mirror of
https://github.com/opencv/opencv.git
synced 2026-07-29 15:23:05 +04:00
Converted multiview calibration sample to application.
This commit is contained in:
Alexander Smorkalov
@@ -0,0 +1 @@
|
||||
matplotlib
|
||||
+2
-1
@@ -158,6 +158,7 @@ if(DOXYGEN_FOUND)
|
||||
set(CMAKE_DOXYGEN_TUTORIAL_JS_ROOT "- @ref tutorial_js_root")
|
||||
set(tutorial_js_path "${CMAKE_CURRENT_SOURCE_DIR}/js_tutorials")
|
||||
set(example_path "${CMAKE_SOURCE_DIR}/samples")
|
||||
set(apps_path "${CMAKE_SOURCE_DIR}/apps")
|
||||
|
||||
set(doxygen_image_path
|
||||
${CMAKE_CURRENT_SOURCE_DIR}/images
|
||||
@@ -179,7 +180,7 @@ if(DOXYGEN_FOUND)
|
||||
string(REPLACE ";" " \\\n" CMAKE_DOXYGEN_EXCLUDE_LIST "${CMAKE_DOXYGEN_EXCLUDE_LIST}")
|
||||
string(REPLACE ";" " " CMAKE_DOXYGEN_ENABLED_SECTIONS "${CMAKE_DOXYGEN_ENABLED_SECTIONS}")
|
||||
# TODO: remove paths_doc from EXAMPLE_PATH after face module tutorials/samples moved to separate folders
|
||||
string(REPLACE ";" " \\\n" CMAKE_DOXYGEN_EXAMPLE_PATH "${example_path} ; ${paths_doc} ; ${paths_sample}")
|
||||
string(REPLACE ";" " \\\n" CMAKE_DOXYGEN_EXAMPLE_PATH "${example_path} ; ${apps_path} ; ${paths_doc} ; ${paths_sample}")
|
||||
string(REPLACE ";" " \\\n" CMAKE_DOXYGEN_INCLUDE_ROOTS "${paths_include}")
|
||||
set(CMAKE_DOXYGEN_LAYOUT "${CMAKE_CURRENT_BINARY_DIR}/DoxygenLayout.xml")
|
||||
set(CMAKE_DOXYGEN_OUTPUT_PATH "doxygen")
|
||||
|
||||
@@ -1,4 +1,4 @@
|
||||
Multi-view Camera Calibration Tutorial {#tutorial_multiview_camera_calibration}
|
||||
Multi-view Camera Calibration Tutorial {#tutorial_multiview_camera_calibration}
|
||||
==========================
|
||||
|
||||
@tableofcontents
|
||||
@@ -71,7 +71,7 @@ Assume we have `N` camera views, for each `i`-th view there are `M` images conta
|
||||
|
||||
Python example
|
||||
--
|
||||
There are two options to run the sample code in Python (`opencv/samples/python/multiview_calibration.py`) either with raw images or provided points.
|
||||
There are two options to run the sample code in Python (`opencv/apps/multiview-calibration/multiview_calibration.py`) either with raw images or provided points.
|
||||
The first option is to prepare `N` files where each file has the path to an image per line (images of a specific camera of the corresponding file). Leave the line empty, if there is no corresponding image for the camera in a certain frame. For example, a file for camera `i` should look like (`file_i.txt`):
|
||||
```
|
||||
/path/to/image_1_of_camera_i
|
||||
@@ -231,25 +231,25 @@ def mutiviewCalibration (pattern_points, image_points, detection_mask):
|
||||
Python sample API
|
||||
----
|
||||
|
||||
To run the calibration procedure in Python follow the following steps (see sample code in `samples/python/multiview_calibration.py`):
|
||||
To run the calibration procedure in Python follow the following steps (see sample code in `apps/multiview-calibration/multiview_calibration.py`):
|
||||
|
||||
-# **Prepare data**:
|
||||
|
||||
@snippet samples/python/multiview_calibration.py calib_init
|
||||
@snippet apps/multiview-calibration/multiview_calibration.py calib_init
|
||||
|
||||
The detection mask matrix is later built by checking the size of image points after detection:
|
||||
|
||||
-# **Detect pattern points on images**:
|
||||
|
||||
@snippet samples/python/multiview_calibration.py detect_pattern
|
||||
@snippet apps/multiview-calibration/multiview_calibration.py detect_pattern
|
||||
|
||||
-# **Build detection mask matrix**:
|
||||
|
||||
@snippet samples/python/multiview_calibration.py detection_matrix
|
||||
@snippet apps/multiview-calibration/multiview_calibration.py detection_matrix
|
||||
|
||||
-# **Finally, the calibration function is run as follows**:
|
||||
|
||||
@snippet samples/python/multiview_calibration.py multiview_calib
|
||||
@snippet apps/multiview-calibration/multiview_calibration.py multiview_calib
|
||||
|
||||
|
||||
C++ sample API
|
||||
@@ -289,24 +289,24 @@ Practical Debugging Techniques
|
||||
|
||||
-# Camera intrinsics can be better estimated when points are more scattered in the image. The following code can be used to plot out the heat map of the observed point
|
||||
|
||||
@snippet samples/python/multiview_calibration.py plot_detection
|
||||
@snippet apps/multiview-calibration/multiview_calibration.py plot_detection
|
||||
|
||||
The left example is not well scattered while the right example shows a better-scattered pattern
|
||||
|
||||
-# Plot out the reprojection error to ensure the result is reasonable
|
||||
-# If ground truth camera intrinsics are available, a visualization of the estimated error on intrinsics is provided.
|
||||
|
||||
@snippet samples/python/multiview_calibration.py vis_intrinsics_error
|
||||
@snippet apps/multiview-calibration/multiview_calibration.py vis_intrinsics_error
|
||||
|
||||
resulting visualization would look similar to
|
||||
|
||||
|
||||
-# **Multiview calibration**
|
||||
-# Use `plotCamerasPosition` in samples/python/multiview_calibration.py to plot out the graph established for multiview calibration. shows positions of cameras, checkerboard (of a random frame), and pairs of cameras connected by black lines explicitly demonstrating tuples that were used in the initial stage of stereo calibration.
|
||||
-# Use `plotCamerasPosition` in apps/multiview-calibration/multiview_calibration.py to plot out the graph established for multiview calibration. shows positions of cameras, checkerboard (of a random frame), and pairs of cameras connected by black lines explicitly demonstrating tuples that were used in the initial stage of stereo calibration.
|
||||
The dashed gray lines demonstrate the non-spanning tree edges that are also used in the optimization.
|
||||
The width of these lines indicates the number of co-visible frames i.e. the strength of connection.
|
||||
It is more desired if the edges in the graph are dense and thick.  For the right tree, the connection for camera four is rather limited and can be strengthened
|
||||
|
||||
-# Visulization method for showing the reprojection error with arrows (from a given point to the back-projected one) is provided (see `plotProjection` in samples/python/multiview_calibration.py). The color of the arrows highlights the error values. Additionally, the title reports mean error on this frame and its accuracy among other frames used in calibration.
|
||||
-# Visulization method for showing the reprojection error with arrows (from a given point to the back-projected one) is provided (see `plotProjection` in apps/multiview-calibration/multiview_calibration.py). The color of the arrows highlights the error values. Additionally, the title reports mean error on this frame and its accuracy among other frames used in calibration.
|
||||
|
||||
|
||||
|
||||
Reference in New Issue
Block a user