realsense viewer output into fullscreen of modify example code to have set depth map

• MartyX Grover

  • January 27, 2021 18:42

  Hi Steve  You can toggle fullscreen view in the Viewer by pressing the F8 key on the keyboard instead of going to the Settings interface to do so.  Once the change is made, it persists and remains active when the Viewer is next launched until toggled back to a windowed screen.

  If you need to program your own fullscreen .cpp application though, I could not find an equivalent method for implementing a fullscreen mode with scripting.  There may be scope for modifying the default fullscreen status of the Viewer program via the model-views.cpp file.

  https://github.com/IntelRealSense/librealsense/blob/master/common/model-views.cpp#L2720

  0

  Comment actions Permalink
• Steve

  • February 08, 2021 19:53

  hi - thanks for the reply - - actually the fulscreen part is the least of my worries - 

  and when i say fulscreen - i mean without the controls/gui ... so just a window with the depth stream in it.

   

  i can manage this so far by going into the .cpp of a particular example and changing the basic window setting .. then i can go into the 'build' folder and 'make' again . and it re-builds the examples - with my new settings ..

   

  so thats great ..

   

  what i CANT manage to do is find a way of getting the output that i have from the realsense-viewer to show up as a standalone window.

  so im trying to modify one of the c++ examples (on the rasp pi 4) to work as i want

  let me explain

   

  the examples all have their own settings .. none of which i want - i want particular settings that i set in the realsense viewer ..

   

  i can, in theory, save these as a .json file ... which is great ... but cant do anything with that file as far as using it to control the stream settings in an existing example ... 

   

  i managed to guess that their were numbered colourizer presets so i could change the colour sets ... by changing numbers in the examples and re-'making' them .. .but there must be a block of code i can use to just open a window up of a set size and have it display the stream based on the /json file in question ? .. 

  0

  Comment actions Permalink
• MartyX Grover

  • February 09, 2021 15:02
  • Edited

  How I interpret your description above is that you would like to create a C++ program that is a standalone window that has no selectable options like the Viewer does (something more like rs-capture) but you would like to apply your own default settings taken from the Viewer to your program by using a json file exported from the Viewer.

  If the above description is correct then it should be relatively straightforward to do so.  You can load a custom json file into your own application to apply the settings in the file to it.  Links to some examples of json file loading scripting for C++ can be found in the link below.

  https://github.com/IntelRealSense/librealsense/issues/7322

  You should be able to test this scripting by modifying an existing C++ example program.

  0

  Comment actions Permalink
• Steve

  • February 11, 2021 23:42

  hi - -thanks so much for getting back to me ..

   im kinda melting down 

  ive been plugging away with realsense cam 435 on raspberry pi 4 .. and having followed MANY paths ive ALMOST achieved what i need to do ... but not quite ...
  i have limited programming experience - - and started off using c# to do what i needed as it was a case of taking the examples and modifying them .. it took me a while to work out how to 'build' them again .. and i just got lucky if im honest .

  i then hit a wall and needed to follow examples which were in python ...

  but i couldnt get the python examples to run .. which was a shame as i figured they would be easier to modify.

  eventually i cracked it - realizing that the build process had put certain files into the wrong folder - so a copy and paste later and hey presto.

  but i feel my luck is running out .. there's only so much searching the internet for bits of code and then blindly hacking them into the programs i can do.

  i dont fully understand the structure -- and i am learning .. but not fast enough 

  ideally i just want a display from the camera in a set colour pallet within specified min and max depth ..

  ive tried all i can to import and use a json file (created in realsense-viewer) but i just cant seem to leverage it .. not knowing exactly where to shoehorn the bits of code i find

  so i then tried to see if there was maybe a way of just calling the functions i want ..

  i decided id just put up with the set pallet presets (it would have been nice to apply a filter of my own but i have no idea how - id also like to apply the prost processing hole flter .. but that too is eluding me) and so i found a thing called clamping .. and then that led me to the realsense docs whch lists code snippets .. but because they are out of context i never seem to manage to get them in the right place or whatnot.

  would you mind looking at my code (its a short program) and telling me what on earth im doing wrong ? .... totally understand if you dont accept .. but it would make a desperate and tired sole very happy ... i am in your hands - kind regards

   

  the following does MOST of what i want .. it just starts up a window - streams the depth image and then resizes it to fill the window .. cool

   

  so then i can only colourise it with presets - - ok .. not bad .. but ideally i want the colours the other way round so when something is close its red and when its far away its blue .. with the colours graduated between that min and max (kind of the reverse of JET

   

  i need to be able to set the min and max distance for the depths .. i cant seem to do anything with the json file as there are too many lines of code in each example i see which means i cant just get a simple command and insert it somewhere obvious .. there's way more to it than that ... so i tracked down 

  __init__(self: pyrealsense2.threshold_filter, min_dist: float=0.15000000596046448, max_dist: float=4.0) → None

   

  but i have no idea where to put that .. and when i try it throws up all manner of errors .. as there are clearly syntax issues in it but i am a beginner and i cant work out which bits to put where - for example - - im pretty sure that arrow symbol before 'none' is wrong .. but why is included in the page:

  https://intelrealsense.github.io/librealsense/python_docs/_generated/pyrealsense2.threshold_filter.html?highlight=threshold

  and why on earth cant it be shown in ANY example anywhere that i can find .. 

   

  really hope you can help .

  losing my mind

   

  steve

   

   

   

  ## License: Apache 2.0. See LICENSE file in root directory.
  ## Copyright(c) 2015-2017 Intel Corporation. All Rights Reserved.

  ###############################################
  ## Open CV and Numpy integration ##
  ###############################################

  import pyrealsense2 as rs
  import numpy as np
  import cv2
  import datetime

  import json
  data = json.load(open('Custom1.json', 'r'))

   

  
  # Configure depth and color streams
  pipeline = rs.pipeline()
  config = rs.config()
  ###config.enable_stream(rs.stream.depth, 848, 480, rs.format.z16, 60)
  ###config.enable_stream(rs.stream.color, 640, 480, rs.format.bgr8, 30)

  # Start streaming
  cfg = pipeline.start(config)
  dev = cfg.get_device()
  depth_sensor = dev.first_depth_sensor()
  depth_sensor.set_option(rs.option.visual_preset, 4)

  
  iteration = 0;
  preset = 1;
  preset_name = '';

  try:
  while True:

  # Wait for a coherent pair of frames: depth and color
  frames = pipeline.wait_for_frames()
  depth_frame = frames.get_depth_frame()
  color_frame = frames.get_color_frame()
  
  iteration = iteration + 0
  if iteration > 100:
  preset = preset + 0
  iteration = 0
  range = depth_sensor.get_option_range(rs.option.visual_preset)
  preset = preset % range.max
  depth_sensor.set_option(rs.option.visual_preset, preset)
  preset_name = depth_sensor.get_option_value_description(rs.option.visual_preset, preset)
  
  # Convert images to numpy arrays
  depth_image = np.asanyarray(depth_frame.get_data())
  color_image = np.asanyarray(color_frame.get_data())

  # Apply colormap on depth image (image must be converted to 8-bit per pixel first)
  depth_colormap = cv2.applyColorMap(cv2.convertScaleAbs(depth_image, None, 0.5, 0), cv2.COLORMAP_RAINBOW)
  # Stack both images horizontally - ive changed this to just the one image - - at some point id like to merge the colour image with it .. but not now

  
  images = depth_colormap
  ###images = depth_image
  
  #font = cv2.FONT_HERSHEY_SIMPLEX
  
  
  

  # Show images
  cv2.namedWindow('RealSense', cv2.WINDOW_NORMAL)
  cv2.resizeWindow('RealSense', 1920,1080)
  cv2.imshow('RealSense', images)
  
  
  key = cv2.waitKey(1)
  # Press esc or 'q' to close the image window
  if key & 0xFF == ord('q') or key == 27:
  cv2.destroyAllWindows()
  break

  finally:

  # Stop streaming
  pipeline.stop()

   

   

   

   

   

  0

  Comment actions Permalink
• Steve

  • February 11, 2021 23:55

  as an aside - her's a classic example of what i keep finding - - for a start its old .. but also the last comments are ALMOST helpful .. but then leave me with sooo many questions ... 

  this modified example has the potential to use a json file if the last comments are to be believed - - but its indicative at best ... :S

   

   

  ## License: Apache 2.0. See LICENSE file in root directory.
  ## Copyright(c) 2015-2017 Intel Corporation. All Rights Reserved.
  
  ###############################################
  ##      Open CV and Numpy integration        ##
  ###############################################
  
  import pyrealsense2 as rs
  import numpy as np
  import cv2
  
  # Configure depth and color streams
  pipeline = rs.pipeline()
  config = rs.config()
  config.enable_stream(rs.stream.depth, 640, 480, rs.format.z16, 30)
  config.enable_stream(rs.stream.color, 640, 480, rs.format.bgr8, 30)
  
  # Start streaming
  cfg = pipeline.start(config)
  dev = cfg.get_device()
  depth_sensor = dev.first_depth_sensor()
  depth_sensor.set_option(rs.option.visual_preset, 4)
  
  iteration = 0;
  preset = 0;
  preset_name = '';
  
  try:
      while True:
  
          # Wait for a coherent pair of frames: depth and color
          frames = pipeline.wait_for_frames()
          depth_frame = frames.get_depth_frame()
          color_frame = frames.get_color_frame()
          
          iteration = iteration + 1
          if iteration > 100:
              preset = preset + 1
              iteration = 0
              range = depth_sensor.get_option_range(rs.option.visual_preset)
              preset = preset % range.max
              depth_sensor.set_option(rs.option.visual_preset, preset)
              preset_name = depth_sensor.get_option_value_description(rs.option.visual_preset, preset)
          
          # Convert images to numpy arrays
          depth_image = np.asanyarray(depth_frame.get_data())
          color_image = np.asanyarray(color_frame.get_data())
  
          # Apply colormap on depth image (image must be converted to 8-bit per pixel first)
          depth_colormap = cv2.applyColorMap(cv2.convertScaleAbs(depth_image, None, 0.5, 0), cv2.COLORMAP_JET)
  
          # Stack both images horizontally
          images = np.hstack((color_image, depth_colormap))
          
          font = cv2.FONT_HERSHEY_SIMPLEX
          cv2.putText(images, preset_name,(60,80), font, 4,(255,255,255),2,cv2.LINE_AA)
  
          # Show images
          cv2.namedWindow('RealSense', cv2.WINDOW_AUTOSIZE)
          cv2.imshow('RealSense', images)
          cv2.waitKey(1)
  
  finally:
  
      # Stop streaming
      pipeline.stop()

  In addition, load_json should also work. After reviewing your code, I had two ideas:

  1. Instead of using find_device_that_supports_advanced_mode, get the device from pipeline configuration after start:

  cfg = pipeline.start(config)
  dev = cfg.get_device()

  2. Perform load_json after you start streaming

   

   

   

   

  0

  Comment actions Permalink
• MartyX Grover

  • February 12, 2021 09:09

  I am not aware of a means to grade depth from red to blue instead of blue to red, so if you are trying to reverse the color scheme with code and it is not working then it is not necessarily a problem with your code.  Depth can though be represented in monochrome as dark in the short range and light in the far range using the Black to White color scheme.  

  A RealSense user created a C# implementation of the Threshold Filter in the link below that you can use as a reference.

  https://support.intelrealsense.com/hc/en-us/community/posts/360039445773-Json-configuration-change-the-min-max-distance-depth-to-get-an-image-with-the-right-colors-D415-camera-C-?page=1#community_comment_360010898533

  0

  Comment actions Permalink
• Steve

  • February 12, 2021 09:38

  thanks for getting back

   

  i am not so focused on the color issue - i realise thats a lot to ask.

   

  but the depth parameters ARE my focus and the example you provide is C#  ... i am unable to translate this into any meaningful python 

   

  which is what i am now working with.

   

  any other ideas ?

  steve

  0

  Comment actions Permalink
• MartyX Grover

  • February 12, 2021 10:46

  If you are now using Python, some recently contributed example code for setting colorization options is here:

  https://github.com/IntelRealSense/librealsense/issues/8299#issuecomment-773833234

  In regard to setting the threshold filter with Python, there isn't a standalone script to refer to that I know of like there is for C++ and C#.  The best references that I am aware of are in the links below.

  https://intelrealsense.github.io/librealsense/python_docs/_generated/pyrealsense2.threshold_filter.html?highlight=threshold

  https://github.com/IntelRealSense/librealsense/commit/84f78c9888a4e2ae48100e52df2e91c5ab82e683

  0

  Comment actions Permalink

Please sign in to leave a comment.