CUDACameraTrackingMultiResRGB.cu is missing

[grungeguitarist]

CUDACameraTrackingMultiResRGB.cu is missing and thus renderCorrespondenceCUDA() is not working.

However, the file from NormalFusion is working: https://github.com/KAIST-VCLAB/normal-fusion/blob/master/Source/CUDACameraTrackingMultiResRGB.cu

[grungeguitarist]

BTW, here is a step-by-step installation that worked for me...:

1. Make original VoxelHashing running

• Download from https://github.com/niessner/VoxelHashing/tree/master/DepthSensingCUDA
• Install Visual Studio 2013 Community
• Install CUDA 7.5
• Make sure that everything works for VoxelHashing

	
2. Insert TextureFusion

- Download from https://github.com/KAIST-VCLAB/texturefusion
- Copy all files to the VoxelHashing root folder "DepthSensingCUDA" and overwrite all existing files
- Open .sln in Visual Studio 2013
- In VS2013, replace content of "DepthSensing.cpp" and "DepthSensing.h" with content of downloaded texFusion.cpp" and "DepthSensing.h", respectively
- In VS2013, rename "DepthSensing.cpp" to "texFusion.cpp" and "DepthSensing.h" to "texFusion.h"
- Drag and drop the following downloaded files to the root (DepthSensing) in the Solution Explorer of VS2013
	- GlobalWarpingState.cpp
	- GlobalWarpingState.h
	- modeDefine.h
	- texFusion_main.cpp
	- texturePool.h
	- TexUpdateUtil.h
- Drag and drop the following downloaded files to the "CUDA" dir in VS2013
	- CUDABuildLinearSystemLocalRGB.cpp
	- CUDABuildLinearSystemLocalRGB.cu
	- CUDABuildLinearSystemLocalRGB.h
	- CUDABuildLinearSystemRGB.cpp
	- CUDABuildLinearSystemRGB.cu
	- CUDABuildLinearSystemRGB.h
	- CUDACameraTrackingLocalRGB.cpp
	- CUDACameraTrackingLocalRGB.cu
	- CUDACameraTrackingLocalRGB.h
	- CUDACameraTrackingMultiResRGB.cpp
	- CUDACameraTrackingMultiResRGB.h
	- cudaDebug.h
	- CUDAImageUtil.cu
	- CUDAImageUtil.h
	- CUDATexPoolParams.h
	- CUDATexUpdate.cpp
	- CUDATexUpdate.cu
	- CUDATexUpdate.h
	- CUDATexUpdateParams.h
- Drag and drop the following downloaded files to the "Sensors" dir in VS2013
	- SensorDataReaderZhou.cpp
	- SensorDataReaderZhou.h
	- sensordatazhou.cpp
	- sensordatazhou.cpp
- One file is missing, copy it from the follow-up paper NormalFusion
	- Download CUDACameraTrackingMultiResRGB.cu from https://github.com/KAIST-VCLAB/normal-fusion/blob/master/Source/CUDACameraTrackingMultiResRGB.cu
	- Move the file to the "Source" dir
	- Drag and drop the file to the "CUDA" dir in VS2013

3. Install required libs
- Install CUDA 10.1
- Install OpenCV 3.4.1
	- Downlaoad from https://github.com/opencv/opencv/archive/3.4.1.zip
	- Open Cmake
		- source code: C:/opencv-3.4.1
		- build the binaries: C:/opencv-3.4.1/build
		- configure
		- choose Visual Studio 2013 (!)
		- BUILD_opencv_world: check
		- WITH_CUDA: check
		- Change path of CUDA_TOOLKIT_ROOT_DIR to "10.1" version if a different one was chosen by Cmake
		- Download Eigen 3.3.4 https://gitlab.com/libeigen/eigen/-/releases/3.3.4
		- Change path of EIGEN_INCLDE_PATH to the downloaded "3.3.4" version 
		- configure
		- generate
		- open project in VS 2013
		- Change to Release Mode
		- right click on INSTALL and "Build"
		
4. Link required libs
	- Right click on DepthSensing -> Build Dependencies -> Build Customizations...
		- Uncheck CUDA 7.5
		- Check CUDA 10.1
	- Right click on DepthSensing -> Proporties -> Configuration Proporties
	- Add "C:\opencv-3.4.1\build\install\include" to...
		- VC++ Directories -> Include Directories 
		- C/C++ -> Additional Include Directories
		- CUDA C/C++ -> Additional Include Directories
	- Add "C:\opencv-3.4.1\build\install\x64\vc12\lib" to...
		- Linker -> Additional Library Directories
	- Add "C:\opencv-3.4.1\build\install\x64\vc12\lib\opencv_world341.lib" to...
		- Linker -> Input -> Additional Dependencies
	- Add "C:\opencv-3.4.1\build\install\x64\vc12\bin" to...
		- Path of Environment Variables on Windows
	
5. Run
	- Change to Release Mode in VS2013
	- Download fountain data set from http://vclab.kaist.ac.kr/cvpr2020p1/fountain_all.zip
	- Put content to D:/dataset/fountain_all
	- Run in VS2013
	- Press 9 when reconstructing was finished to generate .obj file