were placed correctly. Adclip’s Ad Tracker also uses the technology to let brands track competitor advertising in publications.”

“This helps businesses keep up to speed with news in their industries, whether its new launches, price benchmarks, competitor strategies, or market changes.”

In highly contested marketplaces, companies need to stay ahead of their competitors in all aspects, including advertising. The advances in AI technology, and its application in image recognition, makes it easier for businesses to track their adverts and those of other players in the market.