• DepthAI-v2

  • Problems Finding OAK-1-POE on Windows using static IP

I have an OAK-1-POE connected directly to my PC and want to connect using the device's static IP. I have followed the instructions for doing this from here:
https://docs.luxonis.com/en/latest/pages/tutorials/getting-started-with-poe/

It all works fine using Python via PyCharm - I can connect to the camera and the examples all work.

What I want to do is integrate this into our software which is built with VisualStudio and is written in C++. I have downloaded the libraries etc. for version 2.12.1 from here:
https://github.com/luxonis/depthai-core/releases

I have got it to build successfully and it runs up until I create the dai::Device passing it device_info containing the IP address and other state, platform and protocol info. However, it then throws a runtime exception with the message "Cannot find any device with given deviceInfo".

As far as I can see from the source code for DepthAI, this message is only sent from one place, inside DeviceBase::init2. However, it's inside an if then else group which it should not reach if deviceInfo.state is equal to X_LINK_BOOTLOADER. Unless I'm missing something, I have this set in my code so I can't see why it is reaching this point.

Here's my source:

dai::DeviceInfo device_info("169.254.1.222");
device_info.state = X_LINK_BOOTLOADER;
device_info.desc.platform = X_LINK_ANY_PLATFORM;
device_info.desc.protocol = X_LINK_TCP_IP;

// Create pipeline
//dai::Pipeline pipeline;
auto ppipeline = new dai::Pipeline;
auto& pipeline = *ppipeline;

// Define sources and outputs
auto camRgb = pipeline.create<dai::node::ColorCamera>();
auto nn = pipeline.create<dai::node::MobileNetDetectionNetwork>();
auto xoutRgb = pipeline.create<dai::node::XLinkOut>();
auto nnOut = pipeline.create<dai::node::XLinkOut>();

xoutRgb->setStreamName("rgb");
nnOut->setStreamName("nn");

// Properties
camRgb->setPreviewSize(300, 300);  // NN input
camRgb->setInterleaved(false);
camRgb->setFps(40);

// Define a neural network that will make predictions based on the source frames
nn->setConfidenceThreshold(0.5);
nn->setBlobPath(nnPath);
nn->setNumInferenceThreads(2);
nn->input.setBlocking(false);

// Linking
if (syncNN)
    nn->passthrough.link(xoutRgb->input);
else
    camRgb->preview.link(xoutRgb->input);

camRgb->preview.link(nn->input);
nn->out.link(nnOut->input);

// Connect to device and start pipeline
dai::Device* pDevice = nullptr;

try
{
    pDevice = new dai::Device(pipeline, device_info);
}
catch (const std::runtime_error & err)
{
    auto msg = err.what();
}

`

I can't help thinking there's something simple that I've missed or got wrong. Any assistance would be much appreciated!

Many thanks!

  • erik replied to this.

    Hello Malcs , I believe the suggested way is to connect both the host computer and the POE device to the same LAN (POE switch, or normal one + POE injector). Could you try that as well?
    Thanks, Erik

      Hi erik, thanks for your reply.

      I could try that, although my LAN is only 100Mbps so frame rate would be terrible.

      However, it all works fine with Python using the set up I have. What's odd is that it doesn't when I move to using C++. Also, the message I get should not be generated if I specify X_LINK_BOOTLOADER. So something else must surely be amiss?

      So, if I use the camera over the LAN that does indeed work. That's good - at least that's a step forward so thanks for that suggestion.

      The question now is, why doesn't it work on the C++ side when connecting directly? As I mentioned before, it works on Python when connected that way.