EyeCity

About EyeCity - Connecting Eyes, Protecting Lives

During hackathon (which we got 2nd place at) organised by collaboration of Xebia company and EKA.NET students club, our team FlowChain was faced with new challenge, to propose a solution that could ensure safety in society. That’s how EyeCity was born, an AI empowered solution, that utilizes existent CCTV systems, using acquired information to detect wanted/lost person and notify relevant authorities about their last whereabouts. Moreover, to provide citizens with privacy, EyeCity provides face obfuscation for people that are not in scope, so it would be both easier for authorities to analyze the video and comfort providing for people concerned about their privacy. This solution was fully deployed in cloud environment, meaning it could be accesible from all zones that would utilize our application.

What is the purpose of EyeCity?

EyeCity aims to ensure people in public are safe by triggering notifications once the dangerous individual is identified, so police could investigate the area and mitigate the danger. What’s more, this system could be also utilized to make search operations more efficient as each hour passed makes it less probable to find the lost person.

Features

Identifying lost/wanted individuals

EyeCity can be integrated with databases containing pictures of lost/wanted people and extract necessary features to identify them in video streams from CCTV. Once found, system generates a notification, pointing WHO, WHERE, WHEN and HOW probable (The similarity score).

Tracking and prediction

Despite this feature wasn’t fully implemented and we didn’t have enough time and resources to properly verify it, it would allow system to track particular individual “steps” (appears on map) and show their probable path between these locations and even try predicting their next positions!

Tech-stack

• AWS
• Python
• Vue JS
• OpenCV
• ML

Authors: Patryk Wypych, Jakub Olszewski, Krzysztof Galus, Jakub Miara

© 2023 FlowChain. All rights reserved.