• DepthAI
  • Stereo Depth example code crashes abruptly

Hi,

I've a custom script to get depth of items as soon as they enter ROI . On executing the script OAKD gets disconnected after 2-3 minutes. We tried the depthai-python example script, rgb_depth_aligned.py and it throws same error but after 15 to 20 mins.

Error -
rgb_depth_aligned.py: The device has already been disconnected

We get Device not available on running any OAKD script post that.
lsusb | grep 03e7 command still shows that OAKD is connected.

HOST System Information:

  • Intel I7 12Gen
  • NVIDIA A6000
  • Ubuntu 20.04
  • DepthAI Version used- 2.15.2.0
  • DepthAI Repo tag - 2.15.2.0

Camera Model:

  • OAK-D

**Error traceback from custom script when code crashes abruptly- **

2022-04-08 10:33:51.551 | DEBUG | src.Object_depth:__init__:19 - Initializing GetObjectDepth
2022-04-08 10:33:51.554 | DEBUG | src.Object_depth:__init__:88 - StereoDepth config options:
2022-04-08 10:33:51.554 | DEBUG | src.Object_depth:__init__:89 - Resolution:(1280, 800)
2022-04-08 10:33:51.554 | DEBUG | src.Object_depth:__init__:90 - Left-Right check: 0
2022-04-08 10:33:51.554 | DEBUG | src.Object_depth:__init__:91 - Extended disparity: 0
2022-04-08 10:33:51.555 | DEBUG | src.Object_depth:__init__:92 - Subpixel: 1
2022-04-08 10:33:51.555 | DEBUG | src.Object_depth:__init__:93 - Median filtering: MedianFilter.KERNEL_7x7
2022-04-08 10:33:51.555 | INFO | src.Object_depth:__init__:94 - Creating Pipeline
Saving mesh to: ./output
[14442C1031D363D700] [25460.359] [system] [critical] Fatal error. Please report to developers. Log: 'Fatal error on MSS CPU: trap: 00, address: 00000000' '0'
Traceback (most recent call last):
File "main.py", line 171, in <module>
depth_analysis = GetObjectDepth(color_img_res)
File "/home/user/work/capture_Object_labels/tabel_depth_analysis/src/Object_depth.py", line 187, in __init__
img.get().getCvFrame()
AttributeError: 'NoneType' object has no attribute 'getCvFrame'
2022-04-08 10:34:07.968 | WARNING | __main__:exit_handler:21 - closing mq client
2022-04-08 10:34:07.969 INFO [140651989301056] ClientImpl:492 | Closing mq client with 0 producers and 0 consumers
2022-04-08 10:34:07.969 | WARNING | __main__:exit_handler:23 - mq client closed
Stack trace (most recent call last):
#20 Object "python", at 0x5581ccefc98f, in
#19 Object "/lib/x86_64-linux-gnu/libc.so.6", at 0x7fec17f4c0b2, in __libc_start_main
#18 Object "python", at 0x5581ccf540bb, in _Py_UnixMain
#17 Object "python", at 0x5581ccf53d92, in
#16 Object "python", at 0x5581ccf41366, in Py_FinalizeEx
#15 Object "python", at 0x5581ccecb91d, in PyImport_Cleanup
#14 Object "python", at 0x5581cce8c0ce, in PyDict_SetItemString
#13 Object "python", at 0x5581cce8b601, in PyDict_SetItem
#12 Object "python", at 0x5581cce4027c, in
#11 Object "python", at 0x5581cce40266, in
#10 Object "python", at 0x5581cce2f947, in
#9 Object "python", at 0x5581cce2c417, in
#8 Object "python", at 0x5581ccec9faf, in
#7 Object "python", at 0x5581cce3fac7, in
#6 Object "/home/user/anaconda3/lib/python3.7/site-packages/depthai.cpython-37m-x86_64-linux-gnu.so", at 0x7febf667517c, in
#5 Object "/home/user/anaconda3/lib/python3.7/site-packages/depthai.cpython-37m-x86_64-linux-gnu.so", at 0x7febf6690036, in
#4 Object "/home/user/anaconda3/lib/python3.7/site-packages/depthai.cpython-37m-x86_64-linux-gnu.so", at 0x7febf684c218, in dai::Device::~Device()
#3 Object "/home/user/anaconda3/lib/python3.7/site-packages/depthai.cpython-37m-x86_64-linux-gnu.so", at 0x7febf684be03, in dai::Device::~Device()
#2 Object "/home/user/anaconda3/lib/python3.7/site-packages/depthai.cpython-37m-x86_64-linux-gnu.so", at 0x7febf684f1b1, in dai::Device::closeImpl()
#1 Object "/home/user/anaconda3/lib/python3.7/site-packages/depthai.cpython-37m-x86_64-linux-gnu.so", at 0x7febf68a2f3e, in dai::DataOutputQueue::~DataOutputQueue()
#0 Object "/home/user/anaconda3/lib/python3.7/site-packages/depthai.cpython-37m-x86_64-linux-gnu.so", at 0x7febf68a0816, in
Segmentation fault (Address not mapped to object [0x2c9])
[1] 22856 segmentation fault (core dumped) python main.py

The issue looks similar to a one of the comments in forum.

Could anyone please assist me here to figure out the issue ?

  • erik replied to this.

    Hello himansingh ,
    It looks like the firmware crashes. Could you run the same code, but with the debug/trace enabled (docs here)? Hopefully, we will get additional insight that would help our firmware team debug this issue. Could you also share the script you are using (and potential AI models)?
    Thanks, Erik