wrong Pose returned by Camera::GetWorldRotation()

[osrf-migration]

Original report (archived issue) by Eric Chuang (Bitbucket: eric4337).

---

the math::Quaternion returned by Camera::GetWorldRotation() are always '0 0 0'.

no matter what my Camera pose is, it always return '0 0 0'

[osrf-migration]

Original comment by John Hsu (Bitbucket: hsu, GitHub: hsu).

---

I just tried by adding code below to gazebo_2.2 branch:

//////////////////////////////////////////////////
void Camera::Update()
{
  gzerr << this->GetName() << " : " << this->GetWorldRotation() << "\n";

and running

gazebo test/worlds/camera_rotation_test.world

got some non-zero values

Error [Camera.cc:318] default::multicamera_sensor_translated(2) : 0 0 0
Error [Camera.cc:318] default::left(3) : 0 0 0
Error [Camera.cc:318] default::right(4) : 0 0 0
Error [Camera.cc:318] default::left(5) : 1.2 1.3 1.4
Error [Camera.cc:318] default::right(6) : 1.2 1.3 1.4

[osrf-migration]

Original comment by Eric Chuang (Bitbucket: eric4337).

---

Sorry for my unspecific question.

Was your Camera attached to Model?

My Camera is created from SDF, under .
The Pose returned by Model, the parent of Camera, is correct.
But if i access the Pose of Camera directly, it seems to be wrong.

What i want is that i can get the "World Pose" of Camera directly, since the frame of Model & Camera are possibly not aligned.

[osrf-migration]

Original comment by John Hsu (Bitbucket: hsu, GitHub: hsu).

---

I added a unit test here:

https://osrf-migration.github.io/gazebo-gh-pages/#!/osrf/gazebo/branches/compare/issue_1087_gazebo_2.2..gazebo_2.2#diff

you can compile from branch issue_1087_gazebo_2.2 and run the test by:

[build dir]/test/integration/INTEGRATION_multicamera_sensor --gtest_filter="*WorldPose*"

The test fails, indicating there is an issue with Camera::GetWorldRotation() call.

Thanks for the bug report.

[osrf-migration]

Original comment by John Hsu (Bitbucket: hsu, GitHub: hsu).

---

Will issue #1098 make it easier for users to submit a patch similar to this test patch in the future?

[osrf-migration]

Original comment by Ian Chen (Bitbucket: Ian Chen, GitHub: iche033).

---

regarding the test:

line 376

#!c++

  EXPECT_EQ(model1->GetWorldPose() + multicamera1->GetPose(),
    multicamera1->GetCamera(0)->GetWorldPose());

• model1->GetWorldPose() = 0 10 1 0 0 0

• multicamera1->GetPose() = 0 0 0 0 0 0

since camera(0) has an offset pose of 0 1 0 0 0 0, so shouldn't the correct camera world pose be

• multicamera1->GetCamera(0)->GetWorldPose() = 0 11 1 0 0 0

then the EXPECT statement won't evaluate to true right?

[osrf-migration]

Original comment by Steve Peters (Bitbucket: Steven Peters, GitHub: scpeters).

---

Shouldn't the pose addition arguments be reversed as: multicamera1->GetPose() + model1->GetWorldPose()?

[osrf-migration]

Original comment by Ian Chen (Bitbucket: Ian Chen, GitHub: iche033).

---

I made some changes and the test now pass

Here is a diff showing my changes on top of yours
https://osrf-migration.github.io/gazebo-gh-pages/#!/osrf/gazebo/branches/compare/issue_1087_gazebo_2.2_ian..issue_1087_gazebo_2.2#dif

[osrf-migration]

Original comment by Ian Chen (Bitbucket: Ian Chen, GitHub: iche033).

---

@hsu I think the fix + test is ready, let's make a pull request for this?

[osrf-migration]

Original comment by Steve Peters (Bitbucket: Steven Peters, GitHub: scpeters).

---

• changed state from "new" to "resolved"

pull request #1071

[osrf-migration]

Original comment by Nate Koenig (Bitbucket: Nathan Koenig).

---

• changed state from "resolved" to "closed"