Intern Frigate NVR Project – Ongoing Journey and Insights from Our Interns
We’re excited to share updates on our multi-phase intern project, where our interns are gaining hands-on experience with Linux, networking, and advanced CCTV technology. The project offers an invaluable opportunity for them to dive into open-source tools, hardware integration, and practical challenges, fostering skills that will benefit their careers in IT and surveillance industries.
Our Intern Frigate NVR Project is designed to be a continuous, collaborative experience, with each intern building upon the progress of the previous one. As we work through each milestone, we document the contributions, insights, and lessons learned by each intern. This blog will serve as a record of achievements and an inspiration for those interested in CCTV, networking, or the technical side of intern development.
For those in the CCTV or IT industry who have ideas for practical project goals, we invite you to share your suggestions with us! Contact our R&D team, and we’ll assess if it’s something achievable for our interns within their time here.
Project Scope and Constraints
To ensure uniformity and structure, this project is built within defined technical constraints:
- Hardware: We utilise our own hardware boards with ample resources for testing.
- NVR Software: The project is centered around the open-source Frigate NVR.
- Programming Language: Python is the primary language for custom scripts and integrations.
- Cellular Module: Only Telit modules are used for cellular connectivity.
- Camera Hardware: Thanks to Dahua, we’re using their cameras for the project. Other Camera Hardware is available.
- SIM Cards: All SIM cards used are from 3 Ireland – because they seem to have the best network in Ireland.
- Speakers: We’re finalising our choice for speaker hardware.
Milestones and Goals
Below is an outline of the project’s objectives. Each milestone is an essential step toward achieving a sophisticated, AI-enabled NVR solution.
Goal 1 – Hardware Setup and Frigate Installation
Our first milestone was tackled by David Roßkopf from Germany, who was tasked with setting up Frigate on our i-ctrl hardware boards, which are equipped with 4GB RAM and Intel CPUs. This setup required him to work in a Linux environment using the command line in Ubuntu. David successfully installed Frigate, connected a Dahua camera, and shared his reflections on the process.
David’s Reflections
Using Ubuntu, Docker, and the Frigate NVR software, I gained a deeper understanding of containerization, which I see as a valuable skill for my future as a software developer. Initially, accessing the Frigate web GUI was challenging, but by exploring Docker principles, I was able to resolve the issue.
David also shared an insightful suggestion to expand the NVR system by integrating IoT sensors (like temperature or wind sensors) to monitor environmental factors alongside motion detection.
The enthusiasm and fresh perspective that interns bring to projects is refreshing for the Netcelero team.
Goal 2 – Basic Movement Alarms
The next step is to define detection areas within Frigate and enable movement alarms. When movement is detected, the system should record footage and send alerts via email.
Prokop Havlik has now taken ownership of the multiphase intern project. He has progressed really well and has has it working now to send an email with an attached image when a human is detected!
This is our progress so far—this project is a work in progress! Check back here for more updates, or follow Netcelero on our LinkedIn page to stay informed on all developments.
Goal 3 – Human Detection with AI
In this milestone, we aim to detect a human entering the defined area, determine their direction of travel, and track if they are stationary or moving.
Goal 4 – Dog Detection with AI
This goal focuses on AI-based dog detection within the CCTV frame to enhance environmental monitoring.
Goal 5 – Interactive Speaker Alerts
An exciting interactive goal where, upon detecting both a human and a dog (e.g., a dog walker), the system will play pre-recorded messages. Depending on the direction of travel, the system will either welcome or thank them, adding a touch of community engagement to our surveillance system
Goal 6 – Advanced AI Detection (Dog Position)
Our aim here is to enhance AI capabilities to detect a dog in specific postures, useful for public spaces to remind dog owners that they need to clean up after their dogs.
Goal 7 – Real-Time Audio Transcription
With audio recording enabled, we plan to convert real-time audio streams to text, providing enhanced context for recorded footage.
Goal 8 – AI Sentiment Analysis
AI will analyse audio streams to detect any threatening or abusive language, adding another layer of security and awareness.
Goal 9 – Object Detection in Potential Theft Scenarios
This milestone focuses on detecting when a human places an object in a bag, which could signal theft in monitored spaces.
Further Goals and Community Involvement
We have outlined potential goals to keep the project challenging and relevant. Each milestone is carefully selected to provide interns with valuable learning experiences while creating a sophisticated NVR system. As we continue, we encourage industry professionals to share additional ideas that could be achieved within a three-week internship period. Further goals will be added as we move forward with the project.
Stay tuned to this blog for further updates on our interns’ progress, insights, and lessons learned. This project is more than just building a CCTV solution—it’s a collaborative learning journey where each intern leaves a legacy for the next.
