Pixel2ray function with pinhole camera

[nithin.m1]

Please provide the following info (check/uncheck the boxes after creating this topic):
Software Version
[+] DRIVE OS Linux 5.2.6
DRIVE OS Linux 5.2.0
DRIVE OS Linux 5.2.0 and DriveWorks 3.5
NVIDIA DRIVE™ Software 10.0 (Linux)
NVIDIA DRIVE™ Software 9.0 (Linux)
other DRIVE OS version
other

Target Operating System
[+] Linux
QNX
other

Hardware Platform
[+] NVIDIA DRIVE™ AGX Xavier DevKit (E3550)
NVIDIA DRIVE™ AGX Pegasus DevKit (E3550)
other

SDK Manager Version
[+] 1.6.1.8175
1.6.0.8170
other

Host Machine Version
[+] native Ubuntu 18.04
other

Hi team,

I was trying to use my calibrated pinhole camera to convert the image points to world points as shown in the following link,

How to transform points from image space to world space?

Hello, I have a quick question about rendering the points obtained from LaneGraph. We have followed the workflow described in the documentation and are using LaneNet as source with H264 serilized data from one of our drives using Sekonix AR-231-RCCB. The LeaneNet results are correct and we use the lane detection results as input to dwLaneGraph_setLanesMeasurement followed by dwLaneGraph_processHost. The documentation states: “The geometry includes the center paths of lanes or as the dual, thei…

Using the function “dwCameraModel_pixel2Ray” I was able to do the transformation correctly on NVIDIA provided sample videos and generate correct ground/world points. NVIDIA sample was using ‘ocam’ model. (I had tried it on lane detection points).

But when I tried the same for my calibrated pinhole camera it was not giving correct results.
could you please help on this?
Another query is, Whether the “dwCameraModel_pixel2Ray” function work with pinhole camera model also?

[VickNV]

@nithin.m1

[+] DRIVE OS Linux 5.2.6

Hi @nithin.m1,

Are you using DRIVE OS 5.2.6?

[nithin.m1]

I m using drive software 10 ( linux ) only sorry for the error

[VickNV]

I’m thinking maybe some impact from the issue mentioned in the topic.

[calibration-graph-cli] Pinhole calibration bug

When calibrating a camera with a pinhole model, the calibration-graph-cli estimates the distortion with 5 parameters “distortion”: [-0.3046111298320273,0.19613377571040838,0,0,-0.14857597315214646], When exporting the graph to rig, only the first three are exported to the rig: “distortion”: "-3.04611116647720e-1 1.96133777499199e-1 0.000000000000000 ", This seems like a bug, especially because the value of the last parameter is significant (-0.1485759731521464) Questions: Which of the abo…

[nithin.m1]

Hi @VickNV,
Okay I think I need to elaborate it further,
We are yet to receive the drive agx platform, So what I was trying to do is, to use one of the usb camera which we have calibrated externally using MATLAB (not using drive calibration tools) with the lane detection sample and calculate the world points of generated lane points.

lane Detection sample was working correctly on nvidia provided videos and able to generate lane points and corresponding ground points correctly. But when I modified the rig file using our usb camera parameters (both intrinsic and extrinsic) and tried the same with our recorded video, the ground points generated for lane points are not correct.

I recorded the lane points generated by lane detection sample with our video and tried in MATLAB with the same calibration values , we are getting the ground point values correctly (using MATLAB ‘imageToVehicle’ function).

am I doing something wrong here? Could you please guide?

[VickNV]

Please share the files you used to prove MATLAB gets correct ground point values and the file (modified sample_lane_detection source, rig file, video, any other?) you found our DriveWorks gets wrong ground values for us to check with our team. Thanks.