Différences

Ci-dessous, les différences entre deux révisions de la page.

--- intel_realsense [2021/03/18 14:35] – serge
+++ intel_realsense [2022/03/22 14:12] (Version actuelle) – [Les pages sur la détection de squelette] Benjamin Labomedia
@@ Ligne 1: / Ligne 1: @@
 ====== Intel Realsense======
+<WRAP center round box 80% centeralign>
+**{{tagpage>skeleton camera3D|Les pages sur les caméras 3D et la détection de squelette}}**     **[[http://translate.google.com/translate?hl=&sl=auto&tl=en&u=https%3A%2F%2Fressources.labomedia.org%2Fintel_realsense|English Version]]**
+</WRAP>
-=====Installation du driver du Capteur D455=====
+{{ :media_15:img_0005.jpg?600 |}}
+===== Ressources =====
+  * **[[cameras_de_profondeur|Camera Realsense Installation]] Intel® RealSense™ Computer Vision - Depth and Tracking cameras**
+  * [[https://intelrealsense.github.io/librealsense/python_docs/_generated/pyrealsense2.html|API python]]
+  * [[https://pyrealsense.readthedocs.io/en/master/pyrealsense.html|pyrealsense.readthedocs.io]]
+  * [[https://github.com/IntelRealSense/librealsense/tree/development/wrappers/python/examples| Exemples python sur GitHub]]
+=====Installation du driver du Capteur D455=====
   * **[[https://github.com/IntelRealSense/librealsense/blob/master/doc/distribution_linux.md|github.com librealsense doc linux]]**
-  sudo apt-key adv --keyserver keys.gnupg.net --recv-key F6E65AC044F831AC80A06380C8B3A55A6F3EFCDE || sudo apt-key adv --keyserver hkp://keyserver.ubuntu.com:80 --recv-key F6E65AC044F831AC80A06380C8B3A55A6F3EFCDE
+Validé sur Debian 11 Bullseye
-  sudo add-apt-repository "deb http://realsense-hw-public.s3.amazonaws.com/Debian/apt-repo bionic main" -u
-  sudo apt-get install librealsense2-dkms
+<code bash>
-  sudo apt-get install librealsense2-utils
+sudo apt-key adv --keyserver keys.gnupg.net --recv-key F6E65AC044F831AC80A06380C8B3A55A6F3EFCDE || sudo apt-key adv --keyserver hkp://keyserver.ubuntu.com:80 --recv-key F6E65AC044F831AC80A06380C8B3A55A6F3EFCDE
+sudo apt install software-properties-common
+sudo add-apt-repository "deb https://librealsense.intel.com/Debian/apt-repo focal main" -u
+</code>
+Puis\\
+  sudo apt update
+  sudo apt install -f librealsense2-dkms
 Tester avec:
+  sudo apt install librealsense2-utils
   realsense-viewer
+Voir si le module est bien chargé
+  modinfo uvcvideo | grep "version:"            devrait inclure realsense
 {{ ::realsense_viewer.png?400 |}}
@@ Ligne 20: / Ligne 41: @@
   * https://github.com/IntelRealSense/librealsense/tree/master/wrappers/python#installation
-  import pyrealsense2 as rs
+  python3 -m pip install pyrealsense2 --user
+====L'exemple de la doc librealsense====
-===L'exemple de la doc librealsense===
   * [[https://github.com/IntelRealSense/librealsense/blob/master/wrappers/python/examples/opencv_viewer_example.py|opencv_viewer_example.py]]
-<code python opencv_viewer_example.py>
+{{ :media_14:realsense_opencv.png?800 |}}
-## 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()
-# Get device product line for setting a supporting resolution
-pipeline_wrapper = rs.pipeline_wrapper(pipeline)
-pipeline_profile = config.resolve(pipeline_wrapper)
-device = pipeline_profile.get_device()
-device_product_line = str(device.get_info(rs.camera_info.product_line))
-config.enable_stream(rs.stream.depth, 640, 480, rs.format.z16, 30)
-if device_product_line == 'L500':
+====Calibration de la caméra====
-    config.enable_stream(rs.stream.color, 960, 540, rs.format.bgr8, 30)
+**Avec une calibration, la détction du squelette est beaucoup plus stable.**\\
-else:
+Une copie du script de la doc se trouve dans le dossier rs-imu-calibration: [[https://github.com/sergeLabo/rs-opencv/blob/main/rs-imu-calibration/rs-imu-calibration.py|rs-imu-calibration.py]]\\
-    config.enable_stream(rs.stream.color, 640, 480, rs.format.bgr8, 30)
+Suivre le script et le PDF [[https://github.com/sergeLabo/rs-opencv/blob/main/rs-imu-calibration/IMU_Calibration_Tool_for_Intel_RealSense-Depth_Cameras_Whitepaper.pdf|IMU_Calibration_Tool_for_Intel_RealSense-Depth_Cameras_Whitepaper.pdf]]
-# Start streaming
+Exemple d'une calibration
-pipeline.start(config)
+<code bash>
+python3 rs-imu-calibration.py
+waiting for realsense device...
+  Device PID:  0B5C
+  Device name:  Intel RealSense D455
+  Serial number:  053422251581
+  Product Line:  D400
+  Firmware version:  05.12.11.00
+Start interactive mode:
+FOUND ACCEL with fps=63
+FOUND GYRO with fps=200
+-------------------------
+*** Press ESC to Quit ***
+-------------------------
-try:
+Align to direction:  [ 0. -1.  0.]   Mounting screw pointing down, device facing out
-    while True:
+ Status.collect_data[...................]]
-        # Wait for a coherent pair of frames: depth and color
+Direction data collected.
-        frames = pipeline.wait_for_frames()
+Align to direction:  [1. 0. 0.]   Mounting screw pointing left, device facing out
-        depth_frame = frames.get_depth_frame()
+ Status.collect_dataWARNING: MOVING     ]]
-        color_frame = frames.get_color_frame()
+ Status.collect_data[...................]]
-        if not depth_frame or not color_frame:
-            continue
-        # Convert images to numpy arrays
+Direction data collected.
-        depth_image = np.asanyarray(depth_frame.get_data())
+Align to direction:  [0. 1. 0.]   Mounting screw pointing up, device facing out
-        color_image = np.asanyarray(color_frame.get_data())
+ Status.collect_data[...................]]
-        # Apply colormap on depth image (image must be converted to 8-bit per pixel first)
+Direction data collected.
-        depth_colormap = cv2.applyColorMap(cv2.convertScaleAbs(depth_image, alpha=0.03), cv2.COLORMAP_JET)
+Align to direction:  [-1.  0.  0.]   Mounting screw pointing right, device facing out
+ Status.collect_dataWARNING: MOVING     ]]
+ Status.collect_data[...................]]
-        depth_colormap_dim = depth_colormap.shape
+Direction data collected.
-        color_colormap_dim = color_image.shape
+Align to direction:  [ 0.  0. -1.]   Viewing direction facing down
+ Status.collect_data[...................]]
-        # If depth and color resolutions are different, resize color image to match depth image for display
+Direction data collected.
-        if depth_colormap_dim != color_colormap_dim:
+Align to direction:  [0. 0. 1.]   Viewing direction facing up
-            resized_color_image = cv2.resize(color_image, dsize=(depth_colormap_dim[1], depth_colormap_dim[0]), interpolation=cv2.INTER_AREA)
+ Status.collect_dataWARNING: MOVING     ]]
-            images = np.hstack((resized_color_image, depth_colormap))
+ Status.collect_data[...................]]
-        else:
-            images = np.hstack((color_image, depth_colormap))
-        # Show images
+Direction data collected.
-        cv2.namedWindow('RealSense', cv2.WINDOW_AUTOSIZE)
+Would you like to save the raw data? Enter footer for saving files (accel_<footer>.txt and gyro_<footer>.txt)
-        cv2.imshow('RealSense', images)
+Enter nothing to not save raw data to disk. >toto
-        cv2.waitKey(1)
-finally:
-    # Stop streaming
+Writing files:
-    pipeline.stop()
+accel_toto.txt
+gyro_toto.txt
+[ 5.22060069e-04  6.86921142e-05 -2.07340278e-03]
+[1000 1000 1000 1000 1000 1000]
+using 6000 measurements.
+[[ 1.01076975e+00  3.60727590e-02 -8.35014969e-03]
+ [ 4.92538947e-04  1.01871131e+00 -8.72837958e-03]
+ [-6.46373829e-03  1.24138802e-02  1.01959596e+00]
+ [ 2.24575082e-01  1.56413994e-01  2.86333600e-01]]
+residuals: [  7.25572588   3.70199967 114.63373591]
+rank: 4
+singular: [441.23030094 430.20638248 424.20085017  77.26445512]
+norm (raw data  ): 9.655707
+norm (fixed data): 9.805557 A good calibration will be near 9.806650
+Would you like to write the results to the camera? (Y/N)Y
+Writing calibration to device.
+  Device PID:  0B5C
+  Device name:  Intel RealSense D455
+  Serial number:  053422251581
+  Firmware version:  05.12.11.00
+SUCCESS: saved calibration to camera.
+Done.
 </code>
 =====Les pages sur la détection de squelette=====
   * [[skeleton_tracking_de_cubemos_logiciel_proprietaire|]]
-  * [[detection_d_un_squelette_dans_une_image|]]
+  * [[pose_estimation_avec_opencv]]
-{{tag> capteurs sb }}
+{{tag> capteurs opencv python realsense sb skeleton camera3D ia}}