List keys

extract array face adjacencies for each face in the mesh (3 * number of faces); each triple describes the adjacent face triangles for a given face in

extract array of triangles incident to each vertex and check each vertex if it is at the boundary or not

extract array of triangles incident to each vertex and check each vertex if it is at the boundary or not

load a text list of views and their neighbors and assign them to the scene images; each line store the view ID followed by the 3+ closest view IDs, or

Compute SGM stereo for this image and each of the neighbor views: - minResolution is the resolution of the top of the pyramid for tSGM; can be 0 to fo

fuse all depth-maps by simply projecting them in a 3D point-cloud in the world coordinate space

Returns true if this node has no children.

Function to detect AVX availability in operating system.

Function to detect SSE availability in operating system.

parse scene stored in ContextCapture BlocksExchange format containing cameras, images and sparse point-cloud

parse images list containing calibration and pose information and load the corresponding 3D point-cloud

parse scene stored in Polycam format

parse scene stored in MVSNet format composed of undistorted images and camera poses for example see GigaVision benchmark (http://gigamvs.net): |--scen

RTMV scene format: http://www.cs.umd.edu/~mmeshry/projects/rtmv |--sceneX |--images |--xxx.jpg |--xxx.jpg .... |--outputs (optional) |--depthxxxx.exr

parse scene stored either in Metashape images list format or ContextCapture BlocksExchange format

Perform a GPU pick around screen pixel position with given radius (pixels); if a primitive is found returns valid PickResult, where pick.idx is the pr

test MVS stages on a small sample dataset

filter point-cloud based on camera-point visibility intersections

apply a gradient like filter

center a patch of given size on the segment

assuming the mesh is properly oriented, ortho-project it to a camera looking from top to down

See QueryIntAttribute()

generate random depth and normal

optimize mesh using photo-consistency

* Released all used memory for this nod and its children */