Thanks for your answer, I can confirm that openBR can be compiled on Raspbian, i didn't use crosscompile, I compiled directly in the Raspberry following the install instructions for linux in the webpage. Anyway, regarding to the error I mentioned, I ran the AgeGenderDemo and it gives more details, here is the output:
Set algorithm to AgeGenderDemo
Warning: Qt Concurrent has caught an exception thrown from a worker thread.
This is not supported, exceptions thrown in worker threads must be
caught before control returns to Qt Concurrent.
SDK Path: /usr/local
File: thread/qthreadpool.cpp
Function: virtual void QThreadPoolThread::run()
Line: 103
terminate called after throwing an instance of 'std::bad_alloc'
what(): std::bad_alloc
El lunes, 16 de junio de 2014 20:35:58 UTC-5, Scott Klum escribió:
Tony,
I am glad to hear your were able to get things compiled on Raspian. I tried, unsuccessfully, to do the same on an Arch build. I'll probably try to do the same over the next week or so and I will let you know if I can figure out the issue.
Age and gender estimation both involve loading support vectors, but I'm not sure how expensive of an operation that is.
Regards,
Scott
Hi, I'm trying to run OpenBR on a raspberry Pi, with raspbian as OS. Installation was OK, even i ran some basic examples and works ok,but I'm getting this error:
terminate called after throwing an instance of 'std::bad_alloc' what(): std::bad_alloc
when i load the transform.
The C/C++ code that is given this error is:
QSharedPointer<br::Transform> transform = br::Transform::fromAlgorithm("FaceDetection+<FaceClassificationRegistration>+<FaceClassificationExtraction>+<GenderClassifier>+<AgeRegressor>");
Is face detection + gender + age estimation, but if I run only face detection it works ok. adding gender and age estimation produces the error.
It seems that is getting out of memory or something like that. Raspberry Pi has 512 MB RAM
Is there a minimum of RAM to get this works? or maybe the problem is not the RAM?
Does anyone has experienced in a device as raspberry pi?
Thanks.
--
You received this message because you are subscribed to the Google Groups "Open Source Biometric Recognition" group.