roboticschen / sunnycamera

Driver program for Sunny_Camera, based on C++, with python interface

Sunny camera driver

Requirements

# opencv 
git clone https://github.com/opencv/opencv
cd opencv
mkdir build && cd build
cmake ..
make -j12
sudo make install
cd ..

# pybind （ when python shared libraries is wanted）
git clone https://github.com/pybind/pybind11.git
cd pybind11
mkdir build && cd build
cmake ..
make -j12
sudo make install
cd ..

Installation

git clone https://github.com/RoboticsChen/Sunny_camera.git
cd Sunny_camera
chmod +x *.sh

C++ shared libraries only

. build.sh         # default C++ shared libraries only.
sudo ./install.sh

C++ shared libraries and python shared libraries

conda activate your_python_env  
. build.sh -p     # -p option to build Python shared libraries additionally.
sudo ./install.sh
pip install -e .

Note: ● Use command (. build.sh) instead of (./build.sh), otherwise you should open a new terminal for path set activate. ●If build fails, check whether python_env is the desired environment for installing the driver.

demos

Use following command to verify installation

C++ version

cd demo_cpp
mkdir build && cd build
cmake ..
make -j12
./demo_cpp

Python version

cd demo_python
python demo_singleCam.py

remember to specify camera_id accordingly

C++ API Reference

namespace

• SunnyCamera::

Functions

• void device_init() : search and open searched sunny cameras
• void device_deinit() : close all openned sunny cameras

Camera

• Attributes:
  • int WIDTH(std::string img_type): The horizontal width (x-axis) of RGB and depth camera in pixels.
    • The "img_type" can be either "rgb" or "depth," with the default value "rgb" (str type).
  • int HEIGHT(std::string img_type): The vertical height (y-axis) of RGB and depth camera in pixels.
  • cv::Mat INTRINSIC(std::string img_type): The 3x3 intrinsic matrix of the cameras.
  • cv::Mat DISTORTION(std::string img_type): The 1x4(or 1x5) distortion vector of the cameras.
• Methods:
  • Camera(std::string camera_id): Initializes a camera object with the provided camera ID.
  • ~Camera(): Releases the resources occupied by the camera object
  • bool init(): Initialize the backend of the AirBot camera.True when successful, otherwise False.
  • bool deinit(): Uninitialize the camera.
  • cv::Mat get_rgb(): Capture and return the RGB image(H*W*3).
  • std::map<std::string, cv::Mat> get_tof(): Capture and return the captured TOF data.
    • "depth_image": cv::Mat - The depth image(H*W).
    • "aligned_rgb": cv::Mat - The aligned RGB image(H*W*3).
    • "point_cloud": cv::Mat - The point cloud data(N*3).

Python API Reference

Functions

• device_init() : search and open searched sunny cameras
• device_deinit() : close all openned sunny cameras

Camera

• Attributes:
  • WIDTH(img_type): The horizontal width (x-axis) of RGB and depth camera in pixels.
    • The "img_type" can be either "rgb" or "depth," with the default value "rgb" (str type).
  • HEIGHT(img_type): The vertical height (y-axis) of RGB and depth camera in pixels.
  • INTRINSIC(img_type): The 3x3 intrinsic matrix of the cameras.
  • DISTORTION(img_type): The 1x4(or 1x5) distortion vector of the cameras.
• Methods:
  • __init__(self, camera_id): Initializes a camera object with the provided camera ID (str type).
  • init(self): Initialize the backend of the AirBot camera. True when successful, otherwise False.
  • deinit(self): Uninitialize the camera.
  • get_rgb(self): Capture and return the color image.
    • Returns: np.ndarray - The captured RGB image(H*W*3).
  • get_tof(self): Capture and return the depth image.
    • Returns: Dict[str, np.ndarray] - The captured TOF data.
      • "depth_image": np.ndarray - The depth image(H*W).
      • "aligned_rgb": np.ndarray - The aligned RGB image(H*W*3).
      • "point_cloud": np.ndarray - The point cloud data(N*3).