Red Eye

Table of Contents.

• Description.
• System Architecture.
• Hardware.
• Deep Learning Model.
• Email and Sms Notifying.
• Mobile Application.
• Topics used for implementation.
• Awards.

Description.

• This is a complete smart surveillance system which can be deployed on a Rasbperry PI 4 equipped with a Camera module. The system is able to detect Fire and weapons in real time and report the location of the threat to the relevant authorities via a serverless backend, The system can report by SMS, Emails and a Cross-platform application. with a maximum delay time of 4 seconds.

Real-Time, Neatly Build, And extenedable.

---

---

System Architecture

---

Hardware

Rasbperry PI 4, Camera Module V2.

---

Deep Learning Model

Tiny Yolo V3 CNN.

Dataset of labeled images for Guns , Rifles , and fires

---

Email and SMS notifying

Using Amazon AWS Lambda serverless cloud functions it can invoke operations on AWS SNS and AWS Dynamo DB to handle the response of the detected threeat.

---

Mobile Application

Flutter application, Dashboard data and statistical analytics via Amazon AWS DynamoDB.

Topics used for implementation.

• Python, Yolo V3, CNNs.
• Rasbperry PI configuration.
• APIs in Amazaon AWS.
• Flutter, AWS DynamoDB.

---

Awards

Egypt's First Smart Cities Hackathon first place in Smart Surveillance track and Second place overall tracks out of 478 teams.

---