Python 2.6.6
Code:
- https://github.com/som4r/som4r/tree/master/robot
Status:
- 20110821 Funcionando a deteccão de faces na imagem capturada do kinect.
- 20110820 Funcionando exemplo de WSR adaptado do pyfaces.
ToDo:
- Analisar melhor maneira de armazenar as fotos (BD? FS? STM?) usadas no "treino" e na identificacão.
- Verificar melhoria do resultado da identificacão usando somente fotos da mesma pessoa.
- Testar exemplos de pyvision e facel
- Testar exemplos libface
Info:
- PyFaces (Blog)
- PyVision (Documentation)
- Libface (C/CPP)
PyFaces Install:
- sudo apt-get install tix python-imaging-tk
- download pyfaces
- cd ./pyfaces
- python pyfacescontroller.py
PyVision Install:
- svn co https://pyvision.svn.sourceforge.net/svnroot/pyvision/releases/pyvision_0.8.1 pyvision
- cd ./pyvision/src
- sudo cp -r pyvision /usr/local/lib/python2.6/dist-packages/
- ?cd ../R/libsvm?
- ?make? ?(synaptic linsvm2)?
- cd ../../src/FaceL
- python setup.py py2app
- ./FaceL.py (zzz, erro)
Links:
- Python PCA/Eigenfaces function
- FaceL: Facile Face Labeling
- CSU Face Recognition Resources (LRPCA2010 zip)
- Face recognition – much easier than expected
- RTFTR
- Person Recognition (with Python) (ver referências citadas no texto)
- The ImageChops Module
- How do I create an OpenCV image from a PIL image?
- How to open image file via HTTP and save
- cvLoadImage for in-memory images instead of files?