The apply_filter() always causes a Segmentation fault (core dumped) when called.

[DMing4]

Hi, I am using the RealSense D435i on Ubuntu 18.04.2.

I tried to rebuild the rs-capture example in order to extend it.

I copied below the code. It prints indx:0 and then crashes.

If I run the commented part it works fine.

I installed the librealsense packages: the same rs-capture binary works, so it is not due to the firmware or the device.

Do I have to add any symbol at compile time (-D)?

I tried to rebuild the library and I found that the compile options are different.

Any idea?

Thanks in advance.

Diego

#include <iostream>

#include <librealsense2/rs.hpp> // Include RealSense Cross Platform API

int main(int argc, char * argv[]) try

{

 rs2::log_to_console(RS2_LOG_SEVERITY_ERROR);

 rs2::colorizer color_map;

 rs2::pipeline pipe;

 pipe.start();

 for (int indx=0; indx<10; ++indx)

 {  

  std::cout <<"indx:" <<indx <<' ' <<std::endl;

  rs2::frameset data = pipe.wait_for_frames().  // Wait for next set of frames from the camera

                        apply_filter(color_map); // Find and colorize the depth data

  std::cout <<"Size:" <<data.size() <<std::endl;

  /*

  std::cout <<"indx:" <<indx <<' ';

  rs2::frameset data = pipe.wait_for_frames();   // Print each enabled stream frame rate

  std::cout <<"Size:" <<data.size() <<std::endl;

  */

 }

 pipe.stop();

 return 0;

}

catch (const rs2::error & e)

{

  std::cerr << "RealSense error calling " << e.get_failed_function() << "(" << e.get_failed_args() << "):\n  " << e.what() << std::endl;

  return EXIT_FAILURE;

}

catch (const std::exception& e)

{

  std::cerr << e.what() << std::endl;

  return EXIT_FAILURE;

}

[MartyG]

The default set of instructions for building Librealsense are these ones:

https://github.com/IntelRealSense/librealsense/blob/development/doc/distribution_linux.md

There are a different set of instructions that only need to be used if you need to build the project from sources and prepare / patch the OS manually.

https://github.com/IntelRealSense/librealsense/blob/master/doc/installation.md

[DMing4]

Hi MartyG,

thank you for your prompt response.

I rebuilt the library following the instructions.

The result is the same both using the downloaded packages and rebuilding the library.

The only way I made it working is by using the CMakeList.txt files from the librealsense source tree and making sure that my source files are built using the same compile flags.

I am going to wrap the functionalities that I need in a library that I will integrate in my project.

Best regards,

Diego

[MartyG]

Thanks for the update on your progress!

Intel have released documentation for another method for setting up Librealsense on installations that have problems. You may find it useful.

​https://forums.intel.com/s/question/0D50P00004EmJPySAN/setting-up-librealsense-on-linux-versions-with-installation-problems-by-using-libuvcbackend