Different applications of Computer Vision

by | Mar 5, 2019 | Computer Vision Basics

Computer visions in all its forms!

We can distinguish two major categories of computer vision applications. First, online and offline process automation. Secondly, the development of new types of products and services such as autonomous cars, the Internet of Things (IoT), etc.

 

Offline computer vision applications

Simplified industrial inspection

When it comes to industrial inspection, computer vision allows significant time savings and a lower than average error rate. This is especially true in quality control and predictive maintenance because image recognition makes these processes much faster and efficient.

For instance, SNCF has set up a predictive maintenance technology to assess the state of its trains and rails and anticipate potential maintenance needs. This is possible through computer vision technologies. More specifically, the trains transmit data that the algorithms analyze in real-time for predictive maintenance.

 

Automated checkout

Computer vision allows the automation of the invoicing process, thus creating a fluid customer experience in stores, restaurants, etc. There is no need for barcodes! The technology can recognize each product and immediately invoice it.

Thanks to this technology, the Compass Group has significantly reduced queues in its corporate restaurants and improved the customer experience.

 

Intelligent video surveillance

Thanks to computer vision, it is now possible for companies to make their video surveillance systems intelligent. Indeed, the cameras will be able to detect unusual situations, prevent a potential accident or collect important data for process improvement.

For example, Indigo has understood the added value of these monitoring systems and has implemented them in most of its car parks to ensure better security for its users.

 

Online computer vision applications

The marketplace of the future

Marketplaces, such as second-hand markets, need to gather a number of features and photos to create their product sheets. These forms are generally completed by the users of the platform. This is why the use of computer vision is common, even necessary, to simplify and improve completion. Indeed, computer vision makes it possible to recognize the photos posted online, to ensure their consistency with the product description and to offer better referencing on the site. It is important for these marketplaces that products are well categorized and tagged to provide optimal customer experience, in part through quality filtering.

Just like product tagging, visual search has revolutionized the world of online commerce. Indeed, visual shopping has grown enormously in recent years. This new purchase method will allow you to find a product directly from a photo or to propose more precise suggestions for similar products. The potential applications of visual shopping and visual search are still limited but only need to be exploited.

 

Safer social networks

Computer vision can be used for social network moderation. Indeed, the job of moderator has proven to be psychologically challenging because of the amount of content to be processed as well as the type, often inappropriate and shocking.

This is when image recognition can help. Neural networks will learn to recognize content categorized as pornographic, violent, etc. The use of computer vision then allows for more efficient filtering and reduced effort for moderators.

 

From a daily object to an intelligent object: the power of computer vision

The autonomous car is the classic example of an object that becomes intelligent thanks to computer vision. Equipped with cameras, the self-driving car will be able to move around taking into account its environment and interacting with the other vehicles around it. Among other benefits, this technology is likely to significatly reduce the accident rate.

If you are curious to know more about the future of computer vision, don’t hesitate to consult the blog on L’Usine Nouvelle by Augustin Marty, founder and CEO of Deepomatic.

Share This