This work is the DRNN part of OrieNet.
platform:tensorflow
#files: fingerprit.py: structure of the network Prepare_Data.py: making tfrecord files of the network test.py: output test images' orientation fields test_FingerPrint1.py:Training process of the network m_files/getdecmatrix.m show the visualized result.
#running demo: run 'test.py' to get orientation fields. run m_files/getdecmatrix.m to get the visualized result.
#training first convert your database to tfrecord files. labels are of shape [20,20,3]. images are of [160,160], background masked.
then run test_FingerPrint1.py to train your own database. the next steps are the same with 'running demo'.