Proceso de configuración de Jetson Nano (4)

Proceso de configuración de Jetson Nano (4)

Este artículo escribe el proceso de compilación e instalación de opencv 4.1.1 en nano y algunos agujeros de llenado.

1. Desinstale la versión anterior de opencv en el sistema.

sudo apt-get purge libopencv*
sudo apt autoremove
sudo apt-get update

2. Instalar dependencias

sudo apt-get install build-essential

sudo apt-get install libglew-dev libtiff5-dev zlib1g-dev libjpeg-dev\ 
	libavcodec-dev libavformat-dev libavutil-dev libpostproc-dev\
	libswscale-dev libeigen3-dev libtbb-dev libgtk2.0-dev pkg-config libpng-dev
	
sudo apt-get install libatlas-base-dev gfortran
sudo apt update
sudo apt upgrade

3. Descarga el código fuente y compila

Después de descargar el código fuente, descomprima OpenCV-4.1.1 y OpenCV_contirb-4.1.1 y colóquelos en el mismo directorio principal opencv4.
Ingrese al directorio opencv-4.1.1 y ejecute el siguiente comando:

Publique el archivo fuente de opencv modificado aquí :

https://pan.baidu.com/s/1FUWfLriHrqMSDr7l61tBBQ Código de archivo : s7ty

mkdir build
cd build

cmake \

-DCMAKE_BUILD_TYPE=Release \
-DBUILD_PNG=OFF \
-DBUILD_TIFF=OFF \
-DBUILD_TBB=OFF \
-DBUILD_JPEG=OFF \
-DBUILD_JASPER=OFF \
-DBUILD_ZLIB=OFF \
-DBUILD_EXAMPLES=OFF \
-DBUILD_opencv_java=OFF \
-DBUILD_opencv_python2=OFF \
-DBUILD_opencv_python3=ON \
-DENABLE_PRECOMPILED_HEADERS=OFF \
-DWITH_OPENCL=OFF \
-DWITH_OPENMP=OFF \
-DWITH_FFMPEG=ON \
-DWITH_GSTREAMER=OFF \
-DWITH_GSTREAMER_0_10=OFF \
-DWITH_CUDA=ON \
-DWITH_GTK=ON \
-DWITH_VTK=OFF \
-DWITH_TBB=ON \
-DWITH_1394=OFF \
-DWITH_OPENEXR=OFF \
-DCUDA_TOOLKIT_ROOT_DIR=/usr/local/cuda-10.0 \
-DCUDA_ARCH_BIN=5.3 \
-DCUDA_ARCH_PTX="" \
-DINSTALL_C_EXAMPLES=OFF \
-DOPENCV_ENABLE_NONFREE=ON \
-DINSTALL_TESTS=OFF \
-DOPENCV_EXTRA_MODULES_PATH=../../opencv_contrib-4.1.1/modules \

..

sudo make -j4
sudo make install

Los siguientes son errores que ocurrieron durante el proceso de compilación, que se han modificado en el paquete de código fuente:

1.需要将如vgg_generated_48.i、face_landmark_model.dat等，由于是外网的原因，
这些文件会出现超时下载不成功的现象，虽然这里下载不成功，不会出现红色的error报错，
但会影响到后面的make操作，因此必须要确保下载成功。我将我编译时需要下载的文件放在了上面的百度云链接中，需要的同学可以自取。

将这些文件下载好放置在随意目录如/home/ryu/Documents/opencv4/installFile中，
然后更改opencv_contrib-4.1.1/modules/xfeatures2d/cmake/download_vgg.cmake和opencv_contrib-4.1.1/modules/face/CMakeLists.txt中的URL地址，
具体修改如下：（installfile为自己创建）

 "file:///home/djj/opencv-4.1.1/installfile/"
 #"https://raw.githubusercontent.com/opencv/opencv_3rdparty/${OPENCV_3RDPARTY_COMMIT}/"

//
编译68%会报错

需要将上述文件拷贝到  opencv_contrib-4.1.1/modules/xfeatures2d/src/

***********************************************************************************
2.#头文件会报错 #include "features2d/test/test_detectors_regression.impl.hpp"

将opencv-4.1.1/modules/features2d/test/文件下的：
test_descriptors_invariance.impl.hpp
test_descriptors_regression.impl.hpp
test_detectors_invariance.impl.hpp
test_detectors_regression.impl.hpp
test_invariance_utils.hpp

拷贝到opencv_contrib-4.1.1/modules/xfeatures2d/test/文件下。
同时，将opencv_contrib-4.1.1/modules/xfeatures2d/test/test_features2d.cpp文件下的：

#include "features2d/test/test_detectors_regression.impl.hpp"
#include "features2d/test/test_descriptors_regression.impl.hpp"
改为：
#include "test_detectors_regression.impl.hpp"
#include "test_descriptors_regression.impl.hpp"

将opencv_contrib-4.1.1/modules/xfeatures2d/test/test_rotation_and_scale_invariance.cpp文件下的：

#include "features2d/test/test_detectors_invariance.impl.hpp" 
#include "features2d/test/test_descriptors_invariance.impl.hpp"
改为：
#include "test_detectors_invariance.impl.hpp"
#include "test_descriptors_invariance.impl.hpp"

*********************************************************
3.修改：
opencv-4.1.1/samples/gpu/surf_keypoint_matcher.cpp下：
#incloud"opencv2/cuda.hpp"
改为绝对路径
#incloud"/home/djj/../../../opencv2/cuda.hpp"
**********************************************************
4.opencv-4.1.1/samples/gpu/CMakeLists.txt
修改成以下代码：
 foreach(sample_filename ${all_samples})
  ocv_define_sample(tgt ${sample_filename} gpu)
  ocv_target_link_libraries(${tgt} ${OPENCV_LINKER_LIBS} ${OPENCV_CUDA_SAMPLES_REQUIRED_DEPS})
 if(HAVE_CUDA AND NOT ANDROID)
  ocv_target_link_libraries(${tgt} ${CUDA_CUDA_LIBRARY})
 endif()
  if(HAVE_opencv_xfeatures2d)
   ocv_target_link_libraries(${tgt} opencv_xfeatures2d)
  endif()

referencia:

Jetson nano-compila e instala OpenCV4.1.1 y OpenCV_contirb-4.1.1 (suplemento para evitar pits)

Jetson nano compila e instala OpenCV4.1.1 y OpenCV_contirb-4.1.1