Differences

This shows you the differences between two versions of the page.

--- tutorials:camera_arm_calibration [2019/11/08 21:42] – [Executing the calibration script] dgarcia
+++ tutorials:camera_arm_calibration [2022/09/20 00:08] (current) – external edit 127.0.0.1
@@ Line 53: / Line 53: @@
     - Enable auto white Balance
     - Auto Exposure Priority
+    - Set White Balance to: 3760
   * Close ''realsense-viewer''
   * Remember to source tour workspace before executing the following commands.
@@ Line 111: / Line 112: @@
 </code>
-  * THen you need to provide a configuration file for the calibration board, like so:
+  * Then you need to provide a configuration file for the calibration board, like so:
 <code yaml>
@@ Line 118: / Line 119: @@
   cols: 10
   spacing: 0.01861
+</code>
+  * Now we have to create a separate directory for images and poses.
+<code bash>
+mv <catkin_workspace>/src/oms-cylinder/camera2png/data <catkin_ws>/src/oms-cylinder/oms_launcher/cal/configs
+cd <catkin_ws>/src/oms-cylinder/oms_launcher/cal/configs/data
+mkdir images
+mkdir poses
+mv *.yml poses
+mv *.png images
+</code>
+  * Now we have to crate a YAML listing all the files that have to be taken into consideration by the calibration script, in the ''configs/data'' directory.
+<code yaml>
+# TODO: Fix this exxample
+- pose: "000000.yml",
+  image: "000000.png"
+- pose: "000001.yml",
+  image: "000001.png"
+  .
+  .
+  .
+</code>
+  * Launch the static calibration script.
+<code bash>
+roslaunch oms-cylinder/oms_launcher/launch/static_calibration.launch
+</code>
+  * This will start the calibration script. This will open a window that will show you the photos that were taken, you should see some color dots. If your data set is small, you will have to press ''ENTER'' for each image. If your data set is big, only one ''ENTER'' is required (we have no idea what is the threshold, but is seems to be around 2000 poses).
+  * You must wait when you press ''ENTER'' and the program doesn't respond, there are some images that take longer to analyze than others. When you reach the last image, the //iamge-window-interface// will gt stuck, Do Not press ''ENTER''. Wait until the algorithm converges.
+  * Once the algorithm converges you will see a lot of results like the one shown below. The calibration result is the first one.
+<code>
+Did converge?: 1
+Initial cost?: 47853.2 (pixels per dot)
+Final cost?: 0.655122 (pixels per dot)
+BASE TO CAMERA:
+ -0.208397  -0.664739   0.717421 -0.0773392
+ -0.967382  0.0320803  -0.251281   0.220998
+.144021  -0.746387  -0.649742    1.43606
+          0          0          1
+--- URDF Format Base to Camera ---
+xyz="-0.0773392 0.220998 1.43606"
+rpy="0.854511(48.9599 deg) -2.99707(-171.719 deg) 1.35862(77.8429 deg)"
+qxyzw="-0.593562 0.687426 -0.362827 0.208531"
+BASE TO CAMERA:
+ -0.0184899    0.999697  -0.0162778  -0.0817589
+-0.00618554   0.0161659     0.99985    0.139854
+.99981   0.0185878  0.00588475 -0.00619646
+           0           0           1
+--- URDF Format Base to Camera ---
+xyz="-0.0817589 0.139854 -0.00619646"
+rpy="-1.87741(-107.567 deg) -1.59029(-91.1172 deg) 0.322833(18.497 deg)"
+qxyzw="-0.48976 -0.507142 -0.502048 0.500889"
 </code>
 ==== Verify that the calibration is correct ====