* Added stereo calibration using charuco board (backport #976 <https://github.com/ros-perception/image_pipeline/issues/976>) (#1002 <https://github.com/ros-perception/image_pipeline/issues/1002>)
From #972 <https://github.com/ros-perception/image_pipeline/issues/972>
Doing this first for rolling.
This was a TODO in the repository, opening this PR to add this feature.
- The main issue why this wasn't possible imo is the way mk_obj_points
works. I'm using the inbuilt opencv function to get the points there.
- The other is a condition when aruco markers are detected they are
added as good points, This is fine in case of mono but in stereo these
have to be the same number as the object points to find matches although
this should be possible with aruco.<hr>This is an automatic backport of
pull request #976 <https://github.com/ros-perception/image_pipeline/issues/976> done by [Mergify](https://mergify.com).
Co-authored-by: Myron Rodrigues <mailto:41271144+MRo47@users.noreply.github.com>
* Change camera info message to lower case (backport #1005 <https://github.com/ros-perception/image_pipeline/issues/1005>) (#1007 <https://github.com/ros-perception/image_pipeline/issues/1007>)
Change camera info message to lower case since message type had been
change in rolling and humble.
[](https://github.com/ros2/common_interfaces/blob/rolling/sensor_msgs/msg/CameraInfo.msg)<hr>This
is an automatic backport of pull request #1005 <https://github.com/ros-perception/image_pipeline/issues/1005> done by
[Mergify](https://mergify.com).
---------
Co-authored-by: SFhmichael <mailto:146928033+SFhmichael@users.noreply.github.com>
Co-authored-by: Alejandro Hernández Cordero <mailto:ahcorde@gmail.com>
* Contributors: mergify[bot]
depth_image_proc
No changes
image_pipeline
No changes
image_proc
No changes
image_publisher
* [jazzy] image_publisher: Fix loading of the camera info parameters on startup (backport #983 <https://github.com/ros-perception/image_pipeline/issues/983>) (#995 <https://github.com/ros-perception/image_pipeline/issues/995>)
As described in
https://github.com/ros-perception/image_pipeline/issues/965 camera info
is not loaded from the file on node initialization, but only when the
parameter is reloaded.
This PR resolves this issue and should be straightforward to port it to
Humble, Iron and Jazzy.<hr>This is an automatic backport of pull
request #983 <https://github.com/ros-perception/image_pipeline/issues/983> done by [Mergify](https://mergify.com).
Co-authored-by: Krzysztof Wojciechowski <mailto:49921081+Kotochleb@users.noreply.github.com>
* image_publisher: Fix image, constantly flipping when static image is published (backport #986 <https://github.com/ros-perception/image_pipeline/issues/986>) (#987 <https://github.com/ros-perception/image_pipeline/issues/987>)
Continuation of
https://github.com/ros-perception/image_pipeline/pull/984.
When publishing video stream from a camera, the image was flipped
correctly. Yet for a static image, which was loaded once, the flip
function was applied every time ImagePublisher::doWork() was called,
resulting in the published image being flipped back and forth all the
time.
This PR should be straightforward to port it to Humble, Iron and
Jazzy.<hr>This is an automatic backport of pull request #986 <https://github.com/ros-perception/image_pipeline/issues/986> done by
[Mergify](https://mergify.com).
Co-authored-by: Krzysztof Wojciechowski <mailto:49921081+Kotochleb@users.noreply.github.com>
Co-authored-by: Alejandro Hernández Cordero <mailto:ahcorde@gmail.com>
* [jazzy] image_publisher: add field of view parameter (backport #985 <https://github.com/ros-perception/image_pipeline/issues/985>) (#992 <https://github.com/ros-perception/image_pipeline/issues/992>)
Currently, the default value for focal length when no camera info is
provided defaults to 1.0 rendering whole approximate intrinsics and
projection matrices useless. Based on [this
article](https://learnopencv.com/approximate-focal-length-for-webcams-and-cell-phone-cameras/),
I propose a better approximation of the focal length based on the field
of view of the camera.
For most of the use cases, users will either know the field of view of
the camera the used, or they already calibrated it ahead of time.
If there is some documentation to fill. please let me know.
This PR should be straightforward to port it to Humble, Iron and
Jazzy.
<hr>This is an automatic backport of pull request #985 <https://github.com/ros-perception/image_pipeline/issues/985> done by
[Mergify](https://mergify.com).
Co-authored-by: Krzysztof Wojciechowski <mailto:49921081+Kotochleb@users.noreply.github.com>
* Contributors: mergify[bot]
Increasing version of package(s) in repository
image_pipeline
to5.0.3-1
:jazzy/distribution.yaml
0.12.0
5.0.1-3
camera_calibration
depth_image_proc
image_pipeline
image_proc
image_publisher
image_rotate
image_view
stereo_image_proc
tracetools_image_pipeline