You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I'm attempting to offload image processing to workers using the native worker_threads library. The first time my application starts up a worker, runs the computation, and cleans up the worker, things work great. However, the second attempt to do so causes my whole app to crash silently. I noticed that commenting out any imports of @u4/opencv4nodejs in the worker files allows them to start and terminate as often as I like without issue. Upon investigating, I found the below resources related to context aware c++ addons.
Given the computationally intensive nature of image processing, supporting workers in order to not block the main thread seems like it should be a requirement for this library. Creating a long running subprocess is not an option for me due to memory constraints. Happy to provide any more info/help land on a solution. Thanks!
The text was updated successfully, but these errors were encountered:
Any news on this? Electron.js also dropped support for non-context-aware models, so I don't think opencv4nodejs can still run on current electron versions.
Environment
Description of Issue
I'm attempting to offload image processing to workers using the native
worker_threads
library. The first time my application starts up a worker, runs the computation, and cleans up the worker, things work great. However, the second attempt to do so causes my whole app to crash silently. I noticed that commenting out any imports of@u4/opencv4nodejs
in the worker files allows them to start and terminate as often as I like without issue. Upon investigating, I found the below resources related to context aware c++ addons.Additional Resources
Given the computationally intensive nature of image processing, supporting workers in order to not block the main thread seems like it should be a requirement for this library. Creating a long running subprocess is not an option for me due to memory constraints. Happy to provide any more info/help land on a solution. Thanks!
The text was updated successfully, but these errors were encountered: