The concept involves a Saas-Based CCTV surveillance platform integrating face recognition for car drivers. Users can link an IP Camera to distinguish individuals by their faces and car number plates, generating corresponding data. Notably, the system can issue alerts upon the arrival of specific individuals nearby.
A strong face recognition service through CCTV was essential to identify individuals within cars. The primary challenge was achieving quick yet accurate real-time recognition.
The client requires reports on foot traffic within the premises and the number of parked cars in the parking area. This data is then integrated with in/out reports based on car numbers.
The client's preference for a scalable yet economical AI solution led us to develop a custom face recognition feature in-house, eliminating reliance on third-party services like Amazon Rekognition.
To fulfill the requirement of tracking every inbound and outbound vehicle with timestamps, we integrated CCTV footage from entry points with computer vision. This comprehensive approach captured car details such as make, model, color, and number plate, facilitating the generation of desired reports.
An intriguing client request involved identifying potential threats and VIPs. Leveraging the existing face recognition feature, we implemented an alert system. Whenever a recognized individual's face is detected by any surveillance system, the admin receives instant notification.
83% Face recognition accuracy, including demographic breakdowns by gender and race, aligning with NIST’s Facial Recognition Vendor Test.
