Algorithms are part of our daily lives. And if not, how many times have you heard that the algorithm controls what we generate and consume on social networks? And in recruitment and human resources processes? That’s right, these mechanisms are also at work in this area, or do you think that when a company receives hundreds of applications to fill a job position, it reviews one by one the profiles of potential applicants?
As you can see, these are just two examples of the uses they have in our daily lives, but there are many more, or haven’t you heard the term artificial vision? This technology, how algorithms intervene and its applications in urban environments is what we are going to talk about in the following lines.
Concept of artificial vision
First of all, and so as not to put the cart before the horse, let us explain to you as simply as possible what artificial vision is. We could define this concept as a scientific discipline that allows us to collect, process and analyse images of the real world, with the aim of generating information that can later be processed by a machine. And you may ask, what is this for? So that once the algorithms based on self-learning machine learning have trained sufficiently, they are capable, on their own, of distinguishing similar images.
What are the applications of artificial vision?
As you have already seen, this is not a simple technology and constant training is required. However, the evolution is so dizzying and the training, to date, so frequent, that these algorithms, which make up neural networks, are gradually being perfected and require less and less training.
But what are they training for? What are the applications of artificial vision? Let us give you an example: imagine that the town council of Bermeo, in Biscay, needs to know the total number of people who visit the hermitage of San Juan de Gaztelugatxe every year. A great unknown that they must solve, because the protection and care of an environment catalogued as one of the greatest jewels of the Basque Country depends on it. How can they find out? By means of artificial vision, using a camera that detects the flow of people passing through a specific area at a specific time.
And no, that’s not all, because thanks to artificial vision software, Geographic Information Systems are experiencing, together with digital twins, their moment of maximum boom and evolution. Thanks to detection algorithms focused on urban infrastructures, it is also possible to carry out the recognition of lighting entities, vertical and horizontal signage, the state of pavement maintenance and urban traffic, etc.
The importance of artificial vision and the management of urban services
As you have already seen, a true Smart City philosophy requires concrete measures that favour the real protection of the environment and the efficient management of the urban resources that make up cities.
It is in this context that what we could consider as types of artificial vision become particularly relevant, and depending on the complexity of the sensors and the technology used, we can speak of:
- Vision sensors: the essential and basic element that forms part of an automated system and is used precisely to detect the movement of the entity being analysed.
- Advanced vision systems: software and hardware at a higher level of development, which facilitates the capture and processing of huge amounts of data.
- Smart cameras: devices that integrate much more complex calculation systems and whose capacity to process and analyse information remotely makes them one of the most effective solutions of the moment.
Artificial vision projects developed by Fisotec
It is precisely thanks to our 360º multi-lens camera that we use at Fisotec for the execution of urban services inventories that we can go a step further in the detection of entities.
But how do we make this possible? Currently, by using complementary detection algorithms, based on GISand positioning software, we are not only able to identify the location and type of entity we are analysing. For some time now, we have known its conservation status. In this way, we have reliable information in our projects with artificial vision to determine if the element has suffered a blow or is close to some kind of breakdown, if there is a problem of interference with any other entity on the road, etc.
All this, thanks, as we said, to our 360º multi-lens camera, which allows us to sweep all the information simultaneously, collecting images or video from various positions, something that is not possible with other cameras, as they capture information from a single place.