# CloudControlNet

**Repository Path**: YuePanEdward/CloudControlNet

## Basic Information

- **Project Name**: CloudControlNet
- **Description**: No description available
- **Primary Language**: C++
- **License**: GPL-3.0
- **Default Branch**: master
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 2
- **Created**: 2020-03-21
- **Last Updated**: 2020-12-20

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# Cloud Control Net
### Cloud Control Net (CCN): a robust and efficient framework for multiple source point cloud adjustment and fusion

### Under development

### How to use 

1. Install dependent 3rd libraries 

[PCL(>=1.7)](https://github.com/PointCloudLibrary/pcl), [LibLas(Optional for las data IO)](https://github.com/libLAS/libLAS), [Proj4(Optional for geo-projection)](https://proj.org/), [g2o(<=2017version)](https://github.com/RainerKuemmerle/g2o/releases/tag/20170730_git) or [ceres](http://ceres-solver.org/)

```
#you may run the following shell file to install all the dependent libs
sh script/install_dep_lib.sh
```

2. Compile
```
mkdir build
cd build
cmake ..
make 
cd ..
```

3. Prepare data

Use your own data or use the [ISPRS Multiplatform Very High Resolution Photogrammetry Dataset](https://github.com/YuePanEdward/CloudControlNet/blob/master/assets/download_isprs_mpp_dataset.pdf) to test the project.

The data folder structure should be as following:
```
_____ALS_foler
     |___ALS_strip_00.las (single-scanline , optional)
     |___ALS_strip_00
     .   |____ALS_block_0000.las
     .   |____ALS_block_0001.las
     .   |____ALS_block_.....las
     |___ALS_strip_00_dataconfig.txt (optional)
     |___ALS_strip_01.las (single-scanline , optional)
     |___ALS_strip_01
     .   |____ALS_block_0100.las
     .   |____ALS_block_0101.las
     .   |____ALS_block_.....las
     |___ALS_strip_01_dataconfig.txt (optional)
     |___ALS_strip_02.las (single-scanline , optional)
     |___ALS_strip_02
     .   |____ALS_block_.....las
     |___ALS_strip_02_dataconfig.txt (optional)
     |___ALS_strip...

_____MLS_foler
     |___MLS_strip_00.las (single-scanline , optional)
     |___MLS_strip_00
     .   |____MLS_block_0000.las
     .   |____MLS_block_0001.las
     .   |____MLS_block_.....las
     |___MLS_strip_00_dataconfig.txt (optional)
     |___MLS_strip_00_station_pose.txt (multi-scanline , optional)
     |___MLS_strip_01.las (single-scanline , optional)
     |___MLS_strip_01
     .   |____MLS_block_0100.las
     .   |____MLS_block_0101.las
     .   |____MLS_block_.....las
     |___MLS_strip_01_dataconfig.txt (optional)
     |___MLS_strip_01_station_pose.txt (multi-scanline , optional)
     |___MLS_strip...

_____BPLS_foler
     |___BPLS_strip_00.las (single-scanline , optional)
     |___BPLS_strip_00
     .   |____BPLS_block_0000.las
     .   |____BPLS_block_0001.las
     .   |____BPLS_block_.....las
     |___BPLS_strip_00_dataconfig.txt (optional)
     |___BPLS_strip_00_station_pose.txt (multi-scanline , optional)
     |___BPLS_strip...

_____TLS_folder_base
     |___ TLS_folder 
     .    |____TLS_station_0000.las
     .    |____TLS_station_0001.las
     .    |____TLS_station_.....las
     |____TLS_folder_dataconfig.txt (optional)
     |____TLS_folder_station_position.txt (geo-referenced with gnss-spp/ppp/rtk, optional)

```

logs and results folder:

```
_____log
     |____test

_____project_foler
     |____registration_results_experiment_id.txt
     |____pose_result_experiment_id.txt
     |____constraints_experiment_id.txt
     |____...

```

4. configure the data path and parameters

```
#in ./script/run_ccn.sh

#data path
ALS_data_path=ALS_foler;
TLS_data_path=TLS_foler;
MLS_data_path=MLS_folder;
BPLS_data_path=BPLS_folder;
project_folder=project_foler;
log_folder=log_folder;

```

5. Run
```
# for the cloud control net, you can run
sh script/run_ccn.sh

# if you'd like to test the cross-template fine registration module
# edit the script/test_unit_reg.sh file, specify the data path and then run
sh script/test_unit_reg.sh
```

## Cross-template point cloud fine registration

#### Geometric feature points rough classification
![alt text](assets/compare2.png)

Left: ALS , Right: TLS

#### Multiple Metrics Linear Least Sqaure Iterative Closest Points (MMLLS-ICP)
![alt text](assets/compare1.png)

Left: Before registration, Right: After registration

## Cloud Control Net (Pose graph optimization)

### Framework
  ![alt text](assets/framework.jpg)

### Application 1： ALS Point Cloud Refinement
 ALS refinement using TLS or MLS data for highway expansion and reconstruction egineering

 ![alt text](assets/1.jpg)
 Generated Pose Graph (Red: ALS block, Green: TLS station, Blue: MLS block, Yellow: BPLS block, Cyan: Adjacent Smooth Edge, Purple: Overlapping Registration Edge)
 
 ![alt text](assets/2.jpg)
 ALS correction result: (a) Before optimization (b) After optimization
 

 ### Application 2：Multi-Source Point Cloud Fusion
  ![alt text](assets/4.jpg)
  Multi-template point cloud data fusion result (a) TLS (b) TLS+BPLS (c)TLS+BPLS+MLS (d)TLS+BPLS+MLS+ALS