Cross-platform ctypes/Cython wrapper to the librealsense C-library version 1.x. This wrapper is useful for legacy models such as SR300, F200 and R200.
OBS: there is no plan to support librealsense 2.x, as Intel already provides the Python binding through pyrealsense2.
-
install librealsense and run the examples.
-
install the dependencies: pyrealsense uses pycparser for extracting necessary enums and structures definitions from the librealsense API, Cython for wrapping the inlined functions in the librealsense API, and Numpy for generic data shuffling.
-
Windows specifics: set the environment variable PYRS_INCLUDES to the
rs.hdirectory location and the environment variable PYRS_LIBS to the librealsense binary location. You might also need to havestdint.havailable in your path.
from PyPI - (OBS: not always the latest):
pip install pyrealsense
from source:
python setup.py install
## setup logging
import logging
logging.basicConfig(level = logging.INFO)
## import the package
import pyrealsense as pyrs
## start the service - also available as context manager
serv = pyrs.Service()
## create a device from device id and streams of interest
cam = serv.Device(device_id = 0, streams = [pyrs.stream.ColorStream(fps = 60)])
## retrieve 60 frames of data
for _ in range(60):
cam.wait_for_frames()
print(cam.color)
## stop camera and service
cam.stop()
serv.stop()
The server for Realsense devices is started with pyrs.Service() which will printout the number of devices available. It can also be started as a context with with pyrs.Service():.
Different devices can be created from the service Device factory. They are created as their own class defined by device id, name, serial, firmware as well as enabled streams and camera presets. The default behaviour create a device with id = 0 and setup the color, depth, pointcloud, color_aligned_depth, depth_aligned_color and infrared streams.
The available streams are either native or synthetic, and each one will create a property that exposes the current content of the frame buffer in the form of device.<stream_name>, where <stream_name> is color, depth, points, cad, dac or infrared. To get access to new data, Device.wait_for_frames has to be called once per frame.
## with connected device cam
from pyrealsense import offline
offline.save_depth_intrinsics(cam)
## previous device cam now offline
from pyrealsense import offline
offline.load_depth_intrinsics('610205001689') # camera serial number
d = np.linspace(0, 1000, 480*640, dtype=np.uint16)
pc = offline.deproject_depth(d)
The module offline can store the rs_intrinsics and depth_scale of a device to disk by default in the user's home directory in the file .pyrealsense. This can later be loaded and used to deproject depth data into pointcloud, which is useful to store raw video file and save some disk memory.
The wrapper comes with some examples. You will need to install matplotlib, opencv and VTK to run everything. You can also look at the Jupyter Notebook.
To this point, this wrapper has been tested with:
- librealsense v1.12.1
- Ubuntu 16.04 LTS, Mac OS X 10.12.2 w/ SR300 camera
- Mac OS X 10.12.3 w/ R200 camera
- Windows 10
The offline module only supports a single camera.
Ubuntu Trusty, python 2 and 3:
Contributions are always welcome. Make sure to push to the dev branch.
This project has been developed as part of the E.U. Horizon 2020 research and innovation project Moregrasp (award number 643955).
