Installing Caffe

Install dependencies

sudo apt-get update && sudo apt-get upgrade

sudo apt-get install -y gfortran cython

sudo apt-get install -y libprotobuf-dev libleveldb-dev libsnappy-dev libhdf5-serial-dev protobuf-compiler git

sudo apt-get install --no-install-recommends -y libboost-all-dev

sudo apt-get install -y python-dev libgflags-dev libgoogle-glog-dev liblmdb-dev libatlas-base-dev python-skimage

sudo pip3 install pyzmq jsonschema pillow numpy scipy ipython jupyter pyyaml

Install OpenCV

To use the distribution package:

sudo apt install libopencv-dev

For a manually built version, see the OpenCV installation notes.

Clone Caffe

git clone https://github.com/BVLC/caffe

Configure Caffe

Enter its directory and copy the default configuration:

cd caffe

cp Makefile.config.example Makefile.config

nano Makefile.config

Prepare Makefile.config

For Raspberry Pi, change #CPU_ONLY := 1 to CPU_ONLY := 1.
For Python 2 with a pip-installed NumPy, replace /usr/lib/python2.7/dist-packages/numpy/core/include with /usr/local/lib/python2.7/dist-packages/numpy/core/include.

To disable Python 2, change:

   PYTHON_INCLUDE := /usr/include/python2.7 \
  		/usr/lib/python2.7/dist-packages/numpy/core/include

to:

  # PYTHON_INCLUDE := /usr/include/python2.7 \
  #		/usr/lib/python2.7/dist-packages/numpy/core/include

The leading # characters comment out Python 2 support.

For Python 3, uncomment the lines below Uncomment to use Python 3 (default is Python 2):

  PYTHON_LIBRARIES := boost_python3 python3.7m
  PYTHON_INCLUDE := /usr/include/python3.7m \
                   /usr/lib/python3.7/dist-packages/numpy/core/include

For Anaconda, find:

  # Anaconda Python distribution is quite popular. Include path:
  # Verify anaconda location, sometimes it's in root.

and uncomment the lines below it:

  ANACONDA_HOME := $(HOME)/anaconda
  PYTHON_INCLUDE := $(ANACONDA_HOME)/include \
           $(ANACONDA_HOME)/include/python2.7 \
           $(ANACONDA_HOME)/lib/python2.7/site-packages/numpy/core/include

Replace $(HOME)/anaconda with your Anaconda installation path.

Adjust the include and library variables:

INCLUDE_DIRS := $(PYTHON_INCLUDE) /usr/local/include /usr/include/hdf5/serial
LIBRARY_DIRS := $(PYTHON_LIB) /usr/local/lib /usr/lib /usr/lib/arm-linux-gnueabihf/hdf5/serial/

For a manually built OpenCV, the example uses these INCLUDE_DIRS:

INCLUDE_DIRS := $(PYTHON_INCLUDE) /usr/local/include /usr/include/hdf5/serial /usr/local/include/opencv4/

Additional configuration

These workarounds reflect errors encountered when the original notes were written. Consult the linked issues to determine whether a fix still applies.

OpenCV 4.x build error

See the Caffe issue. If the issue is unresolved in the chosen version, the original notes apply:

diff --git a/src/caffe/layers/window_data_layer.cpp b//src/caffe/layers/window_data_layer.cpp
index 1bf3760..f41169d 100644
--- a/src/caffe/layers/window_data_layer.cpp
+++ b/src/caffe/layers/window_data_layer.cpp
@@ -290,7 +290,7 @@ void WindowDataLayer<Dtype>::load_batch(Batch<Dtype>* batch) {
           image_database_cache_[window[WindowDataLayer<Dtype>::IMAGE_INDEX]];
         cv_img = DecodeDatumToCVMat(image_cached.second, true);
       } else {
-        cv_img = cv::imread(image.first, CV_LOAD_IMAGE_COLOR);
+        cv_img = cv::imread(image.first, cv::IMREAD_COLOR);
         if (!cv_img.data) {
           LOG(ERROR) << "Could not open or find file " << image.first;
           return;
diff --git a/src/caffe/test/test_io.cpp b//src/caffe/test/test_io.cpp
index c2c919e..b80df28 100644
--- a/src/caffe/test/test_io.cpp
+++ b/src/caffe/test/test_io.cpp
@@ -20,8 +20,8 @@ class IOTest : public ::testing::Test {};
 bool ReadImageToDatumReference(const string& filename, const int label,
     const int height, const int width, const bool is_color, Datum* datum) {
   cv::Mat cv_img;
-  int cv_read_flag = (is_color ? CV_LOAD_IMAGE_COLOR :
-    CV_LOAD_IMAGE_GRAYSCALE);
+  int cv_read_flag = (is_color ? cv::IMREAD_COLOR :
+    cv::IMREAD_GRAYSCALE);

   cv::Mat cv_img_origin = cv::imread(filename, cv_read_flag);
   if (!cv_img_origin.data) {
diff --git a//src/caffe/util/io.cpp b/caffe-1.0/src/caffe/util/io.cpp
index 835d2d4..941fd38 100644
--- a/src/caffe/util/io.cpp
+++ b/src/caffe/util/io.cpp
@@ -73,8 +73,8 @@ void WriteProtoToBinaryFile(const Message& proto, const char* filename) {
 cv::Mat ReadImageToCVMat(const string& filename,
     const int height, const int width, const bool is_color) {
   cv::Mat cv_img;
-  int cv_read_flag = (is_color ? CV_LOAD_IMAGE_COLOR :
-    CV_LOAD_IMAGE_GRAYSCALE);
+  int cv_read_flag = (is_color ? cv::IMREAD_COLOR :
+    cv::IMREAD_GRAYSCALE);
   cv::Mat cv_img_origin = cv::imread(filename, cv_read_flag);
   if (!cv_img_origin.data) {
     LOG(ERROR) << "Could not open or find file " << filename;
@@ -179,8 +179,8 @@ cv::Mat DecodeDatumToCVMat(const Datum& datum, bool is_color) {
   CHECK(datum.encoded()) << "Datum not encoded";
   const string& data = datum.data();
   std::vector<char> vec_data(data.c_str(), data.c_str() + data.size());
-  int cv_read_flag = (is_color ? CV_LOAD_IMAGE_COLOR :
-    CV_LOAD_IMAGE_GRAYSCALE);
+  int cv_read_flag = (is_color ? cv::IMREAD_COLOR :
+    cv::IMREAD_GRAYSCALE);
   cv_img = cv::imdecode(vec_data, cv_read_flag);
   if (!cv_img.data) {
     LOG(ERROR) << "Could not decode datum ";

C++11 error

To address a compiler-language-mode issue, change the Makefile CXXFLAGS from:

CXXFLAGS += -pthread -fPIC $(COMMON_FLAGS) $(WARNINGS)

to:

CXXFLAGS +=  -std=c++11 -pthread -fPIC $(COMMON_FLAGS) $(WARNINGS)

See the issue.

OpenCV imagecodecs error

For a manually built OpenCV 4.x, change:

LIBRARIES += opencv_core opencv_highgui opencv_imgproc
ifeq ($(OPENCV_VERSION), 3)
        LIBRARIES += opencv_imgcodecs
endif

to:

LIBRARIES += opencv_core opencv_highgui opencv_imgproc  opencv_imgcodecs

Build

Build all sources:

make all

Run tests if needed:

make test

make runtest

Build the Python bindings:

make pycaffe

python3 scripts/download_model_binary.py models/bvlc_googlenet

Install

Install the protobuf Python binding:

cd python

sudo python3 setup.py install