When I try to set up my Oak-1 Lite Camera with DepthAI SDK module and created a code that uses both MobileNetDetectionNetwork and ObjectTracker nodes, it returns the following error:

WARNING [root.__read_device-calibration:351] Calibration data missing, using - AK-D defaults.

I would really appreciate it if anyone could suggest how to solve this problem. Thanks!

    Hi zstarr4098
    Can you add the code as well?
    I think the script should run in spite of the warning you are getting.

    Thanks,
    Jaka

        Hi zstarr4098
        Can you check depthai_sdk version and update it to the newest if possible.
        I will check with Oak-1 tomorrow at the office.

        Thanks,
        Jaka

            Hi jakaskerl
            I just checked. The SDK reports missing calibration data since OAK-1 only has one camera and doesn't need the full calibration. The example you used should run without problems even though you are receiving the warning.

            Thanks,
            Jaka

                Hello jakaskerl ,

                My depthai_sdk should be the newest as I only installed it a few days ago. Also, in terms of my program running, shouldn't it display the results of my detections? Nothing is displayed when I run this code.

                    Hi zstarr4098
                    Then this would mean it's probably a problem with opencv or pyqt. Can you try running cam_test.py (with and without -rs flag). I believe you should get a thread error from QT, in which case: https://discuss.luxonis.com/d/1832-oak-d-lr-support/4 should help you.
                    If this is not the case for you, please paste the terminal output of cam_test.py so we can debug further.

                    Thanks,
                    Jaka

                        5 days later

                        Hi zstarr4098
                        When running it without -rs flag, it showed the GUI as expected? Or was it the same as in script above - meaning no output at all as well as no error in the terminal? Seems a bit strange.

                        Thanks,
                        Jaka