# Static-SLAM-With-PSPnet

**Repository Path**: coffinian/Static-SLAM-With-PSPnet

## Basic Information

- **Project Name**: Static-SLAM-With-PSPnet
- **Description**: No description available
- **Primary Language**: Unknown
- **License**: Not specified
- **Default Branch**: master
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 1
- **Forks**: 0
- **Created**: 2020-06-20
- **Last Updated**: 2022-03-06

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

## Dynamic SLAM based on PSPNet/ICNet
This is a refined version of [Yilei's work](https://github.com/yilei0620/RGBD-Slam-Semantic-Seg-DeepLab). 
We seperately test [PSPNet](https://github.com/hszhao/PSPNet) and [ICNet](https://github.com/hszhao/ICNet) to recognize humans from input images and then mask them out.
For 3D map construction, only the unmasked areas are projected into pointcloud.  

Also since the segmentation's qualities can be pretty poor on blurried images, we first used a dilation filter to slightly
amplify the mask. Some stubborn pixels may still exist after the dilation, a points filter is also employed.

## Installation
The package depends on 
[Boost](https://www.boost.org/),
[Opencv3](https://opencv.org/opencv-3-3.html), 
[PCL1.7](http://mobile.pointclouds.org/http://mobile.pointclouds.org/news/2013/07/23/pcl-1.7/),
[caffe](http://caffe.berkeleyvision.org/)
and [G2O](https://github.com/RainerKuemmerle/g2o). Please follow the official guidance for installation.

For compiling, please follow the <br>
`mkdir build` <br>
`cd build` <br>
`cmake..` <br>
`make`<br>
process. Also please notice that the package only provides CPU version.

## Test
The code has been tested on Ubuntu 16.04. For compling, please `mkdir model` under the generated `bin` folder.
Then copy the `parameters.txt` into `bin` and the `.caffemodel` and `.prototxt` into `bin/model`.
The `parameters.txt` is used for adjusting parameters of the SLAM system and they can be kept as the same for different method. Since the model files are too large, you need to download it from the original PSPNet/ICNet repositories. <br>

Then run `./ICSlam` for the refined result and `./Slam` for the original one.

## Result
#### Origin
![image](https://github.com/SILI1994/static-SLAM-based-on-PSPnet/blob/master/1.png)

#### Refined
![image](https://github.com/SILI1994/static-SLAM-based-on-PSPnet/blob/master/2.png)