septiembre 18, 2022

Artificial vision applied to video surveillance

Artificial vision applied to video surveillance

Vision Analytics has solutions that can monitor 24 hours a day some of the different suspicious behaviors with the use of machine vision. Henry Navarro is the founder of Vision Analytics and a Machine Learning engineer. Unlike other methods such as the use of sensors, machine vision is even able to anticipate human behaviors such as detecting suspicious behavior.

One of the biggest problems in the European trucking industry is fuel theft, which generates losses of millions of euros each year for the sector. Using machine vision and a simple camera installed on the trucks, Vision Analytics is able to detect this behavior almost immediately and send an alert via an application to both the driver and the company responsible.

To do this, a neural network has been trained which, using thousands of real images collected by our team of partners and synthetic data from 3D simulations, can detect these behaviors with high accuracy.

Machine vision in the healthcare sector

An important client of Vision Analytics in the healthcare sector suffered theft of different high-value medical equipment. In the client’s recordings, the alleged criminals followed the same pattern: they tried to cover their faces with caps, hoodies, hats or even carried a backpack, bag or suitcase that would allow them to store the object to be stolen. Together with partner Hikvision, an artificial intelligence capable of detecting these objects, as well as people with their faces covered with a cap or hat, was proposed.

Since the implementation of the solution, the thefts of medical equipment have been reduced in their entirety since, unlike other sensor-based solutions, artificial vision allows alerts to be sent to security officers even before the theft occurs, thus demonstrating the importance of this type of solution in the video surveillance sector.