Update to latest develop from upstream - #1
Closed
smathermather wants to merge 340 commits into
Closed
Conversation
(cherry picked from commit 9d93e51b00def7eccab3eab029427bf0b1789b62)
(cherry picked from commit 0689083e109098bba56f5c733a976216db56fdc0)
(cherry picked from commit 404000b1e3031024ac33ddc50ffee34824347b0e)
(cherry picked from commit 7d10647d6224cafdc2ae472d0987ca6252329c27)
(cherry picked from commit a4405ec4f8d83f119fc7291963c9935be8c328dd)
Drop the FD2R/FR2D and D2R/R2D macros for constexpr function templates over any floating-point type, so every caller states its precision explicitly; CBRT uses std::cbrt instead of pow. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Gravity-aware up-axis detection (sign vote over the subsampled cloud with a camera-plausibility gate), support-driven bounds with scale-invariant point weights, per-axis max-gap growth and a coverage fallback ladder, min-area footprint orientation, and a gravity-aware tower gate so horizontal linear trajectories are not misclassified as towers. OBB fitting gains a fixed-axis mode and defect fixes; exporting an empty ROI now fails loudly. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Validate sub-scene merges with Sim(3) cycle residuals, splitting instead of demoting on failure and rebalancing clusters afterwards; split thin-waist clusters at creation (Fiedler cut) with a coupling diagnostic; optional community-detection clustering behind --cluster-communities; robustified pair weighting; drop the merge-time sub-scene repair and its two-view health check; bound the memory used while sampling track colors. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Move the .dmap codec into the drop-in Interface.h (OpenCV backend plus a no-OpenCV fallback writing byte-identical files); DepthMap.cpp and SFM/InterfaceMVS.cpp only adapt the scene types to it, dropping the SFM duplicate. Densify decodes images on demand through the new gray-storing ImageCache and estimates depth-maps in view-graph order (SortImagesByViewLocality) so consecutive ones share views; fusion reloads pixels through DMapCache and point colors are sampled image-major, so no stage needs every image resident. All caches report hit rates through the shared CacheHitStats. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
… with verification feedback (cdcseacave#1291)
…pped-cluster pixels (cdcseacave#1298)
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
20 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
New fusion approach, possible new meshing with TSDF.