alexgkendall / SegNet-Tutorial

Files for a tutorial to train SegNet for road scenes using the CamVid dataset
http://mi.eng.cam.ac.uk/projects/segnet/tutorial.html
847 stars 517 forks source link

How to feed one channel images to caffe-segnet #90

Open sajjo79 opened 7 years ago

sajjo79 commented 7 years ago

The code written is based on three channel input. When i try to give one channel input (gray images), it automatically converts that into 3 channel (RGB). When i change DenseImageData layer to ImageData, it gives me error in reading files. DenseImageData layer reads files successfully however the results are strange. Any help please. How can i feed one channel images. Here is the output produced by ImageData layer et: "/home/yonatan/PycharmProjects/brats-segnet/Models/segnet_train.prototxt" test_initialization: false I0326 21:29:40.120586 26288 solver.cpp:91] Creating training net from net file: /home/yonatan/PycharmProjects/brats-segnet/Models/segnet_train.prototxt [libprotobuf ERROR google/protobuf/text_format.cc:274] Error parsing text-format caffe.NetParameter: 2:1: Expected identifier. F0326 21:29:40.120620 26288 upgrade_proto.cpp:88] Check failed: ReadProtoFromTextFile(param_file, param) Failed to parse NetParameter file: /home/yonatan/PycharmProjects/brats-segnet/Models/segnet_train.prototxt Check failure stack trace: