调研 May 2020
历史记载
OpenCV4.0貌似GPU加速有点小进步。
Ref: How to use OpenCV’s “dnn” module with NVIDIA GPUs, CUDA, and cuDNN [实操]
Ref: OpenCV ‘dnn’ with NVIDIA GPUs: 1549% faster YOLO, SSD, and Mask R-CNN
Official: https://opencv.org/opencv-4-0/
dnn module was updated with Deep Learning Deployment Toolkit from the OpenVINO™ toolkit R4. See the guide how to build and use OpenCV with DLDT support. dnn module now includes experimental Vulkan backend and supports networks in ONNX format.
一、过去
However, the biggest problem with OpenCV’s dnn module was a lack of NVIDIA GPU/CUDA support
Led by dlib’s Davis King, and implemented by Yashas Samaga, OpenCV 4.2 now supports NVIDIA GPUs for inference using OpenCV’s dnn module, improving inference speed by up to 1549%!
删除老残穷版本：
cd ./build sudo make uninstall sudo rm `find /usr/ ! -path "/usr/local/anaconda*" ! -path "/usr/local/opencv*" -name "*opencv*"` -rf
二、现在
- C++版本
You can then run the following cmake command, making sure you set the CUDA_ARCH_BIN variable based on your NVIDIA GPU architecture version, which you found in the previous section:
$ cmake -D CMAKE_BUILD_TYPE=RELEASE \ -D CMAKE_INSTALL_PREFIX=/usr/local \ -D INSTALL_PYTHON_EXAMPLES=ON \ -D INSTALL_C_EXAMPLES=OFF \ -D OPENCV_ENABLE_NONFREE=ON \ -D WITH_CUDA=ON \ -D WITH_CUDNN=ON \ -D OPENCV_DNN_CUDA=ON \ -D ENABLE_FAST_MATH=1 \ -D CUDA_FAST_MATH=1 \ -D CUDA_ARCH_BIN=7.5 \ -D WITH_CUBLAS=1 \ -D OPENCV_EXTRA_MODULES_PATH=/usr/local/opencv/opencv_contrib/modules \ -D HAVE_opencv_python3=ON \ -D PYTHON_EXECUTABLE=/usr/local/anaconda3/bin/python \ -D BUILD_EXAMPLES=ON ..
NVIDIA GPU architecture version.
GeForce MX250: First Turing GPU for Laptops?
#include "stdafx.h" #include <iostream> #include "opencv2/opencv_modules.hpp" #include "opencv2/core.hpp" #include "opencv2/features2d.hpp" #include "opencv2/highgui.hpp" #include "opencv2/cudafeatures2d.hpp" #include "opencv2/xfeatures2d/cuda.hpp" using namespace std; using namespace cv; using namespace cv::cuda; int main() { //Test the graphics card method 1 (this method can read the graphics card model) printShortCudaDeviceInfo(getDevice()); //Test graphics card method 2 int iDevicesNum = getCudaEnabledDeviceCount(); cout << "Devices Num:"<<iDevicesNum << endl; //Test graphics card method 3 DeviceInfo _deviceInfo; bool _isDeviceOK = _deviceInfo.isCompatible(); cout << "IsGPUDeviceOK : " << _isDeviceOK << endl; return 0; }
Install OpenCV 4 on Ubuntu 18.04 (C++ and Python) [安装依赖包可以借鉴]
How to install OpenCV 4.2 with CUDA 10.0 in Ubuntu 18.04 [看上去不错的样子]
OpenCV使能CUDA加速 [感动的要哭了，更好一些，下面那个更好]
配置成功的话，编译中能看到在编译.cu文件。
[ 20%] Built target opencv_calib3d
[ 20%] Built target opencv_version
[ 20%] Built target opencv_imgcodecs
[ 20%] Built target opencv_quality
[ 20%] Building NVCC (Device) object modules/cudafeatures2d/CMakeFiles/cuda_compile_1.dir/src/cuda/cuda_compile_1_generated_orb.cu.o [ 20%] Building NVCC (Device) object modules/cudaimgproc/CMakeFiles/cuda_compile_1.dir/src/cuda/cuda_compile_1_generated_mean_shift.cu.o [ 20%] Building NVCC (Device) object modules/cudaimgproc/CMakeFiles/cuda_compile_1.dir/src/cuda/cuda_compile_1_generated_bilateral_filter.cu.o [ 20%] Building NVCC (Device) object modules/cudafeatures2d/CMakeFiles/cuda_compile_1.dir/src/cuda/cuda_compile_1_generated_bf_knnmatch.cu.o [ 20%] Building NVCC (Device) object modules/cudaimgproc/CMakeFiles/cuda_compile_1.dir/src/cuda/cuda_compile_1_generated_blend.cu.o [ 20%] Building NVCC (Device) object modules/cudaimgproc/CMakeFiles/cuda_compile_1.dir/src/cuda/cuda_compile_1_generated_build_point_list.cu.o [ 20%] Building NVCC (Device) object modules/cudaimgproc/CMakeFiles/cuda_compile_1.dir/src/cuda/cuda_compile_1_generated_canny.cu.o [ 20%] Building NVCC (Device) object modules/cudaimgproc/CMakeFiles/cuda_compile_1.dir/src/cuda/cuda_compile_1_generated_clahe.cu.o [ 20%] Building NVCC (Device) object modules/cudaimgproc/CMakeFiles/cuda_compile_1.dir/src/cuda/cuda_compile_1_generated_color.cu.o [ 20%] Building NVCC (Device) object modules/cudafeatures2d/CMakeFiles/cuda_compile_1.dir/src/cuda/cuda_compile_1_generated_bf_match.cu.o [ 20%] Building NVCC (Device) object modules/cudafeatures2d/CMakeFiles/cuda_compile_1.dir/src/cuda/cuda_compile_1_generated_bf_radius_match.cu.o [ 20%] Building NVCC (Device) object modules/cudaimgproc/CMakeFiles/cuda_compile_1.dir/src/cuda/cuda_compile_1_generated_corners.cu.o [ 20%] Building NVCC (Device) object modules/cudaimgproc/CMakeFiles/cuda_compile_1.dir/src/cuda/cuda_compile_1_generated_debayer.cu.o [ 20%] Building NVCC (Device) object modules/cudaimgproc/CMakeFiles/cuda_compile_1.dir/src/cuda/cuda_compile_1_generated_generalized_hough.cu.o [ 20%] Building NVCC (Device) object modules/cudafeatures2d/CMakeFiles/cuda_compile_1.dir/src/cuda/cuda_compile_1_generated_fast.cu.o
Scanning dependencies of target opencv_cudafeatures2d
[ 20%] Building CXX object modules/cudafeatures2d/CMakeFiles/opencv_cudafeatures2d.dir/src/brute_force_matcher.cpp.o
[ 20%] Building CXX object modules/cudafeatures2d/CMakeFiles/opencv_cudafeatures2d.dir/src/fast.cpp.o
[ 20%] Building CXX object modules/cudafeatures2d/CMakeFiles/opencv_cudafeatures2d.dir/src/feature2d_async.cpp.o
[ 20%] Building CXX object modules/cudafeatures2d/CMakeFiles/opencv_cudafeatures2d.dir/src/orb.cpp.o
[ 20%] Building NVCC (Device) object modules/cudaimgproc/CMakeFiles/cuda_compile_1.dir/src/cuda/cuda_compile_1_generated_gftt.cu.o
[ 20%] Linking CXX shared library ../../lib/libopencv_cudafeatures2d.so
[ 20%] Built target opencv_cudafeatures2d
[ 20%] Building NVCC (Device) object modules/cudaimgproc/CMakeFiles/cuda_compile_1.dir/src/cuda/cuda_compile_1_generated_hist.cu.o
[ 20%] Built target opencv_cudastereo
Scanning dependencies of target opencv_datasets
[ 20%] Building CXX object modules/datasets/CMakeFiles/opencv_datasets.dir/src/ar_hmdb.cpp.o
[ 20%] Building CXX object modules/datasets/CMakeFiles/opencv_datasets.dir/src/ar_sports.cpp.o
[ 20%] Building CXX object modules/datasets/CMakeFiles/opencv_datasets.dir/src/dataset.cpp.o
[ 20%] Building CXX object modules/datasets/CMakeFiles/opencv_datasets.dir/src/fr_adience.cpp.o
[ 20%] Building CXX object modules/datasets/CMakeFiles/opencv_datasets.dir/src/fr_lfw.cpp.o
[ 20%] Building CXX object modules/datasets/CMakeFiles/opencv_datasets.dir/src/gr_chalearn.cpp.o
[ 20%] Building CXX object modules/datasets/CMakeFiles/opencv_datasets.dir/src/gr_skig.cpp.o
[ 20%] Building CXX object modules/datasets/CMakeFiles/opencv_datasets.dir/src/hpe_humaneva.cpp.o
[ 20%] Building NVCC (Device) object modules/cudaimgproc/CMakeFiles/cuda_compile_1.dir/src/cuda/cuda_compile_1_generated_hough_circles.cu.o
[ 20%] Building CXX object modules/datasets/CMakeFiles/opencv_datasets.dir/src/hpe_parse.cpp.o
[ 20%] Building CXX object modules/datasets/CMakeFiles/opencv_datasets.dir/src/ir_affine.cpp.o
[ 20%] Building NVCC (Device) object modules/cudaimgproc/CMakeFiles/cuda_compile_1.dir/src/cuda/cuda_compile_1_generated_hough_lines.cu.o
[ 20%] Building CXX object modules/datasets/CMakeFiles/opencv_datasets.dir/src/ir_robot.cpp.o
[ 20%] Building CXX object modules/datasets/CMakeFiles/opencv_datasets.dir/src/is_bsds.cpp.o
- 安装python版本
pip install opencv-python
这么做，不是cuda版本。自己编译完后执行才可以，如下所示：
$ sudo ldconfig
The final step is to sym-link the OpenCV library into your Python virtual environment.
- 安装虚拟环境
$ sudo pip install virtualenv virtualenvwrapper # ~./bashrc # virtualenv and virtualenvwrapper export WORKON_HOME=$HOME/.virtualenvs export VIRTUALENVWRAPPER_PYTHON=/usr/bin/python3 source /usr/local/bin/virtualenvwrapper.sh $ mkvirtualenv opencv_cuda -p python3 $ workon opencv_cuda
- 配置参数归总
[小建议]：cmake 的 python for build 会自动找python2，暂时将python2改名，骗过cmake即可。
-- General configuration for OpenCV 4.2.0 ===================================== -- Version control: unknown -- -- Extra modules: -- Location (extra): /usr/local/opencv/opencv_contrib/modules -- Version control (extra): unknown -- -- Platform: -- Timestamp: 2020-05-13T11:02:53Z -- Host: Linux 5.3.0-51-generic x86_64 -- CMake: 3.17.2 -- CMake generator: Unix Makefiles -- CMake build tool: /usr/bin/make -- Configuration: RELEASE -- -- CPU/HW features: -- Baseline: SSE SSE2 SSE3 -- requested: SSE3 -- Dispatched code generation: SSE4_1 SSE4_2 FP16 AVX AVX2 AVX512_SKX -- requested: SSE4_1 SSE4_2 AVX FP16 AVX2 AVX512_SKX -- SSE4_1 (16 files): + SSSE3 SSE4_1 -- SSE4_2 (2 files): + SSSE3 SSE4_1 POPCNT SSE4_2 -- FP16 (1 files): + SSSE3 SSE4_1 POPCNT SSE4_2 FP16 AVX -- AVX (5 files): + SSSE3 SSE4_1 POPCNT SSE4_2 AVX -- AVX2 (29 files): + SSSE3 SSE4_1 POPCNT SSE4_2 FP16 FMA3 AVX AVX2 -- AVX512_SKX (6 files): + SSSE3 SSE4_1 POPCNT SSE4_2 FP16 FMA3 AVX AVX2 AVX_512F AVX512_COMMON AVX512_SKX -- -- C/C++: -- Built as dynamic libs?: YES -- C++ Compiler: /usr/bin/c++ (ver 7.5.0) -- C++ flags (Release): -fsigned-char -ffast-math -W -Wall -Werror=return-type -Werror=non-virtual-dtor -Werror=address -Werror=sequence-point -Wformat -Werror=format-security -Wmissing-declarations -Wundef -Winit-self -Wpointer-arith -Wshadow -Wsign-promo -Wuninitialized -Winit-self -Wsuggest-override -Wno-delete-non-virtual-dtor -Wno-comment -Wimplicit-fallthrough=3 -Wno-strict-overflow -fdiagnostics-show-option -Wno-long-long -pthread -fomit-frame-pointer -ffunction-sections -fdata-sections -msse -msse2 -msse3 -fvisibility=hidden -fvisibility-inlines-hidden -O3 -DNDEBUG -DNDEBUG -- C++ flags (Debug): -fsigned-char -ffast-math -W -Wall -Werror=return-type -Werror=non-virtual-dtor -Werror=address -Werror=sequence-point -Wformat -Werror=format-security -Wmissing-declarations -Wundef -Winit-self -Wpointer-arith -Wshadow -Wsign-promo -Wuninitialized -Winit-self -Wsuggest-override -Wno-delete-non-virtual-dtor -Wno-comment -Wimplicit-fallthrough=3 -Wno-strict-overflow -fdiagnostics-show-option -Wno-long-long -pthread -fomit-frame-pointer -ffunction-sections -fdata-sections -msse -msse2 -msse3 -fvisibility=hidden -fvisibility-inlines-hidden -g -O0 -DDEBUG -D_DEBUG -- C Compiler: /usr/bin/cc -- C flags (Release): -fsigned-char -ffast-math -W -Wall -Werror=return-type -Werror=non-virtual-dtor -Werror=address -Werror=sequence-point -Wformat -Werror=format-security -Wmissing-declarations -Wmissing-prototypes -Wstrict-prototypes -Wundef -Winit-self -Wpointer-arith -Wshadow -Wuninitialized -Winit-self -Wno-comment -Wimplicit-fallthrough=3 -Wno-strict-overflow -fdiagnostics-show-option -Wno-long-long -pthread -fomit-frame-pointer -ffunction-sections -fdata-sections -msse -msse2 -msse3 -fvisibility=hidden -O3 -DNDEBUG -DNDEBUG -- C flags (Debug): -fsigned-char -ffast-math -W -Wall -Werror=return-type -Werror=non-virtual-dtor -Werror=address -Werror=sequence-point -Wformat -Werror=format-security -Wmissing-declarations -Wmissing-prototypes -Wstrict-prototypes -Wundef -Winit-self -Wpointer-arith -Wshadow -Wuninitialized -Winit-self -Wno-comment -Wimplicit-fallthrough=3 -Wno-strict-overflow -fdiagnostics-show-option -Wno-long-long -pthread -fomit-frame-pointer -ffunction-sections -fdata-sections -msse -msse2 -msse3 -fvisibility=hidden -g -O0 -DDEBUG -D_DEBUG -- Linker flags (Release): -Wl,--gc-sections -- Linker flags (Debug): -Wl,--gc-sections -- ccache: NO -- Precompiled headers: NO -- Extra dependencies: m pthread cudart_static dl rt nppc nppial nppicc nppicom nppidei nppif nppig nppim nppist nppisu nppitc npps cublas cudnn cufft -L/usr/local/cuda/lib64 -L/usr/lib/x86_64-linux-gnu -- 3rdparty dependencies: -- -- OpenCV modules: -- To be built: aruco bgsegm bioinspired calib3d ccalib core cudaarithm cudabgsegm cudacodec cudafeatures2d cudafilters cudaimgproc cudalegacy cudaobjdetect cudaoptflow cudastereo cudawarping cudev datasets dnn dnn_objdetect dnn_superres dpm face features2d flann freetype fuzzy gapi hdf hfs highgui img_hash imgcodecs imgproc line_descriptor ml objdetect optflow phase_unwrapping photo plot python2 python3 quality reg rgbd saliency sfm shape stereo stitching structured_light superres surface_matching text tracking ts video videoio videostab xfeatures2d ximgproc xobjdetect xphoto -- Disabled: world -- Disabled by dependency: - -- Unavailable: cnn_3dobj cvv java js matlab ovis viz -- Applications: tests perf_tests examples apps -- Documentation: NO -- Non-free algorithms: YES -- -- GUI: -- GTK+: YES (ver 3.22.30) -- GThread : YES (ver 2.56.4) -- GtkGlExt: NO -- VTK support: NO -- -- Media I/O: -- ZLib: /usr/lib/x86_64-linux-gnu/libz.so (ver 1.2.11) -- JPEG: /usr/lib/x86_64-linux-gnu/libjpeg.so (ver 80) -- WEBP: build (ver encoder: 0x020e) -- PNG: /usr/lib/x86_64-linux-gnu/libpng.so (ver 1.6.34) -- TIFF: /usr/lib/x86_64-linux-gnu/libtiff.so (ver 42 / 4.0.9) -- JPEG 2000: build (ver 1.900.1) -- OpenEXR: build (ver 2.3.0) -- HDR: YES -- SUNRASTER: YES -- PXM: YES -- PFM: YES -- -- Video I/O: -- DC1394: YES (2.2.5) -- FFMPEG: YES -- avcodec: YES (57.107.100) -- avformat: YES (57.83.100) -- avutil: YES (55.78.100) -- swscale: YES (4.8.100) -- avresample: YES (3.7.0) -- GStreamer: YES (1.14.5) -- v4l/v4l2: YES (linux/videodev2.h) -- -- Parallel framework: pthreads -- -- Trace: YES (with Intel ITT) -- -- Other third-party libraries: -- Intel IPP: 2019.0.0 Gold [2019.0.0] -- at: /usr/local/opencv/build/3rdparty/ippicv/ippicv_lnx/icv -- Intel IPP IW: sources (2019.0.0) -- at: /usr/local/opencv/build/3rdparty/ippicv/ippicv_lnx/iw -- Lapack: NO -- Eigen: YES (ver 3.3.4) -- Custom HAL: NO -- Protobuf: build (3.5.1) -- -- NVIDIA CUDA: YES (ver 10.0, CUFFT CUBLAS FAST_MATH) -- NVIDIA GPU arch: 75 -- NVIDIA PTX archs: -- -- cuDNN: YES (ver 7.6.5) -- -- OpenCL: YES (no extra features) -- Include path: /usr/local/opencv/3rdparty/include/opencl/1.2 -- Link libraries: Dynamic load -- -- Python 2: -- Interpreter: /usr/bin/python2.7 (ver 2.7.17) -- Libraries: /usr/lib/x86_64-linux-gnu/libpython2.7.so (ver 2.7.17) -- numpy: /usr/lib/python2.7/dist-packages/numpy/core/include (ver 1.13.3) -- install path: lib/python2.7/dist-packages/cv2/python-2.7 -- -- Python 3: -- Interpreter: /home/jeffrey/.virtualenvs/opencv_cuda/bin/python3 (ver 3.6.9) -- Libraries: /usr/lib/x86_64-linux-gnu/libpython3.6m.so (ver 3.6.9) -- numpy: /home/jeffrey/.virtualenvs/opencv_cuda/lib/python3.6/site-packages/numpy/core/include (ver 1.18.4) -- install path: lib/python3.6/site-packages/cv2/python-3.6 -- -- Python (for build): /usr/bin/python2.7 -- Pylint: /home/jeffrey/.local/bin/pylint (ver: unknown, checks: 163) -- -- Java: -- ant: NO -- JNI: /usr/local/lib/jvm/jdk1.8.0_231/include /usr/local/lib/jvm/jdk1.8.0_231/include/linux /usr/local/lib/jvm/jdk1.8.0_231/include -- Java wrappers: NO -- Java tests: NO -- -- Install to: /usr/local -- ----------------------------------------------------------------- -- -- Configuring done -- Generating done
make install 配置路径
opencv4安装后，头文件显示的是opencv4，建议手动改为opencv。
jeffrey@unsw-ThinkPad-T490:build$ apt-file search opencv.pc libopencv-dev: /usr/lib/x86_64-linux-gnu/pkgconfig/opencv.pc jeffrey@unsw-ThinkPad-T490:build$ pkg-config --cflags opencv -I/usr/local/include/opencv -I/usr/local/include jeffrey@unsw-ThinkPad-T490:build$ pkg-config --libs opencv -L/usr/local/lib -lopencv_dnn -lopencv_ml -lopencv_objdetect -lopencv_shape -lopencv_stitching -lopencv_superres -lopencv_videostab -lopencv_calib3d -lopencv_features2d -lopencv_highgui -lopencv_videoio -lopencv_imgcodecs -lopencv_video -lopencv_photo -lopencv_imgproc -lopencv_flann -lopencv_core jeffrey@unsw-ThinkPad-T490:build$ ls /usr/local/include/opencv opencv/ opencv4/
jeffrey@unsw-ThinkPad-T490:build$ ls /usr/local/include/opencv opencv2 jeffrey@unsw-ThinkPad-T490:build$ ls /usr/local/include/opencv/opencv2/dnn dnn/ dnn.hpp dnn_superres.hpp
jeffrey@unsw-ThinkPad-T490:build$ ls /usr/local/include/opencv/opencv2/dnn/ all_layers.hpp dnn.hpp layer.details.hpp shape_utils.hpp version.hpp dict.hpp dnn.inl.hpp layer.hpp utils
GPU DRIVER 问题集合
Keras 遇到类似问题，可直接替换最新的TF 2.1。
另外的原因就是：安装的版本不匹配！
Found 10 images belonging to 1 classes. Epoch 1/1 Found 10 images belonging to 1 classes. 2020-05-13 20:03:38.972313: I tensorflow/stream_executor/platform/default/dso_loader.cc:42] Successfully opened dynamic library libcudnn.so.7 2020-05-13 20:03:39.424152: E tensorflow/stream_executor/cuda/cuda_dnn.cc:329] Could not create cudnn handle: CUDNN_STATUS_INTERNAL_ERROR 2020-05-13 20:03:39.426938: E tensorflow/stream_executor/cuda/cuda_dnn.cc:329] Could not create cudnn handle: CUDNN_STATUS_INTERNAL_ERROR Traceback (most recent call last): File "./train.py", line 52, in <module> model.fit_generator(myGene,steps_per_epoch=5,epochs=1,callbacks=[model_checkpoint]) File "/usr/local/anaconda3/lib/python3.6/site-packages/keras/legacy/interfaces.py", line 91, in wrapper return func(*args, **kwargs) File "/usr/local/anaconda3/lib/python3.6/site-packages/keras/engine/training.py", line 1732, in fit_generator initial_epoch=initial_epoch) File "/usr/local/anaconda3/lib/python3.6/site-packages/keras/engine/training_generator.py", line 220, in fit_generator reset_metrics=False) File "/usr/local/anaconda3/lib/python3.6/site-packages/keras/engine/training.py", line 1514, in train_on_batch outputs = self.train_function(ins) File "/usr/local/anaconda3/lib/python3.6/site-packages/tensorflow/python/keras/backend.py", line 3292, in __call__ run_metadata=self.run_metadata) File "/usr/local/anaconda3/lib/python3.6/site-packages/tensorflow/python/client/session.py", line 1458, in __call__ run_metadata_ptr) tensorflow.python.framework.errors_impl.UnknownError: 2 root error(s) found. (0) Unknown: Failed to get convolution algorithm. This is probably because cuDNN failed to initialize, so try looking to see if a warning log message was printed above. [[{{node conv2d_1/convolution}}]] [[metrics/accuracy/Identity/_435]] (1) Unknown: Failed to get convolution algorithm. This is probably because cuDNN failed to initialize, so try looking to see if a warning log message was printed above. [[{{node conv2d_1/convolution}}]] 0 successful operations. 0 derived errors ignored.
必备步骤实操
GPU 驱动
一、Cuda 卸载
Ref: Build from source
版本都是对应的。
| tensorflow_gpu-1.14.0 | 2.7, 3.3-3.7 | GCC 4.8 | Bazel 0.24.1 | 7.4 | 10.0 |
| tensorflow_gpu-1.13.1 | 2.7, 3.3-3.7 | GCC 4.8 | Bazel 0.19.2 | 7.4 | 10.0 |
For CUDA 10.1, try:
sudo /usr/local/cuda/bin/cuda-uninstaller
For CUDA 10.0, try:
sudo /usr/local/cuda/bin/uninstall_cuda_10.0.pl
二、Cuda 安装
-
手动安装方式
Ubuntu installs its driver firstly and uses nvidia-smi to check the version.
We can see that Cuda 10.1 is required.
Goto: https://developer.nvidia.com/cuda-toolkit-archive
Old log, useless: [CUDA] 00 - GPU Driver Installation & Concurrency Programming
-
deb安装方式
Ref: GPU support
Add NVIDIA package repositories.
然后，直接从仓库下载。
sudo apt-get install --no-install-recommends cuda-10-2
Cudnn: Download cuDNN v7.6.5 (November 18th, 2019), for CUDA 10.2
三、GPU 未成功启用
tf1.14与cuda10.1的一个常见问题：Compatibility with Cuda 10.1? #26289
这是一个坑，不同的模型使用不同版本的tf可能会更好。
cuda10.0可能会好一些，但是：
2020-05-14 16:17:18.082942: I tensorflow/core/platform/cpu_feature_guard.cc:142] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2 FMA 2020-05-14 16:17:18.247535: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:1005] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero 2020-05-14 16:17:18.247974: I tensorflow/compiler/xla/service/service.cc:168] XLA service 0x558e80a55fa0 executing computations on platform CUDA. Devices: 2020-05-14 16:17:18.247990: I tensorflow/compiler/xla/service/service.cc:175] StreamExecutor device (0): GeForce MX250, Compute Capability 6.1 2020-05-14 16:17:18.275668: I tensorflow/core/platform/profile_utils/cpu_utils.cc:94] CPU Frequency: 1999965000 Hz 2020-05-14 16:17:18.275970: I tensorflow/compiler/xla/service/service.cc:168] XLA service 0x558e7577dc70 executing computations on platform Host. Devices: 2020-05-14 16:17:18.275988: I tensorflow/compiler/xla/service/service.cc:175] StreamExecutor device (0): <undefined>, <undefined> 2020-05-14 16:17:18.276199: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:1005] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero 2020-05-14 16:17:18.276478: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1640] Found device 0 with properties: name: GeForce MX250 major: 6 minor: 1 memoryClockRate(GHz): 1.582 pciBusID: 0000:3c:00.0 2020-05-14 16:17:18.276583: I tensorflow/stream_executor/platform/default/dso_loader.cc:53] Could not dlopen library 'libcudart.so.10.0'; dlerror: libcudart.so.10.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /usr/local/anaconda3/lib/:/usr/local/lib/:/usr/local/cuda/lib64: 2020-05-14 16:17:18.276641: I tensorflow/stream_executor/platform/default/dso_loader.cc:53] Could not dlopen library 'libcublas.so.10.0'; dlerror: libcublas.so.10.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /usr/local/anaconda3/lib/:/usr/local/lib/:/usr/local/cuda/lib64: 2020-05-14 16:17:18.276697: I tensorflow/stream_executor/platform/default/dso_loader.cc:53] Could not dlopen library 'libcufft.so.10.0'; dlerror: libcufft.so.10.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /usr/local/anaconda3/lib/:/usr/local/lib/:/usr/local/cuda/lib64: 2020-05-14 16:17:18.276752: I tensorflow/stream_executor/platform/default/dso_loader.cc:53] Could not dlopen library 'libcurand.so.10.0'; dlerror: libcurand.so.10.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /usr/local/anaconda3/lib/:/usr/local/lib/:/usr/local/cuda/lib64: 2020-05-14 16:17:18.276811: I tensorflow/stream_executor/platform/default/dso_loader.cc:53] Could not dlopen library 'libcusolver.so.10.0'; dlerror: libcusolver.so.10.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /usr/local/anaconda3/lib/:/usr/local/lib/:/usr/local/cuda/lib64: 2020-05-14 16:17:18.276867: I tensorflow/stream_executor/platform/default/dso_loader.cc:53] Could not dlopen library 'libcusparse.so.10.0'; dlerror: libcusparse.so.10.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /usr/local/anaconda3/lib/:/usr/local/lib/:/usr/local/cuda/lib64: 2020-05-14 16:17:18.276886: I tensorflow/stream_executor/platform/default/dso_loader.cc:42] Successfully opened dynamic library libcudnn.so.7 2020-05-14 16:17:18.276898: W tensorflow/core/common_runtime/gpu/gpu_device.cc:1663] Cannot dlopen some GPU libraries. Skipping registering GPU devices...
OpenCV Cuda 编译版本
一、依赖问题
OpenCV编译没问题，编译工程可能会出现该问题。
对‘TIFFReadDirectory@LIBTIFF_4.0’未定义的引用
网上有介绍通过sudo apt-get install libtiff4-dev安装，但是运行该命令后，提示libtiff4-dev已经被libtiff5-dev替代，只可安装libtiff5-dev，安装后并未解决问题。
其原因时安装的libtiff库和opencv和PLC中调用的libtiff库版本与安装的版本不对应，去libtiff管网上下载 tiff-4.0.4.tar.gz压缩包安装即可。
开cuda的一些fps对比数据。
Allow the OpenCV's DNN module to work with GPUs
二、已验证的安装
How to install OpenCV 4.2.0 with CUDA 10.0 in Ubuntu distro 18.04 [感动的要脆了，这个最好]
$ nvidia-smi
Sat May 16 17:13:04 2020
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 435.21 Driver Version: 435.21 CUDA Version: 10.1 |
|-------------------------------+----------------------+----------------------+
| GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. |
|===============================+======================+======================|
| 0 GeForce MX250 Off | 00000000:3C:00.0 Off | N/A |
| N/A 57C P0 N/A / N/A | 1991MiB / 2002MiB | 5% Default |
+-------------------------------+----------------------+----------------------+
+-----------------------------------------------------------------------------+
| Processes: GPU Memory |
| GPU PID Type Process name Usage |
|=============================================================================|
| 0 1634 G /usr/lib/xorg/Xorg 101MiB |
| 0 1798 G /usr/bin/gnome-shell 60MiB |
| 0 2253 C python 1781MiB |
| 0 2519 G ...AAAAAAAAAAAAAAgAAAAAAAAA --shared-files 43MiB |
+-----------------------------------------------------------------------------+
包括了AVX2优化。
build$ cmake -D CMAKE_BUILD_TYPE=RELEASE -D CMAKE_INSTALL_PREFIX=/usr/local -D INSTALL_PYTHON_EXAMPLES=ON -D INSTALL_C_EXAMPLES=OFF -D WITH_TBB=ON -D WITH_CUDA=ON -D BUILD_opencv_cudacodec=OFF -D ENABLE_FAST_MATH=1 -D CUDA_FAST_MATH=1 -D WITH_CUBLAS=1 -D WITH_V4L=ON -D WITH_QT=OFF -D WITH_OPENGL=ON -D WITH_GSTREAMER=ON -D OPENCV_GENERATE_PKGCONFIG=ON -D OPENCV_PC_FILE_NAME=opencv.pc -D OPENCV_ENABLE_NONFREE=ON -D OPENCV_PYTHON3_INSTALL_PATH=~/.virtualenvs/cv/lib/python3.6/site-packages -D OPENCV_EXTRA_MODULES_PATH=../opencv_contrib/modules -D PYTHON_EXECUTABLE=~/.virtualenvs/cv/bin/python -D BUILD_EXAMPLES=ON -D WITH_CUDNN=ON -D OPENCV_DNN_CUDA=ON -D CUDA_ARCH_BIN=6.1 CPU_BASELINE=AVX2 .. -- The CXX compiler identification is GNU 7.5.0 -- The C compiler identification is GNU 7.5.0 -- Check for working CXX compiler: /usr/bin/c++ -- Check for working CXX compiler: /usr/bin/c++ - works -- Detecting CXX compiler ABI info -- Detecting CXX compiler ABI info - done -- Detecting CXX compile features -- Detecting CXX compile features - done -- Check for working C compiler: /usr/bin/cc -- Check for working C compiler: /usr/bin/cc - works -- Detecting C compiler ABI info -- Detecting C compiler ABI info - done -- Detecting C compile features -- Detecting C compile features - done -- Detected processor: x86_64 -- Found PythonInterp: /home/jeffrey/.virtualenvs/cv/bin/python (found suitable version "3.6.9", minimum required is "2.7") CMake Warning at cmake/OpenCVDetectPython.cmake:81 (message): CMake's 'find_host_package(PythonInterp 2.7)' founds wrong Python version: PYTHON_EXECUTABLE=/home/jeffrey/.virtualenvs/cv/bin/python PYTHON_VERSION_STRING=3.6.9 Consider specify 'PYTHON2_EXECUTABLE' variable via CMake command line or environment variables Call Stack (most recent call first): cmake/OpenCVDetectPython.cmake:271 (find_python) CMakeLists.txt:585 (include) -- Found Python2: /usr/bin/python2.7 (found version "2.7.17") found components: Interpreter -- Found PythonInterp: /usr/bin/python2.7 (found version "2.7.17") -- Found PythonLibs: /usr/lib/x86_64-linux-gnu/libpython2.7.so (found suitable exact version "2.7.17") -- Found PythonInterp: /home/jeffrey/.virtualenvs/cv/bin/python3 (found suitable version "3.6.9", minimum required is "3.2") -- Found PythonLibs: /usr/lib/x86_64-linux-gnu/libpython3.6m.so (found suitable exact version "3.6.9") -- Looking for ccache - not found -- Performing Test HAVE_CXX_FSIGNED_CHAR -- Performing Test HAVE_CXX_FSIGNED_CHAR - Success -- Performing Test HAVE_C_FSIGNED_CHAR -- Performing Test HAVE_C_FSIGNED_CHAR - Success -- Performing Test HAVE_CXX_FFAST_MATH -- Performing Test HAVE_CXX_FFAST_MATH - Success -- Performing Test HAVE_C_FFAST_MATH -- Performing Test HAVE_C_FFAST_MATH - Success -- Performing Test HAVE_CXX_W -- Performing Test HAVE_CXX_W - Success -- Performing Test HAVE_C_W -- Performing Test HAVE_C_W - Success -- Performing Test HAVE_CXX_WALL -- Performing Test HAVE_CXX_WALL - Success -- Performing Test HAVE_C_WALL -- Performing Test HAVE_C_WALL - Success -- Performing Test HAVE_CXX_WERROR_RETURN_TYPE -- Performing Test HAVE_CXX_WERROR_RETURN_TYPE - Success -- Performing Test HAVE_C_WERROR_RETURN_TYPE -- Performing Test HAVE_C_WERROR_RETURN_TYPE - Success -- Performing Test HAVE_CXX_WERROR_NON_VIRTUAL_DTOR -- Performing Test HAVE_CXX_WERROR_NON_VIRTUAL_DTOR - Success -- Performing Test HAVE_C_WERROR_NON_VIRTUAL_DTOR -- Performing Test HAVE_C_WERROR_NON_VIRTUAL_DTOR - Success -- Performing Test HAVE_CXX_WERROR_ADDRESS -- Performing Test HAVE_CXX_WERROR_ADDRESS - Success -- Performing Test HAVE_C_WERROR_ADDRESS -- Performing Test HAVE_C_WERROR_ADDRESS - Success -- Performing Test HAVE_CXX_WERROR_SEQUENCE_POINT -- Performing Test HAVE_CXX_WERROR_SEQUENCE_POINT - Success -- Performing Test HAVE_C_WERROR_SEQUENCE_POINT -- Performing Test HAVE_C_WERROR_SEQUENCE_POINT - Success -- Performing Test HAVE_CXX_WFORMAT -- Performing Test HAVE_CXX_WFORMAT - Success -- Performing Test HAVE_C_WFORMAT -- Performing Test HAVE_C_WFORMAT - Success -- Performing Test HAVE_CXX_WERROR_FORMAT_SECURITY -- Performing Test HAVE_CXX_WERROR_FORMAT_SECURITY - Success -- Performing Test HAVE_C_WERROR_FORMAT_SECURITY -- Performing Test HAVE_C_WERROR_FORMAT_SECURITY - Success -- Performing Test HAVE_CXX_WMISSING_DECLARATIONS -- Performing Test HAVE_CXX_WMISSING_DECLARATIONS - Success -- Performing Test HAVE_C_WMISSING_DECLARATIONS -- Performing Test HAVE_C_WMISSING_DECLARATIONS - Success -- Performing Test HAVE_CXX_WMISSING_PROTOTYPES -- Performing Test HAVE_CXX_WMISSING_PROTOTYPES - Failed -- Performing Test HAVE_C_WMISSING_PROTOTYPES -- Performing Test HAVE_C_WMISSING_PROTOTYPES - Success -- Performing Test HAVE_CXX_WSTRICT_PROTOTYPES -- Performing Test HAVE_CXX_WSTRICT_PROTOTYPES - Failed -- Performing Test HAVE_C_WSTRICT_PROTOTYPES -- Performing Test HAVE_C_WSTRICT_PROTOTYPES - Success -- Performing Test HAVE_CXX_WUNDEF -- Performing Test HAVE_CXX_WUNDEF - Success -- Performing Test HAVE_C_WUNDEF -- Performing Test HAVE_C_WUNDEF - Success -- Performing Test HAVE_CXX_WINIT_SELF -- Performing Test HAVE_CXX_WINIT_SELF - Success -- Performing Test HAVE_C_WINIT_SELF -- Performing Test HAVE_C_WINIT_SELF - Success -- Performing Test HAVE_CXX_WPOINTER_ARITH -- Performing Test HAVE_CXX_WPOINTER_ARITH - Success -- Performing Test HAVE_C_WPOINTER_ARITH -- Performing Test HAVE_C_WPOINTER_ARITH - Success -- Performing Test HAVE_CXX_WSHADOW -- Performing Test HAVE_CXX_WSHADOW - Success -- Performing Test HAVE_C_WSHADOW -- Performing Test HAVE_C_WSHADOW - Success -- Performing Test HAVE_CXX_WSIGN_PROMO -- Performing Test HAVE_CXX_WSIGN_PROMO - Success -- Performing Test HAVE_C_WSIGN_PROMO -- Performing Test HAVE_C_WSIGN_PROMO - Failed -- Performing Test HAVE_CXX_WUNINITIALIZED -- Performing Test HAVE_CXX_WUNINITIALIZED - Success -- Performing Test HAVE_C_WUNINITIALIZED -- Performing Test HAVE_C_WUNINITIALIZED - Success -- Performing Test HAVE_CXX_WSUGGEST_OVERRIDE -- Performing Test HAVE_CXX_WSUGGEST_OVERRIDE - Success -- Performing Test HAVE_C_WSUGGEST_OVERRIDE -- Performing Test HAVE_C_WSUGGEST_OVERRIDE - Failed -- Performing Test HAVE_CXX_WNO_DELETE_NON_VIRTUAL_DTOR -- Performing Test HAVE_CXX_WNO_DELETE_NON_VIRTUAL_DTOR - Success -- Performing Test HAVE_C_WNO_DELETE_NON_VIRTUAL_DTOR -- Performing Test HAVE_C_WNO_DELETE_NON_VIRTUAL_DTOR - Failed -- Performing Test HAVE_CXX_WNO_UNNAMED_TYPE_TEMPLATE_ARGS -- Performing Test HAVE_CXX_WNO_UNNAMED_TYPE_TEMPLATE_ARGS - Failed -- Performing Test HAVE_C_WNO_UNNAMED_TYPE_TEMPLATE_ARGS -- Performing Test HAVE_C_WNO_UNNAMED_TYPE_TEMPLATE_ARGS - Failed -- Performing Test HAVE_CXX_WNO_COMMENT -- Performing Test HAVE_CXX_WNO_COMMENT - Success -- Performing Test HAVE_C_WNO_COMMENT -- Performing Test HAVE_C_WNO_COMMENT - Success -- Performing Test HAVE_CXX_WIMPLICIT_FALLTHROUGH_3 -- Performing Test HAVE_CXX_WIMPLICIT_FALLTHROUGH_3 - Success -- Performing Test HAVE_C_WIMPLICIT_FALLTHROUGH_3 -- Performing Test HAVE_C_WIMPLICIT_FALLTHROUGH_3 - Success -- Performing Test HAVE_CXX_WNO_STRICT_OVERFLOW -- Performing Test HAVE_CXX_WNO_STRICT_OVERFLOW - Success -- Performing Test HAVE_C_WNO_STRICT_OVERFLOW -- Performing Test HAVE_C_WNO_STRICT_OVERFLOW - Success -- Performing Test HAVE_CXX_FDIAGNOSTICS_SHOW_OPTION -- Performing Test HAVE_CXX_FDIAGNOSTICS_SHOW_OPTION - Success -- Performing Test HAVE_C_FDIAGNOSTICS_SHOW_OPTION -- Performing Test HAVE_C_FDIAGNOSTICS_SHOW_OPTION - Success -- Performing Test HAVE_CXX_WNO_LONG_LONG -- Performing Test HAVE_CXX_WNO_LONG_LONG - Success -- Performing Test HAVE_C_WNO_LONG_LONG -- Performing Test HAVE_C_WNO_LONG_LONG - Success -- Performing Test HAVE_CXX_PTHREAD -- Performing Test HAVE_CXX_PTHREAD - Success -- Performing Test HAVE_C_PTHREAD -- Performing Test HAVE_C_PTHREAD - Success -- Performing Test HAVE_CXX_FOMIT_FRAME_POINTER -- Performing Test HAVE_CXX_FOMIT_FRAME_POINTER - Success -- Performing Test HAVE_C_FOMIT_FRAME_POINTER -- Performing Test HAVE_C_FOMIT_FRAME_POINTER - Success -- Performing Test HAVE_CXX_FFUNCTION_SECTIONS -- Performing Test HAVE_CXX_FFUNCTION_SECTIONS - Success -- Performing Test HAVE_C_FFUNCTION_SECTIONS -- Performing Test HAVE_C_FFUNCTION_SECTIONS - Success -- Performing Test HAVE_CXX_FDATA_SECTIONS -- Performing Test HAVE_CXX_FDATA_SECTIONS - Success -- Performing Test HAVE_C_FDATA_SECTIONS -- Performing Test HAVE_C_FDATA_SECTIONS - Success -- Performing Test HAVE_CXX_MSSE (check file: cmake/checks/cpu_sse.cpp) -- Performing Test HAVE_CXX_MSSE - Success -- Performing Test HAVE_CXX_MSSE2 (check file: cmake/checks/cpu_sse2.cpp) -- Performing Test HAVE_CXX_MSSE2 - Success -- Performing Test HAVE_CXX_MSSE3 (check file: cmake/checks/cpu_sse3.cpp) -- Performing Test HAVE_CXX_MSSE3 - Success -- Performing Test HAVE_CXX_MSSSE3 (check file: cmake/checks/cpu_ssse3.cpp) -- Performing Test HAVE_CXX_MSSSE3 - Success -- Performing Test HAVE_CXX_MSSE4_1 (check file: cmake/checks/cpu_sse41.cpp) -- Performing Test HAVE_CXX_MSSE4_1 - Success -- Performing Test HAVE_CXX_MPOPCNT (check file: cmake/checks/cpu_popcnt.cpp) -- Performing Test HAVE_CXX_MPOPCNT - Success -- Performing Test HAVE_CXX_MSSE4_2 (check file: cmake/checks/cpu_sse42.cpp) -- Performing Test HAVE_CXX_MSSE4_2 - Success -- Performing Test HAVE_CXX_MF16C (check file: cmake/checks/cpu_fp16.cpp) -- Performing Test HAVE_CXX_MF16C - Success -- Performing Test HAVE_CXX_MFMA -- Performing Test HAVE_CXX_MFMA - Success -- Performing Test HAVE_CXX_MAVX (check file: cmake/checks/cpu_avx.cpp) -- Performing Test HAVE_CXX_MAVX - Success -- Performing Test HAVE_CXX_MAVX2 (check file: cmake/checks/cpu_avx2.cpp) -- Performing Test HAVE_CXX_MAVX2 - Success -- Performing Test HAVE_CXX_MAVX512F (check file: cmake/checks/cpu_avx512.cpp) -- Performing Test HAVE_CXX_MAVX512F - Success -- Performing Test HAVE_CXX_MAVX512F_MAVX512CD (check file: cmake/checks/cpu_avx512common.cpp) -- Performing Test HAVE_CXX_MAVX512F_MAVX512CD - Success -- Performing Test HAVE_CXX_MAVX512F_MAVX512CD_MAVX512VL_MAVX512BW_MAVX512DQ (check file: cmake/checks/cpu_avx512skx.cpp) -- Performing Test HAVE_CXX_MAVX512F_MAVX512CD_MAVX512VL_MAVX512BW_MAVX512DQ - Success -- Performing Test HAVE_CPU_BASELINE_FLAGS -- Performing Test HAVE_CPU_BASELINE_FLAGS - Success -- Performing Test HAVE_CPU_DISPATCH_FLAGS_SSE4_1 -- Performing Test HAVE_CPU_DISPATCH_FLAGS_SSE4_1 - Success -- Performing Test HAVE_CPU_DISPATCH_FLAGS_SSE4_2 -- Performing Test HAVE_CPU_DISPATCH_FLAGS_SSE4_2 - Success -- Performing Test HAVE_CPU_DISPATCH_FLAGS_FP16 -- Performing Test HAVE_CPU_DISPATCH_FLAGS_FP16 - Success -- Performing Test HAVE_CPU_DISPATCH_FLAGS_AVX -- Performing Test HAVE_CPU_DISPATCH_FLAGS_AVX - Success -- Performing Test HAVE_CPU_DISPATCH_FLAGS_AVX2 -- Performing Test HAVE_CPU_DISPATCH_FLAGS_AVX2 - Success -- Performing Test HAVE_CPU_DISPATCH_FLAGS_AVX512_SKX -- Performing Test HAVE_CPU_DISPATCH_FLAGS_AVX512_SKX - Success -- Performing Test HAVE_CXX_FVISIBILITY_HIDDEN -- Performing Test HAVE_CXX_FVISIBILITY_HIDDEN - Success -- Performing Test HAVE_C_FVISIBILITY_HIDDEN -- Performing Test HAVE_C_FVISIBILITY_HIDDEN - Success -- Performing Test HAVE_CXX_FVISIBILITY_INLINES_HIDDEN -- Performing Test HAVE_CXX_FVISIBILITY_INLINES_HIDDEN - Success -- Performing Test HAVE_C_FVISIBILITY_INLINES_HIDDEN -- Performing Test HAVE_C_FVISIBILITY_INLINES_HIDDEN - Failed -- Looking for pthread.h -- Looking for pthread.h - found -- Looking for posix_memalign -- Looking for posix_memalign - found -- Looking for malloc.h -- Looking for malloc.h - found -- Looking for memalign -- Looking for memalign - found -- Check if the system is big endian -- Searching 16 bit integer -- Looking for sys/types.h -- Looking for sys/types.h - found -- Looking for stdint.h -- Looking for stdint.h - found -- Looking for stddef.h -- Looking for stddef.h - found -- Check size of unsigned short -- Check size of unsigned short - done -- Searching 16 bit integer - Using unsigned short -- Check if the system is big endian - little endian -- Found ZLIB: /usr/lib/x86_64-linux-gnu/libz.so (found suitable version "1.2.11", minimum required is "1.2.3") -- Found JPEG: /usr/lib/x86_64-linux-gnu/libjpeg.so (found version "80") -- Found TIFF: /usr/local/lib/libtiff.so (found version "4.1.0") -- Performing Test HAVE_C_STD_C99 -- Performing Test HAVE_C_STD_C99 - Success -- Performing Test HAVE_C_WNO_UNUSED_VARIABLE -- Performing Test HAVE_C_WNO_UNUSED_VARIABLE - Success -- Performing Test HAVE_C_WNO_UNUSED_FUNCTION -- Performing Test HAVE_C_WNO_UNUSED_FUNCTION - Success -- Performing Test HAVE_C_WNO_SHADOW -- Performing Test HAVE_C_WNO_SHADOW - Success -- Performing Test HAVE_C_WNO_MAYBE_UNINITIALIZED -- Performing Test HAVE_C_WNO_MAYBE_UNINITIALIZED - Success -- Performing Test HAVE_C_WNO_MISSING_PROTOTYPES -- Performing Test HAVE_C_WNO_MISSING_PROTOTYPES - Success -- Performing Test HAVE_C_WNO_MISSING_DECLARATIONS -- Performing Test HAVE_C_WNO_MISSING_DECLARATIONS - Success -- Performing Test HAVE_C_WNO_IMPLICIT_FALLTHROUGH -- Performing Test HAVE_C_WNO_IMPLICIT_FALLTHROUGH - Success -- Found Jasper: /usr/lib/x86_64-linux-gnu/libjasper.so (found version "1.900.1") -- Found ZLIB: /usr/lib/x86_64-linux-gnu/libz.so (found version "1.2.11") -- Found PNG: /usr/lib/x86_64-linux-gnu/libpng.so (found version "1.6.34") -- Looking for /usr/include/libpng/png.h -- Looking for /usr/include/libpng/png.h - found -- Looking for semaphore.h -- Looking for semaphore.h - found -- Performing Test HAVE_CXX_WNO_SHADOW -- Performing Test HAVE_CXX_WNO_SHADOW - Success -- Performing Test HAVE_CXX_WNO_UNUSED -- Performing Test HAVE_CXX_WNO_UNUSED - Success -- Performing Test HAVE_CXX_WNO_SIGN_COMPARE -- Performing Test HAVE_CXX_WNO_SIGN_COMPARE - Success -- Performing Test HAVE_CXX_WNO_UNDEF -- Performing Test HAVE_CXX_WNO_UNDEF - Success -- Performing Test HAVE_CXX_WNO_MISSING_DECLARATIONS -- Performing Test HAVE_CXX_WNO_MISSING_DECLARATIONS - Success -- Performing Test HAVE_CXX_WNO_UNINITIALIZED -- Performing Test HAVE_CXX_WNO_UNINITIALIZED - Success -- Performing Test HAVE_CXX_WNO_SWITCH -- Performing Test HAVE_CXX_WNO_SWITCH - Success -- Performing Test HAVE_CXX_WNO_PARENTHESES -- Performing Test HAVE_CXX_WNO_PARENTHESES - Success -- Performing Test HAVE_CXX_WNO_ARRAY_BOUNDS -- Performing Test HAVE_CXX_WNO_ARRAY_BOUNDS - Success -- Performing Test HAVE_CXX_WNO_EXTRA -- Performing Test HAVE_CXX_WNO_EXTRA - Success -- Performing Test HAVE_CXX_WNO_DEPRECATED_DECLARATIONS -- Performing Test HAVE_CXX_WNO_DEPRECATED_DECLARATIONS - Success -- Performing Test HAVE_CXX_WNO_MISLEADING_INDENTATION -- Performing Test HAVE_CXX_WNO_MISLEADING_INDENTATION - Success -- Performing Test HAVE_CXX_WNO_DEPRECATED -- Performing Test HAVE_CXX_WNO_DEPRECATED - Success -- Performing Test HAVE_CXX_WNO_SUGGEST_OVERRIDE -- Performing Test HAVE_CXX_WNO_SUGGEST_OVERRIDE - Success -- Performing Test HAVE_CXX_WNO_INCONSISTENT_MISSING_OVERRIDE -- Performing Test HAVE_CXX_WNO_INCONSISTENT_MISSING_OVERRIDE - Failed -- Performing Test HAVE_CXX_WNO_IMPLICIT_FALLTHROUGH -- Performing Test HAVE_CXX_WNO_IMPLICIT_FALLTHROUGH - Success -- Performing Test HAVE_CXX_WNO_TAUTOLOGICAL_COMPARE -- Performing Test HAVE_CXX_WNO_TAUTOLOGICAL_COMPARE - Success -- Performing Test HAVE_CXX_WNO_MISSING_PROTOTYPES -- Performing Test HAVE_CXX_WNO_MISSING_PROTOTYPES - Failed -- Performing Test HAVE_CXX_WNO_REORDER -- Performing Test HAVE_CXX_WNO_REORDER - Success -- Performing Test HAVE_CXX_WNO_UNUSED_RESULT -- Performing Test HAVE_CXX_WNO_UNUSED_RESULT - Success -- Checking for module 'gtk+-3.0' -- Found gtk+-3.0, version 3.22.30 -- Checking for module 'gthread-2.0' -- Found gthread-2.0, version 2.56.4 -- Found TBB (env): /usr/lib/x86_64-linux-gnu/libtbb.so -- IPPICV: Download: ippicv_2019_lnx_intel64_general_20180723.tgz -- found Intel IPP (ICV version): 2019.0.0 [2019.0.0 Gold] -- at: /usr/local/opencv/build/3rdparty/ippicv/ippicv_lnx/icv -- found Intel IPP Integration Wrappers sources: 2019.0.0 -- at: /usr/local/opencv/build/3rdparty/ippicv/ippicv_lnx/iw -- Looking for pthread.h -- Looking for pthread.h - found -- Performing Test CMAKE_HAVE_LIBC_PTHREAD -- Performing Test CMAKE_HAVE_LIBC_PTHREAD - Failed -- Looking for pthread_create in pthreads -- Looking for pthread_create in pthreads - not found -- Looking for pthread_create in pthread -- Looking for pthread_create in pthread - found -- Found Threads: TRUE -- Found CUDNN: /usr/lib/x86_64-linux-gnu/libcudnn.so (found suitable version "7.6.5", minimum required is "7.5") -- CUDA detected: 10.0 -- CUDA NVCC target flags: -gencode;arch=compute_61,code=sm_61;-D_FORCE_INLINES -- Could not find OpenBLAS include. Turning OpenBLAS_FOUND off -- Could not find OpenBLAS lib. Turning OpenBLAS_FOUND off -- Could NOT find Atlas (missing: Atlas_CLAPACK_INCLUDE_DIR) -- Looking for sgemm_ -- Looking for sgemm_ - not found -- Looking for dgemm_ -- Looking for dgemm_ - found -- Found BLAS: /usr/lib/x86_64-linux-gnu/libblas.so;/usr/lib/x86_64-linux-gnu/libf77blas.so;/usr/lib/x86_64-linux-gnu/libatlas.so -- Looking for cheev_ -- Looking for cheev_ - not found -- Looking for cheev_ -- Looking for cheev_ - found -- A library with LAPACK API found. -- Performing Test HAVE_CXX_WNO_UNUSED_PARAMETER -- Performing Test HAVE_CXX_WNO_UNUSED_PARAMETER - Success -- Performing Test HAVE_CXX_WNO_UNUSED_LOCAL_TYPEDEFS -- Performing Test HAVE_CXX_WNO_UNUSED_LOCAL_TYPEDEFS - Success -- Performing Test HAVE_CXX_WNO_SIGN_PROMO -- Performing Test HAVE_CXX_WNO_SIGN_PROMO - Success -- Performing Test HAVE_CXX_WNO_TAUTOLOGICAL_UNDEFINED_COMPARE -- Performing Test HAVE_CXX_WNO_TAUTOLOGICAL_UNDEFINED_COMPARE - Failed -- Performing Test HAVE_CXX_WNO_IGNORED_QUALIFIERS -- Performing Test HAVE_CXX_WNO_IGNORED_QUALIFIERS - Success -- Performing Test HAVE_CXX_WNO_UNUSED_FUNCTION -- Performing Test HAVE_CXX_WNO_UNUSED_FUNCTION - Success -- Performing Test HAVE_CXX_WNO_UNUSED_CONST_VARIABLE -- Performing Test HAVE_CXX_WNO_UNUSED_CONST_VARIABLE - Success -- Performing Test HAVE_CXX_WNO_SHORTEN_64_TO_32 -- Performing Test HAVE_CXX_WNO_SHORTEN_64_TO_32 - Failed -- Performing Test HAVE_CXX_WNO_INVALID_OFFSETOF -- Performing Test HAVE_CXX_WNO_INVALID_OFFSETOF - Success -- Performing Test HAVE_CXX_WNO_ENUM_COMPARE_SWITCH -- Performing Test HAVE_CXX_WNO_ENUM_COMPARE_SWITCH - Failed -- Found JNI: /usr/local/lib/jvm/jdk1.8.0_231/jre/lib/amd64/libjawt.so -- Found Pylint: /home/jeffrey/.local/bin/pylint -- Could NOT find Flake8 (missing: FLAKE8_EXECUTABLE) -- VTK is not found. Please set -DVTK_DIR in CMake to VTK build directory, or to VTK install subdirectory with VTKConfig.cmake file -- Looking for dlerror in dl -- Looking for dlerror in dl - found -- Performing Test HAVE_C_WNO_UNDEF -- Performing Test HAVE_C_WNO_UNDEF - Success -- Performing Test HAVE_C_WNO_SIGN_COMPARE -- Performing Test HAVE_C_WNO_SIGN_COMPARE - Success -- ADE: Download: v0.1.1f.zip -- OpenCV Python: during development append to PYTHONPATH: /usr/local/opencv/build/python_loader -- Performing Test HAVE_CXX_WNO_STRICT_ALIASING -- Performing Test HAVE_CXX_WNO_STRICT_ALIASING - Success -- Checking for modules 'libavcodec;libavformat;libavutil;libswscale' -- Found libavcodec, version 57.107.100 -- Found libavformat, version 57.83.100 -- Found libavutil, version 55.78.100 -- Found libswscale, version 4.8.100 -- Checking for module 'libavresample' -- Found libavresample, version 3.7.0 -- WARNING: Can't build ffmpeg test code -- Checking for module 'gstreamer-base-1.0' -- Found gstreamer-base-1.0, version 1.14.5 -- Checking for module 'gstreamer-app-1.0' -- Found gstreamer-app-1.0, version 1.14.5 -- Checking for module 'gstreamer-riff-1.0' -- Found gstreamer-riff-1.0, version 1.14.5 -- Checking for module 'gstreamer-pbutils-1.0' -- Found gstreamer-pbutils-1.0, version 1.14.5 -- Checking for module 'libdc1394-2' -- Found libdc1394-2, version 2.2.5 -- Caffe: NO -- Protobuf: NO -- Glog: YES -- Performing Test HAVE_CXX_WNO_UNUSED_VARIABLE -- Performing Test HAVE_CXX_WNO_UNUSED_VARIABLE - Success -- Performing Test HAVE_CXX_WNO_ENUM_COMPARE -- Performing Test HAVE_CXX_WNO_ENUM_COMPARE - Success -- Checking for module 'freetype2' -- Found freetype2, version 21.0.15 -- Checking for module 'harfbuzz' -- Found harfbuzz, version 1.7.2 -- freetype2: YES (ver 21.0.15) -- harfbuzz: YES (ver 1.7.2) -- HDF5: Using hdf5 compiler wrapper to determine C configuration -- Found HDF5: /usr/lib/x86_64-linux-gnu/hdf5/serial/libhdf5.so;/usr/lib/x86_64-linux-gnu/libpthread.so;/usr/lib/x86_64-linux-gnu/libsz.so;/usr/lib/x86_64-linux-gnu/libz.so;/usr/lib/x86_64-linux-gnu/libdl.so;/usr/lib/x86_64-linux-gnu/libm.so (found version "1.10.0.1") -- Module opencv_ovis disabled because OGRE3D was not found -- No preference for use of exported gflags CMake configuration set, and no hints for include/library directories provided. Defaulting to preferring an installed/exported gflags CMake configuration if available. -- Found installed version of gflags: /usr/lib/x86_64-linux-gnu/cmake/gflags -- Detected gflags version: 2.2.1 -- Checking SFM deps... TRUE -- CERES support is disabled. Ceres Solver for reconstruction API is required. -- Checking for module 'tesseract' -- No package 'tesseract' found -- Tesseract: NO -- Performing Test HAVE_CXX_WNO_UNUSED_BUT_SET_VARIABLE -- Performing Test HAVE_CXX_WNO_UNUSED_BUT_SET_VARIABLE - Success -- HDF5: Using hdf5 compiler wrapper to determine C configuration -- Registering hook 'INIT_MODULE_SOURCES_opencv_dnn': /usr/local/opencv/modules/dnn/cmake/hooks/INIT_MODULE_SOURCES_opencv_dnn.cmake -- Performing Test HAVE_CXX_WNO_OVERLOADED_VIRTUAL -- Performing Test HAVE_CXX_WNO_OVERLOADED_VIRTUAL - Success -- xfeatures2d/boostdesc: Download: boostdesc_bgm.i -- xfeatures2d/boostdesc: Download: boostdesc_bgm_bi.i -- xfeatures2d/boostdesc: Download: boostdesc_bgm_hd.i -- xfeatures2d/boostdesc: Download: boostdesc_binboost_064.i -- xfeatures2d/boostdesc: Download: boostdesc_binboost_128.i -- xfeatures2d/boostdesc: Download: boostdesc_binboost_256.i -- xfeatures2d/boostdesc: Download: boostdesc_lbgm.i -- xfeatures2d/vgg: Download: vgg_generated_48.i -- xfeatures2d/vgg: Download: vgg_generated_64.i -- xfeatures2d/vgg: Download: vgg_generated_80.i -- xfeatures2d/vgg: Download: vgg_generated_120.i -- data: Download: face_landmark_model.dat -- No preference for use of exported gflags CMake configuration set, and no hints for include/library directories provided. Defaulting to preferring an installed/exported gflags CMake configuration if available. -- Found installed version of gflags: /usr/lib/x86_64-linux-gnu/cmake/gflags -- Detected gflags version: 2.2.1 -- Checking SFM deps... TRUE -- CERES support is disabled. Ceres Solver for reconstruction API is required. -- NVIDIA_OPTICAL_FLOW: Download: 79c6cee80a2df9a196f20afd6b598a9810964c32.zip -- Performing Test HAVE_CXX_WNO_UNUSED_PRIVATE_FIELD -- Performing Test HAVE_CXX_WNO_UNUSED_PRIVATE_FIELD - Failed -- Pylint: registered 163 targets. Build 'check_pylint' target to run checks ("cmake --build . --target check_pylint" or "make check_pylint") CMake Warning at cmake/OpenCVGenSetupVars.cmake:54 (message): CONFIGURATION IS NOT SUPPORTED: validate setupvars script in install directory Call Stack (most recent call first): CMakeLists.txt:947 (include) -- -- General configuration for OpenCV 4.2.0 ===================================== -- Version control: unknown -- -- Extra modules: -- Location (extra): /usr/local/opencv/opencv_contrib/modules -- Version control (extra): unknown -- -- Platform: -- Timestamp: 2020-05-17T08:40:53Z -- Host: Linux 5.3.0-51-generic x86_64 -- CMake: 3.17.2 -- CMake generator: Unix Makefiles -- CMake build tool: /usr/bin/make -- Configuration: RELEASE -- -- CPU/HW features: -- Baseline: SSE SSE2 SSE3 -- requested: SSE3 -- Dispatched code generation: SSE4_1 SSE4_2 FP16 AVX AVX2 AVX512_SKX -- requested: SSE4_1 SSE4_2 AVX FP16 AVX2 AVX512_SKX -- SSE4_1 (16 files): + SSSE3 SSE4_1 -- SSE4_2 (2 files): + SSSE3 SSE4_1 POPCNT SSE4_2 -- FP16 (1 files): + SSSE3 SSE4_1 POPCNT SSE4_2 FP16 AVX -- AVX (5 files): + SSSE3 SSE4_1 POPCNT SSE4_2 AVX -- AVX2 (29 files): + SSSE3 SSE4_1 POPCNT SSE4_2 FP16 FMA3 AVX AVX2 -- AVX512_SKX (6 files): + SSSE3 SSE4_1 POPCNT SSE4_2 FP16 FMA3 AVX AVX2 AVX_512F AVX512_COMMON AVX512_SKX -- -- C/C++: -- Built as dynamic libs?: YES -- C++ Compiler: /usr/bin/c++ (ver 7.5.0) -- C++ flags (Release): -fsigned-char -ffast-math -W -Wall -Werror=return-type -Werror=non-virtual-dtor -Werror=address -Werror=sequence-point -Wformat -Werror=format-security -Wmissing-declarations -Wundef -Winit-self -Wpointer-arith -Wshadow -Wsign-promo -Wuninitialized -Winit-self -Wsuggest-override -Wno-delete-non-virtual-dtor -Wno-comment -Wimplicit-fallthrough=3 -Wno-strict-overflow -fdiagnostics-show-option -Wno-long-long -pthread -fomit-frame-pointer -ffunction-sections -fdata-sections -msse -msse2 -msse3 -fvisibility=hidden -fvisibility-inlines-hidden -O3 -DNDEBUG -DNDEBUG -- C++ flags (Debug): -fsigned-char -ffast-math -W -Wall -Werror=return-type -Werror=non-virtual-dtor -Werror=address -Werror=sequence-point -Wformat -Werror=format-security -Wmissing-declarations -Wundef -Winit-self -Wpointer-arith -Wshadow -Wsign-promo -Wuninitialized -Winit-self -Wsuggest-override -Wno-delete-non-virtual-dtor -Wno-comment -Wimplicit-fallthrough=3 -Wno-strict-overflow -fdiagnostics-show-option -Wno-long-long -pthread -fomit-frame-pointer -ffunction-sections -fdata-sections -msse -msse2 -msse3 -fvisibility=hidden -fvisibility-inlines-hidden -g -O0 -DDEBUG -D_DEBUG -- C Compiler: /usr/bin/cc -- C flags (Release): -fsigned-char -ffast-math -W -Wall -Werror=return-type -Werror=non-virtual-dtor -Werror=address -Werror=sequence-point -Wformat -Werror=format-security -Wmissing-declarations -Wmissing-prototypes -Wstrict-prototypes -Wundef -Winit-self -Wpointer-arith -Wshadow -Wuninitialized -Winit-self -Wno-comment -Wimplicit-fallthrough=3 -Wno-strict-overflow -fdiagnostics-show-option -Wno-long-long -pthread -fomit-frame-pointer -ffunction-sections -fdata-sections -msse -msse2 -msse3 -fvisibility=hidden -O3 -DNDEBUG -DNDEBUG -- C flags (Debug): -fsigned-char -ffast-math -W -Wall -Werror=return-type -Werror=non-virtual-dtor -Werror=address -Werror=sequence-point -Wformat -Werror=format-security -Wmissing-declarations -Wmissing-prototypes -Wstrict-prototypes -Wundef -Winit-self -Wpointer-arith -Wshadow -Wuninitialized -Winit-self -Wno-comment -Wimplicit-fallthrough=3 -Wno-strict-overflow -fdiagnostics-show-option -Wno-long-long -pthread -fomit-frame-pointer -ffunction-sections -fdata-sections -msse -msse2 -msse3 -fvisibility=hidden -g -O0 -DDEBUG -D_DEBUG -- Linker flags (Release): -Wl,--gc-sections -- Linker flags (Debug): -Wl,--gc-sections -- ccache: NO -- Precompiled headers: NO -- Extra dependencies: m pthread cudart_static dl rt nppc nppial nppicc nppicom nppidei nppif nppig nppim nppist nppisu nppitc npps cublas cudnn cufft -L/usr/local/cuda-10.0/lib64 -L/usr/lib/x86_64-linux-gnu -- 3rdparty dependencies: -- -- OpenCV modules: -- To be built: aruco bgsegm bioinspired calib3d ccalib core cudaarithm cudabgsegm cudafeatures2d cudafilters cudaimgproc cudalegacy cudaobjdetect cudaoptflow cudastereo cudawarping cudev datasets dnn dnn_objdetect dnn_superres dpm face features2d flann freetype fuzzy gapi hdf hfs highgui img_hash imgcodecs imgproc line_descriptor ml objdetect optflow phase_unwrapping photo plot python2 python3 quality reg rgbd saliency sfm shape stereo stitching structured_light superres surface_matching text tracking ts video videoio videostab xfeatures2d ximgproc xobjdetect xphoto -- Disabled: cudacodec world -- Disabled by dependency: - -- Unavailable: cnn_3dobj cvv java js matlab ovis viz -- Applications: tests perf_tests examples apps -- Documentation: NO -- Non-free algorithms: YES -- -- GUI: -- GTK+: YES (ver 3.22.30) -- GThread : YES (ver 2.56.4) -- GtkGlExt: NO -- OpenGL support: NO -- VTK support: NO -- -- Media I/O: -- ZLib: /usr/lib/x86_64-linux-gnu/libz.so (ver 1.2.11) -- JPEG: /usr/lib/x86_64-linux-gnu/libjpeg.so (ver 80) -- WEBP: build (ver encoder: 0x020e) -- PNG: /usr/lib/x86_64-linux-gnu/libpng.so (ver 1.6.34) -- TIFF: /usr/local/lib/libtiff.so (ver 42 / 4.1.0) -- JPEG 2000: /usr/lib/x86_64-linux-gnu/libjasper.so (ver 1.900.1) -- OpenEXR: build (ver 2.3.0) -- HDR: YES -- SUNRASTER: YES -- PXM: YES -- PFM: YES -- -- Video I/O: -- DC1394: YES (2.2.5) -- FFMPEG: NO -- avcodec: YES (57.107.100) -- avformat: YES (57.83.100) -- avutil: YES (55.78.100) -- swscale: YES (4.8.100) -- avresample: YES (3.7.0) -- GStreamer: YES (1.14.5) -- v4l/v4l2: YES (linux/videodev2.h) -- -- Parallel framework: TBB (ver 2017.0 interface 9107) -- -- Trace: YES (with Intel ITT) -- -- Other third-party libraries: -- Intel IPP: 2019.0.0 Gold [2019.0.0] -- at: /usr/local/opencv/build/3rdparty/ippicv/ippicv_lnx/icv -- Intel IPP IW: sources (2019.0.0) -- at: /usr/local/opencv/build/3rdparty/ippicv/ippicv_lnx/iw -- Lapack: NO -- Eigen: YES (ver 3.3.4) -- Custom HAL: NO -- Protobuf: build (3.5.1) -- -- NVIDIA CUDA: YES (ver 10.0, CUFFT CUBLAS FAST_MATH) -- NVIDIA GPU arch: 61 -- NVIDIA PTX archs: -- -- cuDNN: YES (ver 7.6.5) -- -- OpenCL: YES (no extra features) -- Include path: /usr/local/opencv/3rdparty/include/opencl/1.2 -- Link libraries: Dynamic load -- -- Python 2: -- Interpreter: /usr/bin/python2.7 (ver 2.7.17) -- Libraries: /usr/lib/x86_64-linux-gnu/libpython2.7.so (ver 2.7.17) -- numpy: /usr/lib/python2.7/dist-packages/numpy/core/include (ver 1.13.3) -- install path: lib/python2.7/dist-packages/cv2/python-2.7 -- -- Python 3: -- Interpreter: /home/jeffrey/.virtualenvs/cv/bin/python3 (ver 3.6.9) -- Libraries: /usr/lib/x86_64-linux-gnu/libpython3.6m.so (ver 3.6.9) -- numpy: /home/jeffrey/.virtualenvs/cv/lib/python3.6/site-packages/numpy/core/include (ver 1.18.4) -- install path: /home/jeffrey/.virtualenvs/cv/lib/python3.6/site-packages/cv2/python-3.6 -- -- Python (for build): /usr/bin/python2.7 -- Pylint: /home/jeffrey/.local/bin/pylint (ver: unknown, checks: 163) -- -- Java: -- ant: NO -- JNI: /usr/local/lib/jvm/jdk1.8.0_231/include /usr/local/lib/jvm/jdk1.8.0_231/include/linux /usr/local/lib/jvm/jdk1.8.0_231/include -- Java wrappers: NO -- Java tests: NO -- -- Install to: /usr/local -- ----------------------------------------------------------------- -- -- Configuring done
三、已验证的应用
代码来自于作者的github代码集：
https://github.com/spmallick/learnopencv/tree/master/OpenPose-Multi-Person
代码来自于作者邮件自动发送：
How to use OpenCV’s “dnn” module with NVIDIA GPUs, CUDA, and cuDNN
OpenCV ‘dnn’ with NVIDIA GPUs: 1549% faster YOLO, SSD, and Mask R-CNN
/* continue */