segnet 编译与测试

参考：http://sunxg13.github.io/2015/09/10/caffe/

http://m.blog.csdn.net/lemianli/article/details/76687508

http://blog.h5min.cn/u010069760/article/details/75258539

（注意：nakefile而非makefile.config）

1、编译caffe-segnet：

1.1下载caffe-segnet(适用于segnet的caffe版本，下成scaffe)

git clone https://github.com/alexgkendall/caffe-segnet

1.2更改一些编译选项：

Makefile.config:

#USE_CUDNN := 1 (scaffe不支持高版本cudnn)

INCLUDE_DIRS := $(PYTHON_INCLUDE) /usr/local/include /usr/include/hdf5/serial

LIBRARY_DIRS := $(PYTHON_LIB) /usr/local/lib /usr/lib /usr/lib/x86_64-linux-gnu/hdf5/serial

WITH_PYTHON_LAYER := 1(需要PYTHON_LAYER)

修改Makefile，在

LIBRARIES += glog gflags protobuf leveldb snappy \

lmdb boost_system hdf5_hl hdf5 m \

opencv_core opencv_highgui opencv_imgproc opencv_imgcodecs

处加入后面的opencv_imgcodecs，因为opencv3.0.0把imread相关函数放到imgcodecs.lib中了（原来是imgproc.lib

make -j8

make pycaffe

1.3、下载segnet,建议放置在caffe-segnet文件中：

git clone https://github.com/alexgkendall/SegNet-Tutorial

文件很大，因为其中包含一些图片

下载模型文件：[http://mi.eng.cam.ac.uk/~agk34/resources/SegNet/segnet_weights_driving_webdemo.caffemodel

这是用于摄像头的模型文件，不过图片也能使用,不过需要改变测试文件

使用文件中自带的图片测试结果

Example——Moudel下有相应的模型描述文件prototxt

Scripts文件夹下有相应的测试文件：*p

更改相应路径即可显示结果，这里我更改了使用上面那个摄像头模型的测试文件，可以用于测试单张图片：

# -*- coding: utf-8 -*import numpy as npimport matplotlib.pyplot as pltimport os.pathimport scipyimport argparseimport mathimport cv2import sysimport timesys.path.append('/usr/local/lib/python2.7/site-packages')# Make sure that caffe is on the python path:caffe_root = '/media/lbk/娱乐/seg-env/caffe-segnet/'sys.path.insert(0, caffe_root + 'python')import caffe# Import arguments#deploy='Example_Models/segnet_model_driving_webdemo.prototxt'#weights='Example_Moudels/segnet_weights_driving_webdemo.caffemodel'#colours='Scripts/camvid12.png'#net = caffe.Net(deploy,weights,caffe.TEST)# Import argumentsparser = argparse.ArgumentParser()parser.add_argument('--model', type=str, required=True)parser.add_argument('--weights', type=str, required=True)parser.add_argument('--colours', type=str, required=True)args = parser.parse_args()net = caffe.Net(args.model,args.weights,caffe.TEST)#caffe.set_mode_gpu()input_shape = net.blobs['data'].data.shapeoutput_shape = net.blobs['argmax'].data.shapelabel_colours = cv2.imread(args.colours).astype(np.uint8)#cv2.namedWindow("Input")#cv2.namedWindow("SegNet")cap = cv2.VideoCapture(0) # Change this to your webcam ID, or file name for your video filerval = Trueframe = cv2.imread('/media/lbk/娱乐/seg-env/caffe-segnet/segnet/Example_Models/123.png')frame = cv2.resize(frame, (input_shape[3],input_shape[2]))input_image = frame.transpose((2,0,1))# input_image = input_image[(2,1,0),:,:] # May be required, if you do not open your data with opencvinput_image = np.asarray([input_image])out = net.forward_all(data=input_image)segmentation_ind = np.squeeze(net.blobs['argmax'].data)segmentation_ind_3ch = np.resize(segmentation_ind,(3,input_shape[2],input_shape[3]))segmentation_ind_3ch = segmentation_ind_3ch.transpose(1,2,0).astype(np.uint8)segmentation_rgb = np.zeros(segmentation_ind_3ch.shape, dtype=np.uint8)cv2.LUT(segmentation_ind_3ch,label_colours,segmentation_rgb)#这里应该变成小数存储了，看来opencv对于小数也是热图显示，但是保存还是黑白的图segmentation_rgb = segmentation_rgb.astype(float)/255#cv2.imwrite('output.jpg',segmentation_rgb)#cv2.imshow("Input", frame)#cv2.imshow("SegNet", segmentation_rgb)#cv2.imwrite('output.jpg',segmentation_rgb)#这里使用plt显示与保存，比cv2好点，并且不会出现进程卡住的情况plt.imshow(segmentation_rgb)plt.savefig('output.png')plt.show()

运行：进入到SegNet-Tutorial-master文件夹

python Scripts/*.py --model Example_Models/segnet_model_driving_webdemo.prototxt --weights Example_Models/segnet_weights_driving_webdemo.caffemodel --colours Scripts/camvid12.png

即可得到结果