GmAPD Updates

[kdamkjer]

Draft enhancements to support GmAPD LiDAR. Addresses Issue #35.

[hobu]

After looking over these changes, my only suggestion is that the "aggregate" bit flag not only correspond to GmAPD systems, and it could also just as easily apply to dense imagery situations. I don't think we need to say anything about algorithms or consensus models on the description of the flag.

[rapidlasso]

I suggest that where ever the current version states "This is often the case for data generated by GmAPD systems." we replace this overly specific terminology with a much more general wording that can be used to any point cloud that is derived by combining multiple observations from some sensor (multiple images, photon counting arrays, averaged SAR scatterings, multiple linear full waveform LiDAR measurements accumulated in a voxel grid, sonar echo composited, etc) in contrast to the direct range observation by an active sensing system that LAS was originally designed for.

[rapidlasso]

I suggest that where ever the current version states "This is often the case for data generated by GmAPD systems." we replace this overly specific terminology with a much more general wording that can be used to any point cloud that is derived by combining multiple observations from some sensor (multiple images, photon counting arrays, averaged SAR scatterings, multiple linear full waveform LiDAR measurements accumulated in a voxel grid, sonar echo composited, etc) in contrast to the direct range observation by an active sensing system that LAS was originally designed for.

[anayegandhi]

Agree with comments from hobu and rapidlasso. Evon - can you please make these changes and then we seek final approval. Thanks.

[kdamkjer]

I made the changes requested w.r.t. GmAPD LiDAR systems language. I agree that the aggregate bit should apply to more than just GmAPD systems (which was hopefully reflected in the original language), but I think it is also important to acknowledge that it is being added so that GmAPD and other aggregate point clouds can be accepted as valid formatted LAS data. With the introduction of FWF, SPL, and GmAPD, perhaps there is a case for providing examples of valid formatted files to act as "Gold Standards".

[rapidlasso]

Looks like the main specification document has changed too much since that pull request was made and now needs manual tweaking to be merged. Maybe someone neutral should volunteer. Please keep the language system neutral.

It occurred to me that if any system needs "special mention" it should be the SPL system because their "single shot of many beamlets" could actually benefit from special consideration whereas GML is just another point cloud derived from multiple observations similar to SfM / dense-matching ... (-:

[kdamkjer]

To be clear, there's nothing preventing a GmAPD system from employing beamlets nor is it a defining characteristic of SPL.

[esilvia]

I should have time tomorrow morning to clear up the merge conflicts from the structural changes I've been doing on the draft. Thanks for your patience while we iron out the final kinks from the GitHub transition.

[esilvia]

I managed to resolve the merge conflicts WRT the source code reorganization and pushed the updates to @kdamkjer's branch. With those corrections this pull request looks good to me.

I have a couple questions about the wording that came up during my review, but that won't stop this pull request. I refreshed the issue branch and we can continue the conversation about details on #35.