使用VScode编译和调试Caffe源码，vscodecaffe

使用VScode编译和调试Caffe源码，vscodecaffe

文章目录

• 一.caffe源码编译
• 二.调试运行cpp_classification.cpp

一.caffe源码编译

• 可选修改项

DEBUG := 1
CPU_ONLY := 1
USE_OPENCV := 1

- 必须修改的项

  #找到下面两行
  INCLUDE_DIRS := $(PYTHON_INCLUDE) /usr/local/include
  LIBRARY_DIRS := $(PYTHON_LIB) /usr/local/lib /usr/lib
  #修改为:
  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 /usr/lib/x86_64-linux-
  gnu/hdf5/serial

3. sudo make all -j8(使用8个线程执行)

4. sudo make test -j8

5. sudo make runtest -j8

二.调试运行cpp_classification.cpp

./build/examples/cpp_classification/classification.bin \
  models/bvlc_reference_caffenet/deploy.prototxt \
  models/bvlc_reference_caffenet/bvlc_reference_caffenet.caffemodel \
  data/ilsvrc12/imagenet_mean.binaryproto \
  data/ilsvrc12/synset_words.txt \
  examples/images/cat.jpg

如果上面的例子执行正常，那我们下面来调试该例子。

3. 使用VSCODE打开caffe文件目录，修改launch.json
   修改下面一行内容，也就是让程序运行时去执行classification.bin

"program": "${workspaceFolder}/build/examples/cpp_classification/${fileBasenameNoExtension}.bin",
"args": ["models/bvlc_reference_caffenet/deploy.prototxt",
"models/bvlc_reference_caffenet/bvlc_reference_caffenet.caffemodel",
         "data/ilsvrc12/imagenet_mean.binaryproto",
         "data/ilsvrc12/synset_words.txt",
         "examples/images/cat.jpg"
        ],

4. classfication.cpp增加断点，开始调试运行

5. 如果编辑了classfication.cpp文件，想要重新生成怎么办？

   例如，给例子添加计时函数

double time1 = static_cast<double>( cv::getTickCount());
std::vector<Prediction> predictions = classifier.Classify(img);
double timeuse = (static_cast<double>( cv::getTickCount()) - time1)/cv::getTickFrequency();
std::cout<<"time use:"<<timeuse<<std::endl;

执行sudo make example -j8可以重新编译examples中的*.cpp，并生成*.bin执行文件