Calibration.bin not found

[shaikster101]

Issue Summary:

So i am trying to use the inbuild calibration toolbox, however I cannot seem to find the calibration.bin

Executed Command (if any)

Note: add --logging_level 0 --disable_multi_thread to get higher debug information.

build\examples\calibration.bin --help

Type of Issue

You might select multiple topics, delete the rest:

• Help wanted

Your System Configuration

1. General configuration:

   • Installation mode: CMake
   • Operating system (lsb_release -a in Ubuntu): Windows 10
   • Release or Debug mode? (by defualt: release): release
   • Compiler (gcc --version in Ubuntu or VS version in Windows): VS2015 Enterprise Update 3

2. Non-default settings:

   • 3-D Reconstruction module added? (by default: no): Yes
   • Any other custom CMake configuration with respect to the default version? (by default: no): No

3. 3rd-party software:

   • Caffe version: Default from OpenPose
   • CMake version (cmake --version in Ubuntu): 3.12.0
   • OpenCV version: OpenPose default

4. If GPU mode issue:

   • CUDA version (cat /usr/local/cuda/version.txt in most cases):8.0
   • cuDNN version: 5.1
   • GPU model (nvidia-smi in Ubuntu):Nvidia Titan X

5. If Windows system: Compiled Library

[gineshidalgo99]

For Windows it's a different path, it's in build\x64\Release\calibration.exe. Make sure you are able to run basic OpenPose before trying to run this utility (doc/quick_start.md)

[MarkU39]

I am having the same issue. I am able to run basic OpenPose, but the calibration.exe is neither in the location given in the documentation nor in this location. Any other suggestions for where to look for that file?

[W-LH]

Firstly You need to release the calibration part, like the steps you release the OpenPoseDemo. Then you will find it in the path.

[W-LH]

Besides, I also have a question of the calibration in Win10. I use the image the I took ahead of time to do the intrinsic calibration of my IP camera with the proper serial number, I got the right result and the xml file in "models\calibrationParameters\flir".
Then I run Openpose as usual, but there is totally not different between the images before and after the calibration. Therefore, I want to know that if I use the images that is token ahead of time to do the calibration, the Openpose will need more parameter to detect the IP camera I use, OR it can do it automatically just with the serial number? BTW, is there any problem with the camera_serial_number I got, mine is 17110902, which seems a right one.

[gineshidalgo99]

Camera un-distortion from OpenPose demo is only for point-grey cameras at the moment. Apologies if the doc is not clear about this.

[W-LH]

What a pity. Anyway thanks for your answer!

[W-LH]

@gineshidalgo99 So only the un-distortion is for point-grey camera. Can the intrinsic and extrinsic camera parameters calibration work well with IP camera?

[MarkU39]

If the un-distortion doesn't work, how much (if at all) does this hinder performance (e.g. accuracy, lag, etc.)? Is it just something I would have to write a routine for myself? I would like to assemble a working 3D setup to evaluate before I invest in a FLIR camera. Thanks for taking the time to answer our questions.

[gineshidalgo99]

Extrinsics are only required for 3-D setup, we only created 3-D set up for point-grey cameras as they include SDK for camera sync. If you wanna create your own 3D setup, you need your own code to make sure they are sync. That's why we have only provided it only for the point grey 3D camera setup