Building a Network Video Recorder with Nvidia Jetson Nano
As a seasoned photographer and technical expert, I’ve always been fascinated by the intersection of technology and creativity. In this article, I’ll share my experience in building a network video recorder (NVR) using the Nvidia Jetson Nano, a powerful and affordable AI-powered computer.
Why Nvidia Jetson Nano?
I purchased the Nvidia Jetson Nano for $99, a tiny circuit board that packs a punch. Its power consumption is minimal, making it easy to carry around, and its integrated GPU AI capabilities far surpass most large NAS devices. With the Nano preloaded with a large number of AI tools, it’s an ideal choice for those who want to play with object recognition.
Step 1: Configure Your Nano
Configuring the Nvidia Nano is a breeze. All you need is a microSD card and a computer to refresh the L4T (Linux For Tegra) image. I used a 64GB high-speed card, which occupied most of the system. Once connected to a keyboard, mouse, and monitor, insert a mini-USB power, and you’ll be running Ubuntu 18.04 in no time. The Nano comes with a variety of AI tools and demonstration applications preinstalled.
Step 2: Connect Your Nano
The Nvidia Nano comes with a CPU, GPU, RAM, and multiple I/O options, including an Ethernet port. However, if you want to use Wi-Fi, you’ll need to add it via an M.2 slot or one of the four USB ports. For heavy calculations, use the appropriate safety barrel connector, and make sure to connect a fan for cooling. Nvidia provides links to compatible components on its Jetson Nano Developer website.
Step 3: Pick Your NVR Software
Since the Nvidia Nano requires software that can run on Linux and ARM processors, I opted for ZoneMinder, a free and out-of-the-box solution. ZoneMinder is easy to install and has a native Web UI, allowing you to manage your NVR from any location in the network. However, I found the user interface to be not very intuitive.
Step 4: Pick Your Camera
When selecting a camera, I chose the inexpensive but powerful Honic 4K Outdoor PoE IP camera for $80. This camera supports open RTSP and ONVIF protocols, allowing you to use it with ZoneMinder without the need for a cloud subscription.
Step 5: Configure Your NVR
To test your camera with ZoneMinder, add it as a monitor and configure the “monitoring” operation. To start recording a video, change the camera to “Mocord” or “Modect,” which enables motion detection in a specified area. I strongly recommend continuous recording, where all content is recorded, and prominent sports events are highlighted. This allows you to fast-forward to see sports events detected by the recording, which is really helpful.
Conclusion
Building a network video recorder with Nvidia Jetson Nano is a fun and rewarding experience. With its powerful AI capabilities and affordable price, it’s an ideal choice for those who want to play with object recognition and build a custom NVR solution. By following these steps, you can create a robust and portable video surveillance system that meets your needs.
Recommendations
- Use a high-speed microSD card to ensure smooth performance.
- Opt for a 64GB or larger card to accommodate the system.
- Choose a camera that supports open RTSP and ONVIF protocols.
- Consider using a fan for cooling, especially for heavy calculations.
- Use ZoneMinder or other Linux-based NVR software to manage your system.
- Experiment with custom AI features to enhance your NVR solution.
Future Development
The Nvidia Jetson Nano is a versatile device that can be used for various applications, including AI-powered video surveillance, object recognition, and more. With its powerful GPU AI capabilities and affordable price, it’s an ideal choice for developers and hobbyists who want to explore the world of AI and computer vision.