Inheritance diagram for w2kmeans::KMeansSegmenter< T >:
Public Types | |
| typedef Cluster< SingleFeature > | RegionProperty |
| The simplest region property is a cluster of a single feature. | |
| typedef std::vector< int > | Replacement |
| Given a region with label i, Replacement[i] provides the new label, usually at a particular scale. | |
Public Member Functions | |
| virtual w2img::SegmenterResult | segment (const code::Image< int > &orig) |
| The resulting image has pixels marked with the region number, starting at 1. | |
| virtual | ~KMeansSegmenter () |
| KMeansSegmenter (int min_val, int max_val, size_t number_of_clusters, int max_depth) | |
| Pass in the range of values to consider in thresholding. | |
| w2img::SegmenterResult | getScale (size_t scale, bool rawlabels=false) const |
| The different scales. | |
| const Replacement & | getReplacement (size_t scale) const |
| Gets the replacement at different scales. | |
| size_t | getNumScales () const |
| Returns the total number of scales in the segmentation. | |
Static Public Member Functions | |
| static void | setMarkovWeight (float lambda) |
| Set the weight of the Markov energy relative to the Discontiguity measure. | |
| static void | setMinimumRegionSizes (const std::string &minimum_allowed) |
| Set the size thresholds for the scales. | |
| static void | keepAllSizes (bool flag) |
| By default, isolated, small regions will be removed at successive scales. | |
| static void | setSizeParams (const std::string ¶ms) |
| Convenient way to set the sizes and allSizes option through a single string. | |
| static size_t | getMaxScales () |
| Returns the maximum number of scales in the segmentation. | |
| static void | computeRegionProperty (const code::Image< int > &orig, const SegmenterResult &label, std::vector< RegionProperty > *fillme) |
| Fills in the passed in vector with a vector of computed region properties. | |
| static void | computeRegionProperty (const SegmenterResult &label, std::vector< RegionProperty > *fillme) |
| Fills in the passed in vector with a vector of computed region properties. | |
| static void | normalize (Segmenter::Replacement &replacement) |
| completes the cycle | |
| static SegmenterResult | relabel (const SegmenterResult &label_img) |
| Will relabel the image, removing unused labels. | |
Protected Member Functions | |
| void | doCurvatureCheck (w2img::SegmenterResult &label) const |
| If there is a highly curved or elongated large object, split it into smaller objects. | |
| code::Image< int > | doKMeans (const w2img::ImageAttr< T > &attr, code::Image< int > *init_clusters, std::vector< w2img::Cluster< T > > *clusters, int iter_no=0) const |
| Perform K means segmentation, returns a better set of clusters. | |
| code::Image< int > | initialize (const code::Image< int > &orig, const w2img::ImageAttr< T > &attr, std::vector< w2img::Cluster< T > > *clusters) const |
| Get an initial cut of clusters, somehow. | |
| void | computeClusterAttr (const w2img::ImageAttr< T > &attr, const code::Image< int > &clusterid, std::vector< w2img::Cluster< T > > *clusters, int max_clusterid) const |
| Compute cluster attributes from the image attribute and labeling. | |
| int | findBestCluster (const T ¤t, const std::vector< w2img::Cluster< T > > &clusters, const code::Image< int > &clusterid, int x, int y) const |
| Return label corresponding to best cluster available. | |
| float | computeMeasureDistance (const w2img::Cluster< T > &cluster, const T ¤t, int x, int y) const |
| Way to compute distance between a point (value,x,y) and a cluster's mean. | |
| float | computeMeasureDistance (const T &other, const T ¤t) const |
| Finds distance between two statistics. | |
| float | computeDiscontiguityDistance (const code::Image< int > &label, int x, int y, int candidate_label) const |
| Way to compute discontiguity if the current label were to be changed to the candidate. | |
| virtual void | computeReplacements (Pruner &pruner) |
| This function may be over-ridden to provide a different way of choosing how to combine/replace regions at higher levels. | |
| void | expand_segment (int x, int y, const code::Image< int > &orig, code::Image< int > &segimg, int label) const |
| Set the segimg at x,y to be label and then call this function. | |
Static Protected Attributes | |
| static const int | Unset = -1000 |
| static const int | Invalid = -2000 |
Classes | |
| class | Pruner |
| Controls pruning of regions based on size. More... | |
| struct | SurrRegion |
| Used by the pruner to control pruning of regions based on size. More... | |
typedef Cluster< SingleFeature > w2img::Segmenter::RegionProperty [inherited] |
The simplest region property is a cluster of a single feature.
typedef std::vector<int> w2img::Segmenter::Replacement [inherited] |
Given a region with label i, Replacement[i] provides the new label, usually at a particular scale.
The Replacement structure is such that replacement[i] provides the replacement label for the current label i. This replacement may be 0 or negative, which means that no replacement exists.
| virtual w2kmeans::KMeansSegmenter< T >::~KMeansSegmenter | ( | ) | [inline, virtual] |
| w2kmeans::KMeansSegmenter< T >::KMeansSegmenter | ( | int | min_val, | |
| int | max_val, | |||
| size_t | number_of_clusters, | |||
| int | max_depth | |||
| ) |
Pass in the range of values to consider in thresholding.
Specify the number of initial quantization levels and the maximum depth (in quantization levels) that a cluster can range over without being too dissimilar. Specify a -ve value for max_depth if you don't want dissimilarity to affect cluster combination.
| void w2kmeans::KMeansSegmenter< T >::computeClusterAttr | ( | const w2img::ImageAttr< T > & | attr, | |
| const code::Image< int > & | clusterid, | |||
| std::vector< w2img::Cluster< T > > * | clusters, | |||
| int | max_clusterid | |||
| ) | const [protected] |
Compute cluster attributes from the image attribute and labeling.
During K-means, the computation is always done with K levels, but pass in false if you are using this after K-means and relabeling.
| float w2kmeans::KMeansSegmenter< T >::computeDiscontiguityDistance | ( | const code::Image< int > & | label, | |
| int | x, | |||
| int | y, | |||
| int | candidate_label | |||
| ) | const [protected] |
Way to compute discontiguity if the current label were to be changed to the candidate.
Returns a number between 0 and 1.
| float w2kmeans::KMeansSegmenter< T >::computeMeasureDistance | ( | const T & | other, | |
| const T & | current | |||
| ) | const [protected] |
Finds distance between two statistics.
| float w2kmeans::KMeansSegmenter< T >::computeMeasureDistance | ( | const w2img::Cluster< T > & | cluster, | |
| const T & | current, | |||
| int | x, | |||
| int | y | |||
| ) | const [protected] |
Way to compute distance between a point (value,x,y) and a cluster's mean.
The base class does not use the location information, but derived classes may want to. Between 0 and 1.
| static void w2img::Segmenter::computeRegionProperty | ( | const SegmenterResult & | label, | |
| std::vector< RegionProperty > * | fillme | |||
| ) | [static, inherited] |
Fills in the passed in vector with a vector of computed region properties.
The input label image should be a labeled image such as that returned by the segmenter. WARNING: The resulting properties will not have any of the value elements set because the original image was not passed in.
| static void w2img::Segmenter::computeRegionProperty | ( | const code::Image< int > & | orig, | |
| const SegmenterResult & | label, | |||
| std::vector< RegionProperty > * | fillme | |||
| ) | [static, inherited] |
Fills in the passed in vector with a vector of computed region properties.
The input label image should be a labeled image such as that returned by the segmenter.
| virtual void w2kmeans::KMeansSegmenter< T >::computeReplacements | ( | Pruner & | pruner | ) | [protected, virtual] |
This function may be over-ridden to provide a different way of choosing how to combine/replace regions at higher levels.
| label | The image with the K-means segmentation done and labels computed based on 8-neighborhood, but no selection among the regions performed. | |
| pruner | to be used. | |
| replacements_list | The list of replacements; the first element (most detail) should be filled already. Coarser detail will be filled in here. |
| void w2kmeans::KMeansSegmenter< T >::doCurvatureCheck | ( | w2img::SegmenterResult & | label | ) | const [protected] |
If there is a highly curved or elongated large object, split it into smaller objects.
changes the labels in the image passed in.
| code::Image<int> w2kmeans::KMeansSegmenter< T >::doKMeans | ( | const w2img::ImageAttr< T > & | attr, | |
| code::Image< int > * | init_clusters, | |||
| std::vector< w2img::Cluster< T > > * | clusters, | |||
| int | iter_no = 0 | |||
| ) | const [protected] |
Perform K means segmentation, returns a better set of clusters.
These can be segmented using the ContourSegmenter to yield a w2img::SegmenterResult.
| void w2img::Segmenter::expand_segment | ( | int | x, | |
| int | y, | |||
| const code::Image< int > & | orig, | |||
| code::Image< int > & | segimg, | |||
| int | label | |||
| ) | const [protected, inherited] |
Set the segimg at x,y to be label and then call this function.
It will expand out segimg to all contiguous values that match
| int w2kmeans::KMeansSegmenter< T >::findBestCluster | ( | const T & | current, | |
| const std::vector< w2img::Cluster< T > > & | clusters, | |||
| const code::Image< int > & | clusterid, | |||
| int | x, | |||
| int | y | |||
| ) | const [protected] |
Return label corresponding to best cluster available.
| static size_t w2kmeans::KMeansSegmenter< T >::getMaxScales | ( | ) | [static] |
| size_t w2kmeans::KMeansSegmenter< T >::getNumScales | ( | ) | const [virtual] |
Returns the total number of scales in the segmentation.
The returned values correspond to the previous call to segment()
Implements w2kmeans::MultiscaleSegmenter.
| const Replacement& w2kmeans::KMeansSegmenter< T >::getReplacement | ( | size_t | scale | ) | const |
Gets the replacement at different scales.
0 is the most detail, 1 is less and so on. If the scale passed in exceeeds the number of scales in the segmentation, an invalid Replacement ( .size() == 0 ) is returned.
The returned values correspond to the previous call to segment()
| w2img::SegmenterResult w2kmeans::KMeansSegmenter< T >::getScale | ( | size_t | scale, | |
| bool | rawlabels = false | |||
| ) | const [virtual] |
The different scales.
The returned values correspond to the previous call to segment()
| scale | The scale to return. 0 is the most detail, 1 is less ... | |
| rawlabels | If true, then no relabeling of the original image is done. If false, the labeling corresponds to that of raw_segimg. |
Implements w2kmeans::MultiscaleSegmenter.
| code::Image<int> w2kmeans::KMeansSegmenter< T >::initialize | ( | const code::Image< int > & | orig, | |
| const w2img::ImageAttr< T > & | attr, | |||
| std::vector< w2img::Cluster< T > > * | clusters | |||
| ) | const [protected] |
Get an initial cut of clusters, somehow.
The implementation here is to divide the measurement space (max-min) into equal parts.
| static void w2kmeans::KMeansSegmenter< T >::keepAllSizes | ( | bool | flag | ) | [static] |
By default, isolated, small regions will be removed at successive scales.
Set this flag to 'true' if you want larger scales to contain the isolated regions from smaller scales.
| static void w2img::Segmenter::normalize | ( | Segmenter::Replacement & | replacement | ) | [static, inherited] |
completes the cycle
| static SegmenterResult w2img::Segmenter::relabel | ( | const SegmenterResult & | label_img | ) | [static, inherited] |
Will relabel the image, removing unused labels.
| virtual w2img::SegmenterResult w2kmeans::KMeansSegmenter< T >::segment | ( | const code::Image< int > & | orig | ) | [virtual] |
The resulting image has pixels marked with the region number, starting at 1.
Zero values do not belong to any region.
What gets returned is the segmentation with the most detail. To obtain the segmentation at coarser levels of detail, use the getScale() function.
This is not really a constant function -- getScale() and getReplacement()'s results will change because of this function.
Implements w2kmeans::MultiscaleSegmenter.
| static void w2kmeans::KMeansSegmenter< T >::setMarkovWeight | ( | float | lambda | ) | [static] |
Set the weight of the Markov energy relative to the Discontiguity measure.
The default value is 0.6 and should be left alone unless you know what you are doing. If wish to change it, provide any number in the range [0,1].
This is more that "just" a static function -- you are setting it for all values of the template parameter T as well!
| static void w2kmeans::KMeansSegmenter< T >::setMinimumRegionSizes | ( | const std::string & | minimum_allowed | ) | [static] |
Set the size thresholds for the scales.
By default, this string is 20,160,480
This is more that "just" a static function -- you are setting it for all values of the template parameter T as well!
| static void w2kmeans::KMeansSegmenter< T >::setSizeParams | ( | const std::string & | params | ) | [static] |
Convenient way to set the sizes and allSizes option through a single string.
The default values are 20,160,480:0
const int w2img::Segmenter::Invalid = -2000 [static, protected, inherited] |
const int w2img::Segmenter::Unset = -1000 [static, protected, inherited] |
1.4.7