Is Computer Vision the Way Forward to Deliver Commercial Value?
Article submitted by Montse Llort, IoT & Analytics Specialist at Tech Data
Whether you are in retail, manufacturing or just running offices, having a solution to help you understand your environment is crucial for many reasons. It allows you to make decisions based on what can be seen, helping you to get to know your customers better and their habits; keeping an area safe or even ensuring compliance when it comes to rules such as wearing face masks.
This is where computer vision is can truly add value. Without even being noticed, it is enhancing our lives. It is now seen as one of the main technologies that enables the digital world to interact with the physical world. By using digital imagery, machines can accurately identify and classify objects and react to what they see. For a more in-depth understanding of computer vision, read this: Introduction to Computer Vision.
Accuracy rates for object identification and classification have jumped from 50% a decade ago to 99% by 2019, and today’s systems can even be more accurate than humans. (Source: TechSee)
Computer vision started in the 1950s when neural networks began to detect the edges of objects and sorted them into shapes. As the internet matured in the 1990s, large sets of images became available online for analysis, driving the development of facial recognition programs. Today, computer vision technology has become easily accessible, making it more appealing to enterprises. The real-life use cases of computer vision now mean it can provide actionable insights for many businesses.
This real-life commercial imperative gained by using computer vision modelling has resulted in the growth of a number of specialist companies creating computer vision modelling to drive commercial agility within businesses. Organisations are benefiting from an increase pool of specialist companies in computer vision modelling, enhanced further by specialising in vertical markets as well. For example, if you are looking for a specialist provider for warehouse intelligence like automated pallet counting, there will be specialists that will provide this.
There are a multitude of available commercial applications including next gen footfall; PPE detection, warehouse intelligence, customer waiting, suspicious person detection, loss prevention, social safety, fire detection, face mask detection, age and gender demographics, traffic management and access control to name a few, not forgetting organisations like Tesla who are really pushing the boundaries with vision AI technology.
Computer vision applications can be home grown by your own data scientists, but to get the best ROI on computer vision to provide the actionable insight you need, this will be achieved either through a suite of developed applications that make sense of the data you are collecting or through working with companies that can truly train your computer vision models using techniques called synthetic modelling.
So, is computer vision the way forward? According to this report, the computer vision market is expected to reach $48.6 billion by 2022, making it an extremely promising UX technology (Source: XD Adobe).
