-
Notifications
You must be signed in to change notification settings - Fork 38
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Lidar points are not accurate towards scene representation. #153
Comments
@sriceumich @Autumn60 |
@sriceumich |
@Autumn60 I am using the DepthBufferLiDARSensor currently for this and the 32 beam configuration. The RaycastLiDARSensor is presenting it's own set of challenges. We noticed some additional issues in the sensor this past week as we prepared for a paper. As we started moving the ego, the sensor started exhibiting some broken details the further away the beams are. We attempted to use the points for our methods, but with the noise we were unable to produce good results from them. |
As you say, DepthBufferLiDAR has poor accuracy at long distances. Also, the lower the resolution of the RenderTexture used, the less accurate the point cloud. |
It is possible that the part that calculates the corresponding pixel index from the laser direction is wrong, but I don't know the exact cause yet now. |
Based on some experiments, I am convinced that the calculation to find the texture pixels from the laser direction is wrong. |
We are evaluating this as an alternative to our lidar simulator and seem to be having a few issues with the points generated. The lidar points are not accurately representing the scene and have a curve to them, points closer to the lidar have more of a curve. The noise doesn't seem to be representative of a lidar in the scene causing detection to seem inaccurate. Noise adjustment does not seem to compensate for this issue. I tried the various lidar that were represented and all seemed to exhibit this issue.
Curved floor with wide noise paths.
Extra thick points which should be a wall.
The text was updated successfully, but these errors were encountered: