CDEHP Dataset

• Calibration
• Action Capture
• Convertion
• Annotation

OpenCV

sudo su root
apt install build-essential cmake git libgtk2.0-dev pkg-config libavcodec-dev libavformat-dev libswscale-dev python3-dev python3-numpy libtbb2 libtbb-dev libjpeg-dev libpng-dev libtiff-dev

cd /usr/local/src
wget https://github.com/opencv/opencv/archive/refs/tags/3.3.0.tar.gz -O opencv_3.3.0.tar.gz
wget https://github.com/opencv/opencv_contrib/archive/refs/tags/3.3.0.tar.gz -O opencv_contrib_3.3.0.tar.gz
tar -xzvf opencv_3.3.0.tar.gz
tar -xzvf opencv_contrib_3.3.0.tar.gz
cd opencv-3.3.0/

sed -i "1i\#define AV_CODEC_FLAG_GLOBAL_HEADER (1 << 22)" modules/videoio/src/cap_ffmpeg_impl.hpp
sed -i "2i\#define CODEC_FLAG_GLOBAL_HEADER AV_CODEC_FLAG_GLOBAL_HEADER" modules/videoio/src/cap_ffmpeg_impl.hpp
sed -i "3i\#define AVFMT_RAWPICTURE 0x0020" modules/videoio/src/cap_ffmpeg_impl.hpp
sed -i "854s/PyString/(char *)PyString/" modules/python/src2/cv2.cpp

mkdir build
cd build/
cmake -D CMAKE_BUILD_TYPE=Release -D CMAKE_INSTALL_PREFIX=/usr/local/ -D OPENCV_EXTRA_MODULES=../../opencv_contrib-3.3.0/modules/ -D BUILD_DOCS=ON -D BUILD_EXAMPLES=ON -D WITH_CUDA=OFF ..
make -j8
make install

ldconfig /usr/local/lib

Calibration

1. Use calibration/save_intrinsics.py to get the camera intrinsic parameters

   使用calibration/save_intrinsics.py程序获取深度相机内参

2. Use CelePixel/Calibration-Tools to online calibrate the camera to compute the camera intrinsic parameters.

   使用CelePixel/Calibration-Tools程序在线拍摄标定事件相机内参

3. In MATLAB, print open checkerboardPattern.pdf with the actual size，the length of each checker should be 23mm or 470mm.

   MATLAB打印棋盘open checkerboardPattern.pdf,打印时选择实际大小，方格长度应为23mm或470mm

4. Use the depth and event cameras to capture the calibration board simultaneously.

   同时拍摄深度相机、事件相机

5. Open two 'Camera Calibrator' programs, and set the option of Radial Distortionto be 3 Coefficients

   打开两个Camera Calibrator程序，畸变参数Radial Distortion设为三参数3 Coefficients

6. Calibrate the depth camera and event camera respectively, and remove those unrecognized images.

   分别标定深度相机、事件相机，去除无法识别和部分识别的图片，保持两者一致

7. Output the coordinates

   导出坐标点

8. Computation

   计算