Video analytics on the edge or on servers? What is better and why
Both running video analytics on cameras and on servers have their own advantages and disadvantages, and the choice largely depends on the specific use case and requirements.
Here are some factors to consider:
Running video analytics on a camera:
Pros:
- Faster processing: Running video analytics on a camera can be faster because the processing is done locally, without the need to transmit video data over a network.
- Reduced bandwidth usage: By processing the video locally, less data needs to be transmitted over the network, reducing bandwidth usage.
- Increased privacy: If the video analytics are sensitive or involve private data, running them locally on the camera can be more secure, as the data doesn’t leave the camera. And that’s the mistake — because you stream the results, not just video. The last thing you want to be taken is the results, or what we refer to as “sensitive data.”
Cons:
- Limited processing power: Cameras generally have limited processing power, which can limit the complexity of the video analytics that can be performed.
- Limited storage: Cameras also have limited storage, which can be a constraint if the video needs to be stored for later analysis.
- Difficult to scale: Running video analytics on many cameras can be difficult to scale because each camera needs to be individually configured and managed.
Running video analytics on a server:
Pros:
- More processing power: Servers generally have more processing power than cameras, which can enable more complex and sophisticated video analytics.
- More storage: Servers can also have more storage capacity than cameras, which can enable longer-term storage of video data for analysis.
- Easier to scale: Running video analytics on a server can be easier to scale because new cameras can be added to the network without needing to configure each one individually.
Cons:
- Higher bandwidth usage: Running video analytics on a server requires video data to be transmitted over the network, which can increase bandwidth usage.
- Increased latency: Processing video on a server introduces additional latency because the data needs to be transmitted to the server and the results need to be transmitted back to the camera.
In summary, the choice between running video analytics on cameras or servers depends on factors such as the desired complexity of the analytics, the processing power and storage requirements, and security and privacy considerations. At issivs.com, we specialise in designing and supporting complex video analytics projects, and can help you make the right choice for your specific needs.
