主要是基于opencv, 学习下图像融合,并延伸到视频融合.
Sync opencv repo
当前opencv最新版本为4.6.0
$ git clone https://github.com/opencv/opencv.git -b 4.6.0
##
Build Opencv with Windows
使用wsl是不错的选择.
Build Opencv with Android
compile with ndk
VS Code + NDK
修改cmake tools kit json, 添加android clang编译器
[
{
"name": "Android-Clang",
"environmentVariables": {
"ANDROID_NDK": "path/to/Android/Sdk/ndk/"
},
"compilers": {
"C": "path/to/Android/Sdk/ndk/toolchains/llvm/prebuilt/windows-x86_64/bin/clang.exe",
"CXX": "path/to/Android/Sdk/ndk/toolchains/llvm/prebuilt/windows-x86_64/bin/clang++.exe"
}
},
,
{
"name": "Visual Studio Community 2022 Release - amd64",
"visualStudio": "9b8c874c",
"visualStudioArchitecture": "x64",
"preferredGenerator": {
"name": "Visual Studio 17 2022",
"platform": "x64",
"toolset": "host=x64"
}
}
]
添加workspace的settings json
//.vscode/settings.json
{
"cmake.configureArgs": [
"-DCMAKE_TOOLCHAIN_FILE=${env:ANDROID_NDK}/build/cmake/android.toolchain.cmake",
"-DANDROID_SDK=${env:ANDROID_NDK}/../..",
"-DCMAKE_ANDROID_ARCH_ABI=arm64-v8a",
"-DANDROID_ABI=arm64-v8a",
"-DCMAKE_SYSTEM_VERSION=28",
"-DANDROID_PLATFORM=android-28",
"-DCMAKE_BUILD_TYPE=Release",
"-DBUILD_DOCS=OFF",
"-DBUILD_PERF_TESTS=OFF",
"-DBUILD_TESTS=OFF",
"-DBUILD_opencv_python=OFF",
"-DBUILD_opencv_js=OFF",
"-DBUILD_ANDROID_PROJECTS=OFF",
"-DBUILD_ANDROID_EXAMPLES=OFF",
"-DBUILD_JAVA=OFF"
]
}
for mingw
{
"cmake.configureArgs": [
"-DBUILD_PERF_TESTS=OFF",
"-DBUILD_TESTS=OFF",
"-DBUILD_opencv_python=OFF",
"-DBUILD_opencv_js=OFF",
"-DWITH_CUDA=ON",
"-DWITH_FFMPEG=ON",
"-DOPENCV_FFMPEG_USE_FIND_PACKAGE=ON",
"-DOPENCV_FFMPEG_SKIP_BUILD_CHECK=OFF",
"-DOPENCV_EXTRA_MODULES_PATH=/home/kiah/worktmp/opencv_contrib/modules/"
],
"cmake.generator": "MinGW Makefiles",
"files.associations": {
"iosfwd": "cpp"
}
}
启动cmake configure\build
[cmake] --
[cmake] -- GUI: GTK2
[cmake] -- GTK+: YES (ver 2.24.33)
[cmake] -- GThread : YES (ver 2.72.1)
[cmake] -- GtkGlExt: NO
[cmake] -- VTK support: NO
[cmake] --
[cmake] -- Media I/O:
[cmake] -- ZLib: /usr/lib/x86_64-linux-gnu/libz.so (ver 1.2.11)
[cmake] -- JPEG: /usr/lib/x86_64-linux-gnu/libjpeg.so (ver 80)
[cmake] -- WEBP: build (ver encoder: 0x020f)
[cmake] -- PNG: /usr/lib/x86_64-linux-gnu/libpng.so (ver 1.6.37)
[cmake] -- TIFF: /usr/lib/x86_64-linux-gnu/libtiff.so (ver 42 / 4.3.0)
[cmake] -- JPEG 2000: build (ver 2.4.0)
[cmake] -- OpenEXR: build (ver 2.3.0)
[cmake] -- HDR: YES
[cmake] -- SUNRASTER: YES
[cmake] -- PXM: YES
[cmake] -- PFM: YES
[cmake] --
[cmake] -- Video I/O:
[cmake] -- DC1394: NO
[cmake] -- FFMPEG: YES (find_package)
[cmake] -- avcodec: YES (58.134.100)
[cmake] -- avformat: YES (58.76.100)
[cmake] -- avutil: YES (56.70.100)
[cmake] -- swscale: YES (5.9.100)
[cmake] -- avresample: NO
[cmake] -- GStreamer: NO
[cmake] -- v4l/v4l2: YES (linux/videodev2.h)
[cmake] --
[cmake] -- Parallel framework: pthreads
[cmake] --
[cmake] -- Trace: YES (with Intel ITT)
[cmake] --
[cmake] -- Other third-party libraries:
[cmake] -- Intel IPP: 2020.0.0 Gold [2020.0.0]
[cmake] -- at: /home/kiah/worktmp/opencv/build/3rdparty/ippicv/ippicv_lnx/icv
[cmake] -- Intel IPP IW: sources (2020.0.0)
[cmake] -- at: /home/kiah/worktmp/opencv/build/3rdparty/ippicv/ippicv_lnx/iw
[cmake] -- VA: NO
[cmake] -- Lapack: NO
[cmake] -- Eigen: NO
[cmake] -- Custom HAL: NO
[cmake] -- Protobuf: build (3.19.1)
[cmake] --
[cmake] -- NVIDIA CUDA: YES (ver 11.8, CUFFT CUBLAS)
[cmake] -- NVIDIA GPU arch: 35 37 50 52 60 61 70 75 80 86
[cmake] -- NVIDIA PTX archs:
[cmake] --
[cmake] -- cuDNN: NO
[cmake] --
[cmake] -- OpenCL: YES (no extra features)
[cmake] -- Include path: /home/kiah/worktmp/opencv/3rdparty/include/opencl/1.2
[cmake] -- Link libraries: Dynamic load
[cmake] --
[cmake] -- Python (for build): /usr/bin/python3
[cmake] --
[cmake] -- Java:
[cmake] -- ant: NO
[cmake] -- JNI: NO
[cmake] -- Java wrappers: NO
[cmake] -- Java tests: NO
[cmake] --
[cmake] -- Install to: /usr/local
FFMPEG
使用apt 安装ffmpeg相关库
sudo apt-get install libavdevice-dev
[cmake] -- Video I/O:
[cmake] -- DC1394: NO
[cmake] -- FFMPEG: YES (find_package)
[cmake] -- avcodec: YES (58.134.100)
[cmake] -- avformat: YES (58.76.100)
[cmake] -- avutil: YES (56.70.100)
[cmake] -- swscale: YES (5.9.100)
[cmake] -- avresample: NO
[cmake] -- GStreamer: NO
查询ffmpeg相关库信息
$ dpkg -l libswresample-dev
Desired=Unknown/Install/Remove/Purge/Hold
| Status=Not/Inst/Conf-files/Unpacked/halF-conf/Half-inst/trig-aWait/Trig-pend
|/ Err?=(none)/Reinst-required (Status,Err: uppercase=bad)
||/ Name Version Architecture Description
+++-=======================-========================-============-=========================================================================
ii libswresample-dev:amd64 7:4.4.2-0ubuntu0.22.04.1 amd64 FFmpeg library for audio resampling, rematrixing etc. - development files
$ dpkg -L libswresample-dev
/.
/usr
/usr/include
/usr/include/x86_64-linux-gnu
/usr/include/x86_64-linux-gnu/libswresample
/usr/include/x86_64-linux-gnu/libswresample/swresample.h
/usr/include/x86_64-linux-gnu/libswresample/version.h
/usr/lib
/usr/lib/x86_64-linux-gnu
/usr/lib/x86_64-linux-gnu/libswresample.a
/usr/lib/x86_64-linux-gnu/pkgconfig
/usr/lib/x86_64-linux-gnu/pkgconfig/libswresample.pc
/usr/share
/usr/share/doc
/usr/share/doc/libswresample-dev
/usr/share/doc/libswresample-dev/copyright
/usr/lib/x86_64-linux-gnu/libswresample.so
/usr/share/doc/libswresample-dev/changelog.Debian.gz
CUDA
主要是安装nvidia drivers， cuda toolkit以及sdk
nvidia-smi确认是否成功安装
NDK only
使用r21e版本为例,其它参考.
$ export ANDROID_NDK=/path/to/ndk
$ export ANDROID_SDK_ROOT=/path/to/android-sdk
$ export ANDROID_NATIVE_API_LEVEL=28
$ export STRIP=/path/to/ndk/toolchains/llvm/prebuilt/linux-x86_64/bin/llvm-strip
创建build目录, 并在build 目录执行
$ cmake -DCMAKE_TOOLCHAIN_FILE=/path/to/ndk/build/cmake/android.toolchain.cmake -DANDROID_TOOLCHAIN=clang++ -DANDROID_ABI=arm64-v8a -D CMAKE_BUILD_TYPE=Release -D ANDROID_NATIVE_API_LEVEL=28 -D WITH_CUDA=OFF -D WITH_MATLAB=OFF -D BUILD_ANDROID_EXAMPLES=OFF -D BUILD_DOCS=OFF -D BUILD_PERF_TESTS=OFF -D BUILD_TESTS=OFF -D ANDROID_STL=c++_shared -D BUILD_SHARED_LIBS=ON -D BUILD_opencv_objdetect=OFF -D BUILD_opencv_video=OFF -D BUILD_opencv_videoio=OFF -D BUILD_opencv_features2d=OFF -D BUILD_opencv_flann=OFF -D BUILD_opencv_highgui=ON -D BUILD_opencv_ml=ON -D BUILD_opencv_photo=OFF -D BUILD_opencv_python=OFF -D BUILD_opencv_shape=OFF -D BUILD_opencv_stitching=OFF -D BUILD_opencv_superres=OFF -D BUILD_opencv_ts=OFF -D BUILD_opencv_videostab=OFF -DBUILD_ANDROID_PROJECTS=OFF ..
-G ninja
进行编译
$ make -j [nproc]
或者是使用ninja
$ninja -j [nproc]
Opencv image stitch
https://docs.opencv.org/4.x/d1/d46/group__stitching.html
GTK 2.0 install
./build/bin/example_cv_read_image data/pic/test_monkey.jpg
[ INFO:0@0.004] global opencv/modules/highgui/src/registry.impl.hpp (114) UIBackendRegistry UI: Enabled backends(3, sorted by priority): GTK(1000); GTK3(990); GTK2(980)
OpenCV(4.6.0) Error: Unspecified error (The function is not implemented. Rebuild the library with Windows, GTK+ 2.x or Cocoa support. If you are on Ubuntu or Debian, install libgtk2.0-dev and pkg-config, then re-run cmake or configure script) in cvShowImage, file /home/kiah/worktmp/opencv/modules/highgui/src/window.cpp, line 1267
terminate called after throwing an instance of 'cv::Exception'
what(): OpenCV(4.6.0) /home/kiah/worktmp/opencv/modules/highgui/src/window.cpp:1267: error: (-2:Unspecified error) The function is not implemented. Rebuild the library with Windows, GTK+ 2.x or Cocoa support. If you are on Ubuntu or Debian, install libgtk2.0-dev and pkg-config, then re-run cmake or configure script in function 'cvShowImage'
Aborted (core dumped)
sudo apt-get install libgtk2.0-dev并重新构建opencv 库
文档信息
- 本文作者：Kiah Zhao
- 本文链接：/2022/10/09/cv_video_stitching/
- 版权声明：自由转载-非商用-非衍生-保持署名（创意共享3.0许可证）
