feature_matchers

Classes

FeatureMatcher	Abstract base class for any image matching process.
KeypointMatcher	Abstract base class defining the interface for keypoint detection and matching.
KeypointMatcherOptions	Options for configuring a KeypointMatcher
OrbKeypointMatcher	Implementation of KeypointMatcher using ORB for detection and FLANN for matching.
ORBKeypointMatcherOptions	Options for configuring the ORB keypoint matcher
SIFTKeypointMatcher	Implementation of KeypointMatcher using SIFT for detection and FLANN for matching.
SIFTKeypointMatcherOptions	Options for configuring the ORB keypoint matcher
RoMaFeatureMatcher	Implementation of a matcher using RoMa.
RoMaFeatureMatcherOptions	RoMaFeatureMatcherOptions(device: torch.device = <factory>, coarse_res: int | tuple[int, int] = 560, upsample_res: tuple[int, int] = (864, 864), sample_thresh: float = 0.05)

Enums and Config

FLANNCentersInit	Enum specifying the algorithm to use to initialize kmeans centers
FLANNIndexAlgorithmType	An enum specifying the FLANN algorithm to use.
FLANNIndexAutotunedParams	Parameters for confi of automatically selecting the best algorithm based on data
FLANNIndexCompositeParams	Parameters for config of combination of randomized k-d trees and hierarchical k-means
FLANNIndexHierarchicalParams	Parameters for config of hierarchical clustering
FLANNIndexKdTreeParams	Parameters for configuring k-dimensional tree
FLANNIndexKdTreeSingleParams	Parameters for configuring single randomized k-d tree (vs multiple trees in regular KDTREE)
FLANNIndexKMeansParams	Parameters for config of hierarchical k-means clustering
FLANNIndexLSHParams	Paramters for config of locality sensitive hashing
FLANNIndexLinearParams	Parameters for initializing brute force linear search
FLANNSearchParams	Parameters for configuring how FLANN does a search
FLANNDistance