Darwin  1.10(beta)
 All Classes Namespaces Files Functions Variables Typedefs Enumerations Enumerator Friends Groups Pages
Functions
drwnNNGraphMoves Namespace Reference

Templated drwnNNGraph search moves. More...

Functions

template<class DistanceMetric >
void initialize (drwnNNGraph &graph, const DistanceMetric &M)
 randomly initialize edges (matched) for all active images (keeps existing matches unless an improvement is found)
 
void initialize (drwnNNGraph &graph)
 randomly initialize edges (matched) for all active images (keeps existing matches unless an improvement is found) using default metric
 
template<class DistanceMetric >
void initialize (drwnNNGraph &graph, unsigned imgIndx, const DistanceMetric &M)
 randomly initialize edges (matched) for a given image (keeps existing matches unless an improvement is found)
 
template<class DistanceMetric >
double rescore (drwnNNGraph &graph, const DistanceMetric &M)
 rescore all matches based on current features (which may have changed since graph was created)
 
double rescore (drwnNNGraph &graph)
 rescore all matches based on current features (which may have changed since graph was created) using default metric
 
template<class DistanceMetric >
double rescore (drwnNNGraph &graph, unsigned imgIndx, const DistanceMetric &M)
 rescore all matches for a given image based on current features (which may have changed since graph was created) More...
 
template<class DistanceMetric >
void flann (drwnNNGraph &graph, const DistanceMetric &M)
 update initial edges using FLANN (with possible self-matches in both source and destination matchable)
 
template<class DistanceMetric >
void update (drwnNNGraph &graph, const DistanceMetric &M)
 perform one update iteration (including enrichment and exhaustive) on all active images
 
void update (drwnNNGraph &graph)
 perform one update iteration (including enrichment and exhaustive) on all active images using default metric
 
template<class DistanceMetric >
bool propagate (drwnNNGraph &graph, const drwnNNGraphNodeIndex &u, const DistanceMetric &M)
 propagate good matches from supeprixel u to its neighbours (returns true if a better match was found)
 
template<class DistanceMetric >
bool search (drwnNNGraph &graph, const drwnNNGraphNodeIndex &u, const DistanceMetric &M)
 random search across all images for superpixel u (returns true if a better match was found)
 
template<class DistanceMetric >
bool local (drwnNNGraph &graph, const drwnNNGraphNodeIndex &u, const DistanceMetric &M)
 local neighbourhood search around current match for superpixel u (returns true if a better match was found)
 
template<class DistanceMetric >
bool randproj (drwnNNGraph &graph, const DistanceMetric &M)
 random projection moves that project all superpixels onto a random direction, sort and compare adjacent superpixels if from different images (returns true if a better match was found)
 
template<class DistanceMetric >
bool enrichment (drwnNNGraph &graph, const DistanceMetric &M)
 enrichment: inverse (from target to source) and forward (from source to target's target) (returns true if a better match was found)
 
template<class DistanceMetric >
bool exhaustive (drwnNNGraph &graph, const drwnNNGraphNodeIndex &u, const DistanceMetric &M)
 exhaustive search for best match across entire graph — use sparingly (returns true if a better match was found)
 

Detailed Description

Templated drwnNNGraph search moves.

Function Documentation

template<class DistanceMetric >
double drwnNNGraphMoves::rescore ( drwnNNGraph graph,
unsigned  imgIndx,
const DistanceMetric &  M 
)

rescore all matches for a given image based on current features (which may have changed since graph was created)

Todo:
delete the node