I am running the main.py script from the depthai-expirements/gen2-pointcloud/device-pointcloud. I amh have a error that I have not seen beofre.s:\DEPT\SVM4\Shared\Crossfunctional_Work\Projects\DepthCameras\LuxonisDepthAI\test_run\depthai-experiments\gen2-pointcloud\device-pointcloud\main.py:153: DeprecationWarning: RIGHT is deprecated, use CAM_C or address camera by name instead.

M_right = calibData.getCameraIntrinsics(dai.CameraBoardSocket.RIGHT,

Stack trace (most recent call last):

#26 Object "", at 00007FFAA3EC7344, in BaseThreadInitThunk

#25 Object "", at 00007FF7F41E1490, in OPENSSL_Applink

#24 Object "", at 00007FFA4C1C493A, in Py_Main

#23 Object "", at 00007FFA4C1C4552, in Py_gitidentifier

#22 Object "", at 00007FFA4C1C3ADA, in Py_gitidentifier

#21 Object "", at 00007FFA4C3C72E5, in PyRun_SimpleFileObject

#20 Object "", at 00007FFA4C3CA398, in PyRun_StringFlags

#19 Object "", at 00007FFA4C3CA7C8, in PyRun_FileExFlags

#18 Object "", at 00007FFA4C3CA6FE, in PyRun_FileExFlags

#17 Object "", at 00007FFA4C3481F1, in PyEval_EvalCode

#15 Object "", at 00007FFA4C246DB1, in PyObject_Vectorcall

#15 Object "", at 00007FFA4C246DB1, in PyObject_Vectorcall

#14 Object "", at 00007FFA4C246B4A, in PyObject_MakeTpCall

Attempt to access invalid address.

#13 Object "", at 00007FF9B690570A, in ??

#12 Object "", at 00007FFA4C2AB23F, in PyType_Name

#11 Object "", at 00007FFA4C2B7879, in PyType_Ready

#10 Object "", at 00007FFA4C246E38, in PyObject_Call

#9 Object "", at 00007FFA4C246D1D, in PyVectorcall_Function

#8 Object "", at 00007FFA4C24984A, in PyMethod_Self

#7 Object "", at 00007FFA4C246B4A, in PyObject_MakeTpCall

#6 Object "", at 00007FFA4C28E459, in PyCFunction_GetFlags

Attempt to access invalid address.

#5 Object "", at 00007FF9B6909DCE, in ??

#4 Object "", at 00007FF9B6B8D5E4, in librealsense::ptr_option<int>::is_enabled

#3 Object "", at 00007FF9B6B8D521, in librealsense::ptr_option<int>::is_enabled

#2 Object "", at 00007FF9B6B4B1AD, in librealsense::ptr_option<int>::is_enabled

#1 Object "", at 00007FF9B6A8B7BF, in librealsense::frame::unpublish

The specified module could not be found.

#0 Object "", at 0000000000000000, in ??

Does this relates to memory or specific modules? I see that this is output to the screen: The specified module could not be found.

    jakaskerl
    I currently have numpy version 2.0.0 with open3d version 0.18.0. I believe I had installed numpy version 1.26.4 and it the point cloud worked. So I would assume that I need to downgrade from the updated version. However can you explain why some of the point cloud scripts do not work, is it because of module compatibility?

        jakaskerl
        Okay that worked and it allowed me to run the point cloud script but a stack trace message printed to the terminal. I am somewhat curious on what it means. Here is the message:
        Starting pipeline

        S:\DEPT\SVM4\Shared\Crossfunctional_Work\Projects\DepthCameras\LuxonisDepthAI\test_run\main.py:215: RuntimeWarning: invalid value encountered in cast

        depth = (disp_levels * baseline * focal / disp).astype(np.uint16)

        Stack trace (most recent call last):

        #18 Object "", at 00007FFB22D77374, in BaseThreadInitThunk

        #17 Object "", at 00007FF7F25F1490, in OPENSSL_Applink

        #16 Object "", at 00007FFAD2BC493A, in Py_Main

        #15 Object "", at 00007FFAD2BC4552, in Py_gitidentifier

        #14 Object "", at 00007FFAD2BC3ADA, in Py_gitidentifier

        #13 Object "", at 00007FFAD2DC72E5, in PyRun_SimpleFileObject

        #12 Object "", at 00007FFAD2DCA398, in PyRun_StringFlags

        #11 Object "", at 00007FFAD2DCA7C8, in PyRun_FileExFlags

        #10 Object "", at 00007FFAD2DCA6FE, in PyRun_FileExFlags

        #9 Object "", at 00007FFAD2D481F1, in PyEval_EvalCode

        #8 Object "", at 00007FFAD2D4CF43, in PyEval_EvalFrameDefault

        #7 Object "", at 00007FFAD2C46DB1, in PyObject_Vectorcall

        #6 Object "", at 00007FFAD2C46B4A, in PyObject_MakeTpCall

        #5 Object "", at 00007FFAD2C8E459, in PyCFunction_GetFlags

        Attempt to access invalid address.

        #4 Object "", at 00007FFA32C39DCE, in ??

        #3 Object "", at 00007FFA32FD8204, in PyInit_pybind

        #2 Object "", at 00007FFA32FD7FF9, in PyInit_pybind

        #1 Object "", at 00007FFA32E1207B, in librealsense::ptr_option<int>::is_enabled

        The specified module could not be found.

        #0 Object "", at 0000000000000000, in ??

          @gdeanrexroth it's in the error message: invalid value encountered in cast (likely casting INF to int16). GPT4 is usually helpful with such logs.

            erik

            Okay, I am not sure how to fix this error as I am still learning about the camera and its functionality. However whenever I ran my python debugger, there was an issue at the at this line; depth = value.astype(np.uint16)

            Here is the repo link that i got the code from :cupoch/examples/python/depthai/depthai_pcd_viewer.py at main · neka-nat/cupoch · GitHub

            On another note, I have several other pointcloud scripts and all of them result with the same error as above. I did not changed anything at all, I even copied and paste the same the code that is from the luxonis documentation page about pointcloud. That also printed about the same exact Stack Trace.

              Hi @gdeanrexroth , this is more relevant to numpy (conversion) than to depthai/camera itself. My suggestion would be to check what values are in value, as casting those values to uint16 results in error.

              erik
              UPDATE
              For some reason my depthai_sdk was not installed. So I ran the command "python3 -m pip install --upgrade depthai-sdk". I then ran the scripts and the errors above are gone. I think due to my job proxy, I am still unable to run the depthai viewer tool at it full ability. However I did run the script,

              from depthai_sdk import OakCamera

              with OakCamera() as oak:

              color = oakcreate_.camera('color')

              stereo = oak.create_stereo()

              stereo.config_stereo(align=color)

              pcl = oak.create_pointcloud(depth_input=stereo, colorize=color)

              oak.visualize(pcl, visualizer='depthai-viewer')

              oak.start(blocking=True)

              Here is the output of that script

              Is there for me to display or generate better pointcloud to show the image on the right side of the screen?