Image recognition technologies




Visual Engines is able to meet any need for security, research, statistical analysis and more, in multiple contexts, through the creation of visual analysis solutions based on Artificial Intelligence technologies that comply with the rules on the protection of personal data established by the Regulation EU 2016/679 and the Workers’ Statute. This important feature is ensured by the use of specialized neural networks that operate “at the edge”, ie where the image is acquired. The access and transmission of what is taken up are therefore not necessary and after the required analysis, the images can be immediately destroyed.

Here are some contexts for which Visual Engines has realized or intends to implement solutions.


The solutions in this context have the aim of increasing the safety of workers in working environments such as construction sites, industrial plants, factories, as they allow the detection in real time of anomalous situations and the quick alerting to the recipients of the accident prevention accountables , executives, RSPPs and supervisors. Using Visual Artificial Intelligence techniques, they detect in particular the correct use of Personal Protective Equipment (PPE) by workers and the control of industrial plants. The typical solution is based on a system composed of one or more smart devices, ie equipped with a microcomputer board, with on board neural networks applications.The system allows to perform a complex visual analysis and to detect if a worker wears the individual prevention devices (PPE) foreseen for carrying out the work. The device, through an internal camera, acquires images of the working environment, the neural network analyzes, checks and, if necessary, signals the lack of use of PPE to the company figures responsible for supervising workers’ safety. The analysis takes place directly on the device, without the possibility of access from outside, thus ensuring compliance with the rules on work and privacy. Visual analysis systems can also be combined with those of video surveillance to detect dangerous situations (eg man on the floor).


Visual analysis is applicable to increase the safety of citizens in cities, stations, shopping centers, public offices, institutional offices, spaces for sporting and musical events. The solutions under study intend to automatically detect and report to the competent offices, the presence of suspect, wanted individuals, anomalous behaviors and other specific situations in the individual contexts. Solutions based on neural networks for visual analysis, can be designed from scratch or integrated into the video surveillance infrastructures already existing in the various areas, processing the flows either on cloud systems or directly in the control areas. In these cases visual analysis helps the operators.


The solutions in this context intend to detect, automatically, possible situations of danger for passengers and service personnel, on public transport: trains, metro, bus, taxi, as the identification of people with suspicious behavior, wanted individuals, disputes, aggression, violence, people on the ground. The solutions can also be used to detect statistical data such as number of passengers present, seated and standing or crowding of vehicles.


Through facial recognition, people identification operations are speeded up, reducing waiting times and passenger queues at airport departure gates.With visual analysis it is also possible to monitor and control the movements of people inside the airport, for security purposes but also for statistical purposes.


In this case, these are solutions for parking and transport in general. For example, a solution is already available that allows you to know the free parking spaces by analyzing the information provided by the video sensors installed in the area. It applies to car parks arranged on public roads or open spaces, and can be integrated with existing video surveillance and parking management systems. The system is used on Lake Garda in Lazise and in the CNR research area of Pisa. Other solutions to the study concern the counting of vehicles that transit in a public road (roads and motorways), detection of stationary vehicles, slowed traffic, queue situations, anomalous behavior of means, U-turn, countermove means.


In this field of application, our solutions can provide, for example, for the prevention of ATM fraud, the control of branch accesses, the recognition of customers, the control of strangers within branches, in bank main branches, in unattended self-help desks.


In this context, the solutions provide for the control of people’s access, the monitoring of movements, the control of perimeter security, access to unauthorized areas.


In the industrial field, solutions can speed up and optimize production processes and provide for activities such as product classification and identification of defective products. Visual analysis systems are part of the National Industry 4.0 Plan of the Ministry of Economic Development.


Also in the field of sports, Visual Engines solutions can contribute both to speeding up access to systems and their security.


In the field of Retail, the application of visual analysis solutions can allow an analysis of customer behavior and customer satisfaction, and can gather information about people regarding gender, age, distance, satisfaction, permanence, passages.