Artificial vision as a safety measure for driving safety
Artificial intelligence applied to automobiles is on everyone's lips: every day, hundreds of news items appear about amazing autonomous cars driving around our cities, leading us to imagine a revolutionary future.
Probably this projection is due to memories of our childhood, where movies full of robots capable of interacting with humans made us dream of a fantastic future. In fact everyone is fascinated by the idea of machines capable of doing our work, helping us at home or what is already a reality: driving for us.
But the reality is much simpler and more complex at the same time: "machine vision" is just an expression referring to algorithms capable of recognizing patterns within an image.
In other words, machine vision devices used by the automotive industry base their technology on their ability to recognize patterns of "pedestrians", "cyclists" and other vehicles.
Machine vision is based on recognizing patterns of pedestrians, cyclists and other actors involved in driving.
Although it sounds simple, it is important to consider that these devices never get tired or distracted: that is their main advantage. Statistics reveal that 80% of car accidents are due to driver distractions. In this sense, it is evident how an artificial vision device can help us in those cases in which distractions occur while driving and we fail to detect a dangerous situation in time.
When the machine surpasses human capacity
In addition, in recent years, artificial vision has come to surpass human capabilities thanks to the latest generation of sensors: there are in fact cameras such as the thermal imaging cameras that can see through smoke, fog and snow.
The Lidar cameras measure with a very small margin of error the distance between objects.
In addition, we are all used to the images from infrared cameras that see in complete darkness.
