i/rtabmap
1
0
Fork 0
mirror of https://github.com/introlab/rtabmap.git synced 2026-10-05 17:47:49 +08:00
Code Issues Packages Projects Releases Wiki Activity

Parallelize cloud generation in export/view clouds dialog and rtabmap-export (#1757)

Browse Source 
* Parallelize cloud generation in export/view clouds dialog and rtabmap-export

* Simplified NodeExportData by holding Signature directly. Added --threads option (default max cores) for CLI and UI.

* Parallelized texturing phase the same way than cloud generation

* Fixed not cancelable (right away) cloud generation in UI

---------

Co-authored-by: matlabbe <[email protected]>
This commit is contained in:
Torjus Iveland
2026-08-30 17:49:28 -07:00
committed by GitHub
co-authored by matlabbe
parent 9279ab68ca
commit 7eb26e1dc2
8 changed files with 901 additions and 515 deletions
Download Patch File Download Diff File Expand all files Collapse all files
corelib/include/rtabmap/core/util3d_surface.h
+4 -2
View File
@@ -150,7 +150,8 @@ pcl::TextureMesh::Ptr RTABMAP_CORE_EXPORT createTextureMesh(
const std::vector<float> & roiRatios = std::vector<float>(), // [left, right, top, bottom] region of interest (in ratios) of the image projected.
const ProgressState * state = 0,
std::vector<std::map<int, pcl::PointXY> > * vertexToPixels = 0, // For each point, we have a list of cameras with corresponding pixel in it. Beware that the camera ids don't correspond to pose ids, they are indexes from 0 to total camera models and texture's materials.
bool distanceToCamPolicy = false);
bool distanceToCamPolicy = false,
int numThreads = 1); // number of threads used to compute the visible faces of the cameras (1=sequential)
pcl::TextureMesh::Ptr RTABMAP_CORE_EXPORT createTextureMesh(
const pcl::PolygonMesh::Ptr & mesh,
const std::map<int, Transform> & poses,
@@ -163,7 +164,8 @@ pcl::TextureMesh::Ptr RTABMAP_CORE_EXPORT createTextureMesh(
const std::vector<float> & roiRatios = std::vector<float>(), // [left, right, top, bottom] region of interest (in ratios) of the image projected.
const ProgressState * state = 0,
std::vector<std::map<int, pcl::PointXY> > * vertexToPixels = 0, // For each point, we have a list of cameras with corresponding pixel in it. Beware that the camera ids don't correspond to pose ids, they are indexes from 0 to total camera models and texture's materials.
bool distanceToCamPolicy = false);
bool distanceToCamPolicy = false,
int numThreads = 1); // number of threads used to compute the visible faces of the cameras (1=sequential)
/**
* Remove not textured polygon clusters. If minClusterSize<0, only the largest cluster is kept.