Face Anonymization and Passenger Counting Solution

Summary

A global technology solutions provider sought to collaborate with a company specializing in computer vision to develop an AI solution for anonymizing faces encountered while docking and undocking aircraft. Folio3 AI crafted face anonymization and people counting solutions, which effectively blurred and anonymized all detected faces in docking videos while also accurately tallying the number of passengers offboarding from the aircraft.

About the Customer

Our client is a US-based global technology solutions provider catering to high-value sectors within the food processing and air transportation industries. They specialize in delivering cutting-edge equipment, systems, and services tailored to airports and airlines, with a keen focus on enhancing operational efficiency, safety, and customer experience.

Team composition
6 members
Expertise used

Machine Learning, Computer Vision, Vehicle Detection. Face Anonymization, Anomaly Detection
Duration

4 months
Services provided

Data cleaning and annotations, Model development and finetuning, Performance optimization for scalability, Application integration and reporting
Region

US
Industry

Aviation and Aerospace

Understanding the Challenge

Our client needed close monitoring of the aircraft docking sequence, and ensure compliance with privacy regulations during passenger ingress and egress. An intelligent system was required to detect anomalies in the docking process while also accurately counting passengers without recording their faces in the video feed. To overcome these challenges our client needed an AI solution capable of providing real-time monitoring, precise counting, and face anonymization to meet regulatory requirements and enhance operational efficiency.

Solution

Folio3 AI developed a comprehensive on-premises Face Anonymization & Counting solution tailored to the specific needs of our client's project. This solution is designed to efficiently process video feeds from multiple camera sources simultaneously, ensuring seamless integration into the existing infrastructure.

Face Anonymization

Blurred faces in video feeds using advanced face detection and anonymization techniques so that the remaining content was not compromised. Ensures compliance with privacy regulations and protects the identity of individuals captured in the footage.

Ingress and Egress Counting

Enables precise counting of both ingress and egress movements separately, allowing for detailed analysis of passenger flow.

Custom Vehicle Detection

Developed a custom feature that seamlessly integrates with the camera systems. Automatically classifies vehicles entering the designated regions, including cars, trucks, and tow trucks, providing additional insights into airport traffic and operations.

Docking Sequence Detection
Accurately detects the docking sequence, specifically ensuring that the airplane bridge is correctly attached to the airplane. Also providing real-time insights into the aircraft docking process and sending alerts on any anomalies in the docking sequence.

Customization Options
Customizable settings that can adapt to varying operational requirements and environmental conditions. This flexibility ensures optimal performance and usability across different airport settings and scenarios.

Book a Demo

Result

With an accuracy of up to 90%, the system detects any anomalies in the docking process and sends alerts for prompt actions. The system also anonymized all faces of the passengers, ensuring all compliance and regulations were met.