An error occurred when installing pyrealsense2 on ubuntu
I want to use the D435i lens on nvidia tx2. At present, I am trying to install pyrealsense2, but no matter it is python2.7 or python 3.7, I can’t install it. What might be the problem?
This is the error I encountered
-
Hi B2992872 pip install of pyrealsense2 will not work on Jetson because it has an Arm processor chip, and the pyrealsense2 pip packages are for x86 / x64 processors like the ones in PCs. The pyrealsense2 wrapper can be built from source code on Jetson.
https://github.com/IntelRealSense/librealsense/tree/master/wrappers/python#building-from-source
-
Have you got the CMake program installed please?
-
I believe that you go to the root librealsense directory of the librealsense SDK and do mkdir build && cd build to create a sub-folder named build (librealsense > build). The command then changes the directory so that you are inside the build folder.
Then input the CMake build command whilst in that build folder:
cmake ../ -DBUILD_PYTHON_BINDINGS:bool=true
Then after that, input these commands:
make -j4
sudo make install
-
The error message suggests that you are launching the build command from the /home/Nvidia/build directory. If that is the case then you are not at the correct directory path (the instructions for building the pyrealsense2 wrapper from source are admittedly not clear about this).
If you have the librealsense SDK built then you should have a directory called librealsense or something similar. This is the folder that the SDK files are in. So you should enter this directory and use the command mkdir build && cd build to create a 'build' directory and navigate to it. Then input the CMake build command.
-
To the best of my knowledge, once CMake is installed somewhere on your computer then it should be accessible from anywhere. You can download it for Linux as a pre-built "binary" file (cmake-3.18.2-Linux-x86_64.tar.gz) that just installs when you run it, so you do not need to build CMake from source code.
-
Hi Marty
Here is my requirement
1. Initially the IR emitter will be turned off
2. I need to then turn on the emitter and get depth frame
3. Turn off the emitter and get IR frame
I believe when we call pipeline.wait_for_frames(), we will get all the enabled streams with either emitter on or off
Could u please tell me whether it is possible to get each individual stream with different emitter options ie ON/OFF??
Thanks in advance
-
Hi Sheethal Ng The IR Emitter (projector) is a separate physical camera component from the IR imager components. It can be enabled / disabled independently of the current status of the depth, color and IMU streams. So the IR Emitter is either On or Off, and you cannot set an individual on / off state for each individual stream type.
If you are needing to remove the IR dot pattern projection from the IR image, the D415 camera model provides a Visual Preset (json) configuration file called Left Imager Color w/o IR Pattern that removes the pattern. This pattern removal feature is also coming to the new D455 model in an upcoming firmware version.
https://github.com/IntelRealSense/librealsense/tree/master/examples/sensor-control
The visibility of the dot pattern can also be reduced whilst keeping the IR Emitter on by reducing the value of the Laser Power option. This can make the depth image more sparse in detail though.
If you prefer to toggle the IR Emitter on-off then you can find example code to do so in the link below, under the librealsense2 heading:
https://github.com/IntelRealSense/librealsense/wiki/API-How-To#controlling-the-laser
Please sign in to leave a comment.
Comments
15 comments