FAQ

Intenseye is an AI platform capable of connecting existing cameras in manufacturing facilities to prevent accidents. Our deep learning-powered AI models can predict and analyze body posture, ergonomics, danger zone violations, personal protective equipment usage in real-time. Based on our instant notification system facilities can stop violations before they turn into accidents.

Camera feed is processed and deleted on the fly and never stored. Our system never identifies people, nor stores identities. All the output is anonymized and aggregated and reported by our dashboard and API as visual or tabular data. We don’t rely on facial recognition, instead taking in visual cues from all features across the body. Intenseye is fully compliant with GDPR and US State laws.

Security is always our top priority. We manage our security in multiple ways.

• Camera stream security

  The data fed into Intenseye from cameras are secured by TLS protocol between client and server. The frames from cameras are never physically stored on hardware. They are stored in memory which is closed to outside access and they are deleted shortly after sending it. (Happens every in 0.1 - 0.2 ms ).

• Encryption

  To encrypt the communication internally and externally, we use the most secure encryption algorithms. Each service has it’s own identity and it’s own access rights. A service can whitelist the services which can connect to itself and validate using the central authentication mechanism.

• Server side security

  Intenseye servers use hash based algorithm to determine if the correct version of the operating system and images are used inside the machine.

• Cloud Infrastructure Security

  Services inside Intenseye infrastructure uses HTTP or RPC protocols to communicate with each other. Since the internal connection is encrypted, if someone is able to get inside the network, they will not be able to decode the message.

Good news! We are compatible with over 90% percent of the IP Cameras produced and distributed worldwide.

For more information check out our detailed brand list.

We have various endpoints for you to integrate your software or hardware into our real-time notification system.

You can check out our API documentation from here! And make sure to let us know about your clever use cases.

When the telnet response indicates that the RTSP port of the camera is accessible next step is building the RTSP URL for the system we will stream to. Since Finding RTSP URL structure is the most challenging part because it is different for each brand and model.

But don’t worry, you can find it in less than 10 minutes with a Google search. Could not find it? You can raise a support ticket to your camera provider or write to us. We’re here to help :)

Here are some examples:

• HikVision

  rtsp://admin:admin123@111.22.333.44:5001/Streaming/Channels/101/

• Bosch

  rtsp://1.222.333.44:5002/rtsp_tunnel

Once the RTSP URL is generated it can be tested with software like on VLC

You can check out the full article about how to setup and configure the RTSP stream with port forwarding from here!

1. Download VLC media player

2. Open VLC

3. Click on the File menu

4. Select Open Network

5. Paste the generated RTSP URL to the URL textbox

6. Click Open

The video stream will start in less than 10 seconds

You can check out the full article about how to setup and configure the RTSP stream with port forwarding from here!

1. Open your terminal

2. Install FFmpeg

   - brew install ffmpeg

3. To test the accessibility run:

   - ffprobe rtsp://admin:admin@111.22.333.44:5001/Streaming/Channels/101/

4. To play the video stream run:

   - ffplay -rtsp_transport tcp rtsp://1.222.333.44:5002/rtsp_tunnel