Find a collection of Deep Learning Use Cases and Computer Vision Applications across different industries.
A Czech Gas and Oil company is building unmanned LNG stations. Due to the inherent risks of LNG, using proper PPE like Cryogenic Gloves is essential. However, many drivers weren't using them.
Reviewing security footage was tedious and on-site checks costly, so they needed a cost-effective way to detect non-compliant drivers.
DEKI suggested using the station's existing security cameras (K1, K2, & K3) to detect Cryogenic Gloves a Face Shields via RTSP video streams. These streams were processed by detection models on an edge device for a proof of concept.
The system achieved 83% PPE detection and 82% license plate accuracy, exceeding the 80% benchmark. It offers 24/7 monitoring, enabling corrective actions for safety breaches. After the PoC's success, the client has decided to continue enhancing the system to provide proactive notifications to the drivers as well.
The Semmering Ski Resort faced the challenge of adhering to government regulations during the pandemic while remaining operational. Requiring all guests to wear FFP2 masks meant hiring additional security personnel to monitor and enforce compliance, which increased costs and lowered profits.
The FFP2 mask detection solution was seamlessly integrated into the resort's existing camera system, reducing the need for investment in additional equipment or personnel. When the system detected a guest without a mask, it triggered a sound notification, prompting the individual to put on a mask.
By implementing an automated FFP2 mask detection solution, the Semmering Ski Resort successfully adhered to government regulations while minimizing additional staffing costs. This innovative approach to mask enforcement not only enhanced the resort's ability to maintain a safe environment for all, but also demonstrated the potential for technology to address challenges in a rapidly changing world.
Despite constant advances in technology, standard motion alarm systems can be a victim of their own success in that they often generate unwanted false alarms as a consequence of an inability to accurately distinguish between intruders and other sources of movement.
Deep learning video analytics of DEKI software by default ignores video noise, waving trees, moving clouds and animals, all of which might normally be the cause of false alarms when pixel based motion detection technology or motion sensors are being used to detect activity.
By the end of the testing, our DEKI SAFE successfully detected intrusion events as a simple solution of giving existing IP cameras object recognition capabilities with the use of artificial intelligence, while saving the costs of purchase and manual installation of motion sensor systems.