boost thread: trying joining itself: Resource deadlock avoided

Hi, every now and then, I get the following error:

terminate called after throwing an instance of 'baslerboost::exception_detail::clone_impl<baslerboost::exception_detail::error_info_injector<baslerboost::thread_resource_error> >'
  what():  boost thread: trying joining itself: Resource deadlock avoided

I haven’t figured out a way to reproduce it. I have 4 cameras running in two threads (two cameras per thread).

Any idea?

[thiesmoeller]

always catching exception and return true

yes this is what i do. decide in the exception handler if can proceed -> true if this is really a stop -> false

and output the exception contents to a logger channel

[simonaertssen]

Hi, I’m having some difficulties with subclassing the pylon.ImageEventHandler, as I cannot find an implementation with a more elaborate example than in the sample grabusinggrabloopthread.py

I have made an implementation as found below, and the grabResult is saved as an attribute of the class, which I can access from outside the class, as I wish to process the result somewhere else. I would expect some sort of signal to be available when a new grabResult is registerd (eg when OnImageGrabbed is called). How can I make this more robust?

class ImageEventHandler (pylon.ImageEventHandler):

def __init__(self, callback):
    super().__init__()
    self.returnImageCallback = callback
    self.grabbedImage = None

def OnImageGrabbed(self, camera, grabResult):
    if grabResult.GrabSucceeded():
        with grabResult.GetArrayZeroCopy() as ZCArray:
            self.grabbedImage = ZCArray
        self.returnImageCallback(self.grabbedImage)