gl_depth_sim - GitLite

KO | EN

gitlite — search

#javascript #typescript #golang #android #machine-learning #rust #nodejs #java #windows #awesome #kubernetes #vue #pytorch #mcp

gl_depth_sim

(★ 82)

Stand-alone depth camera simulation using opengl for hardware acceleration

File Explorer

Download Latest Version (.zip)

- - main.yml
- depth.gif
- laser.gif
- - camera_properties.h
  - glfw_guard.h
  - mesh.h
  - mesh_loader.h
  - renderable_mesh.h
  - shader_program.h
  - sim_depth_camera.h
  - sim_laser_scanner.h
- laser_example.launch
- laser_example.rviz
- - glad.c
  - glfw_guard.cpp
  - mesh.cpp
  - mesh_loader.cpp
  - renderable_mesh.cpp
  - shader_program.cpp
  - sim_depth_camera.cpp
  - sim_laser_scanner.cpp
- - opencv_interface.cpp
  - pcl_interface.cpp
- camera_orbit_example.cpp
- camera_ros_example.cpp
- laser_example.cpp
- usage_example.cpp
- cylinder_r1.stl
- stanford_dragon.stl
- utest.cpp
CMakeLists.txt
LICENSE.txt
package.xml
README.md

# Installation Guide

1. Get the code

git clone https://github.com/Jmeyer1292/gl_depth_sim

Downloads the entire project code from GitHub to your computer.

cd gl_depth_sim

Moves into the project folder you just downloaded.

2. Official Install Script

Easy Recommended

Prerequisites

APT (Debian/Ubuntu 계열) Built into Debian/Ubuntu-based Linux distributions.

sudo apt install libglfw3-dev

Installs directly from the APT package repository (Debian/Ubuntu-based).

✅ After installing, open a new terminal and run the program's version command (e.g. --version) to confirm it worked.

Pulled directly from this repo's README.

3. CMake

Medium

Prerequisites

Git Needed to download the project code from GitHub.
CMake The tool used to generate build configuration.
C/C++ 컴파일러 Windows needs Visual Studio (Community edition, free), macOS needs Xcode Command Line Tools, Linux needs the gcc/g++ package.

mkdir build && cd build

Creates a folder to hold the build output and moves into it.

cmake ..

Analyzes the source code and generates build configuration files (must be run inside the build folder).

make

Compiles the code based on the generated build configuration to produce an executable.

✅ Check that an executable was created inside the build folder, then run it directly (e.g. ./build/app_name).

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# GL Depth Simulator
This library simulates an organized depth camera and laser scanner by reading back a depth buffer after calling OpenGL to render a scene.
I tried to write this package so that the core library had only a couple of dependencies and could stand on its own. The primary use case, however, is simulating depth data and laser scan data WITH hardware acceleration and WITHOUT needing to use Gazebo for ROS robotics development.

## Installation
### Dependencies
The core library requires:
 - GLFW (for GL context creation/windowing)
 - Eigen (linear algebra)
 - Assimp (for model loading in examples)

I provide extensions for using the resulting depth data (and thus have dependencies) with:
 - OpenCV (cv::Mat in both floating point and unsigned-16 formats)
 -  PCL (PointCloud\<PointXYZ>)

If you have ROS, you have everything but GLFW already. To get running:
```
sudo apt install libglfw3-dev
```
## Depth Camera Example
![Stanford Dragon](https://raw.githubusercontent.com/Jmeyer1292/gl_depth_sim/master/docs/depth.gif)
This scene shows the classic Stanford Dragon seen by a depth camera orbiting around the model. Rviz is used to display the clouds. This was taken directly from the ros example in the src/ directory. While the GIF is slow, the scan data is published at hundreds of frames per second.

You can run this example in ROS by building this package in your workspace and running:
```
rosrun gl_depth_sim ros_example _mesh:=<PATH_TO_YOUR_MESH>
```

You can also set the `_z` and `_radius` parameters.

## Laser Scanner Example
![Stanford Dragon](https://raw.githubusercontent.com/Jmeyer1292/gl_depth_sim/master/docs/laser.gif)
This scene shows a line-scan of the same Stanford Dragon produced by a laser scanner with 360 degree field-of-view travelling of the surface of the mesh.
Under the hood, a GL depth camera renders the scene 3 times at 120 degree rotational offsets to produce a single line of points in the simulated laser scan with 360 degree field of view.
Various properties of the laser scanner, such as minimum and maximum range and angular resolution, are configurable.

You can run this example in ROS by building this package in your workspace and running:
```
roslaunch gl_depth_sim laser_example.launch
```

## Usage
To use in a ROS context,  add a catkin dependency on `gl_depth_sim` and follow the idea of the following example:
```c++
#include "gl_depth_sim/sim_depth_camera.h"
#include "gl_depth_sim/mesh_loader.h"
#include "gl_depth_sim/interfaces/pcl_interface.h" // For converting to PCL cloud

int main(int argc, char** argv)
{
  // Create a camera properties structure that defines the intrinsic
  // properties of the camera
  gl_depth_sim::CameraProperties props;
  props.width = 640; // In pixels
  props.height = 480;
  props.fx = 550.0f; // In pixels
  props.fy = 550.0f;
  props.cx = props.width / 2; // In pixels
  props.cy = props.height / 2;
  props.z_near = 0.25f; // In "world units"
  props.z_far = 10.0f;

// Create the depth camera itself
  gl_depth_sim::SimDepthCamera sim (props);

// Use convienence function to load mesh from a file. Returns nullptr on failure.
  std::unique_ptr<gl_depth_sim::Mesh> mesh_ptr = gl_depth_sim::loadMesh(argv[1]);

// Adds mesh to the renderable scene at the origin
  sim.add(*mesh_ptr, Eigen::Isometry3d::Identity());

// Define a camera location from which you want to "take a picture"
  // Uses ROS conventions for optical frames (+Z down the camera field of view, Y down the image)
  Eigen::Isometry3d camera_pose = Eigen::Isometry3d::Identity();
  camera_pose.translation() = Eigen::Vector3d(-1, 0, 0);

// Render the image and download the depth data
  gl_depth_sim::DepthImage depth_data = sim.render(camera_pose);

// Lastly, convert it to a point cloud and do something with it...
  pcl::PointCloud<pcl::PointXYZ> cloud;
  gl_depth_sim::toPointCloudXYZ(props, depth_data, cloud);

// Your code here...
```

## Issues
 1. Needs better scene management. Add, remove, move, etc...
 2. Needs qualification for depth accuracy. GPU depth buffers are fairly limited and have precision problems at distance. There are techniques to manage this and I'd like to add them
 3. I'd like to support color image generation too. It shouldn't be too bad but we need to have meshes with texture.