cvkit.pose_estimation.processors.filter package

Submodules

cvkit.pose_estimation.processors.filter.distance_statistics_filter module

class DistanceStatisticsFilter(distance_matrix_mean, distance_matrix_sd, threshold=0.5, sd_factor=1.25, likelihood_threshold=0.6)

Bases: Processor

META_DATA = {'distance_matrix_mean': <cvkit.pose_estimation.processors.processor_interface.ProcessorMetaData object>, 'distance_matrix_sd': <cvkit.pose_estimation.processors.processor_interface.ProcessorMetaData object>, 'likelihood_threshold': <cvkit.pose_estimation.processors.processor_interface.ProcessorMetaData object>, 'sd_factor': <cvkit.pose_estimation.processors.processor_interface.ProcessorMetaData object>, 'threshold': <cvkit.pose_estimation.processors.processor_interface.ProcessorMetaData object>}

Processor parameters’ metadata. UI generator uses this information to generate a form.

PROCESSOR_ID = 'cvkit_distance_stats'

Unique Processor Identifier.

PROCESSOR_NAME = 'Distance Statistics Filter'

Name of the processor. Used as label by the UI generator.

PROCESSOR_SUMMARY = 'Uses Mean and Standard deviation of distance among body parts to filter outliers'

Processor summary.

cvkit.pose_estimation.processors.filter.kalman_filter module

class KalmanFilter(target_column, framerate, skip=True, threshold=0.6)

Bases: Processor

DISTRIBUTED = True

Indicates whether the processor operates at body part level or skeleton level. If true, this processor can be parallelized for each body part.

META_DATA = {'framerate': <cvkit.pose_estimation.processors.processor_interface.ProcessorMetaData object>, 'skip': <cvkit.pose_estimation.processors.processor_interface.ProcessorMetaData object>, 'target_column': <cvkit.pose_estimation.processors.processor_interface.ProcessorMetaData object>, 'threshold': <cvkit.pose_estimation.processors.processor_interface.ProcessorMetaData object>}

Processor parameters’ metadata. UI generator uses this information to generate a form.

PROCESSOR_ID = 'cvkit_kalman_filter'

Unique Processor Identifier.

PROCESSOR_NAME = 'Kalman Filtering'

Name of the processor. Used as label by the UI generator.

PROCESSOR_SUMMARY = 'Constant acceleration Kalman Filter'

Processor summary.

cvkit.pose_estimation.processors.filter.linear_interpolation module

class LinearInterpolationFilter(target_column, threshold=0.6, max_cluster_size=10)

Bases: Processor

DISTRIBUTED = True

Indicates whether the processor operates at body part level or skeleton level. If true, this processor can be parallelized for each body part.

META_DATA = {'max_cluster_size': <cvkit.pose_estimation.processors.processor_interface.ProcessorMetaData object>, 'target_column': <cvkit.pose_estimation.processors.processor_interface.ProcessorMetaData object>, 'threshold': <cvkit.pose_estimation.processors.processor_interface.ProcessorMetaData object>}

Processor parameters’ metadata. UI generator uses this information to generate a form.

PROCESSOR_ID = 'cvkit_interpolation'

Unique Processor Identifier.

PROCESSOR_NAME = 'Linear Interpolation'

Name of the processor. Used as label by the UI generator.

PROCESSOR_SUMMARY = 'Performs linear interpolation to fill missing body parts.'

Processor summary.

REQUIRES_STATS = True

Indicates whether this processor requires data statistics generated by py:class:.ClusterAnalysis

cvkit.pose_estimation.processors.filter.median_distance_culling module

class MedianDistanceFilter(threshold=0.6, distance_threshold=400)

Bases: Processor

META_DATA = {'distance_threshold': <cvkit.pose_estimation.processors.processor_interface.ProcessorMetaData object>, 'threshold': <cvkit.pose_estimation.processors.processor_interface.ProcessorMetaData object>}

Processor parameters’ metadata. UI generator uses this information to generate a form.

PROCESSOR_ID = 'cvkit_median_distance_culling'

Unique Processor Identifier.

PROCESSOR_NAME = 'Median Distance Culling'

Name of the processor. Used as label by the UI generator.

PROCESSOR_SUMMARY = 'Computes distance matrix among all body parts and filters outliers based on median distance.'

Processor summary.

cvkit.pose_estimation.processors.filter.moving_average module

class MovingAverageFilter(target_column, window_size, threshold=0.6)

Bases: Processor

DISTRIBUTED = True

Indicates whether the processor operates at body part level or skeleton level. If true, this processor can be parallelized for each body part.

META_DATA = {'target_column': <cvkit.pose_estimation.processors.processor_interface.ProcessorMetaData object>, 'threshold': <cvkit.pose_estimation.processors.processor_interface.ProcessorMetaData object>, 'window_size': <cvkit.pose_estimation.processors.processor_interface.ProcessorMetaData object>}

Processor parameters’ metadata. UI generator uses this information to generate a form.

PROCESSOR_ID = 'cvkit_moving_average'

Unique Processor Identifier.

PROCESSOR_NAME = 'Moving Average'

Name of the processor. Used as label by the UI generator.

PROCESSOR_SUMMARY = 'Runs a moving average filter to reduce noise.'

Processor summary.

cvkit.pose_estimation.processors.filter.region_filter_2d module

class RegionFilter2D(uncertainty_regions)

Bases: Processor

META_DATA = {'uncertainty_regions': <cvkit.pose_estimation.processors.processor_interface.ProcessorMetaData object>}

Processor parameters’ metadata. UI generator uses this information to generate a form.

PROCESSOR_ID = 'cvkit_2d_region_filter'

Unique Processor Identifier.

PROCESSOR_NAME = '2D Region Filter'

Name of the processor. Used as label by the UI generator.

PROCESSOR_SUMMARY = 'Deletes body parts lying in provided 2D regions of uncertainty.'

Processor summary.

REQUIRES_STATS = True

Indicates whether this processor requires data statistics generated by py:class:.ClusterAnalysis

cvkit.pose_estimation.processors.filter.velocity_filter module

class VelocityFilter(target_column, threshold, framerate, threshold_velocity)

Bases: Processor

DISTRIBUTED = True

Indicates whether the processor operates at body part level or skeleton level. If true, this processor can be parallelized for each body part.

META_DATA = {'framerate': <cvkit.pose_estimation.processors.processor_interface.ProcessorMetaData object>, 'target_column': <cvkit.pose_estimation.processors.processor_interface.ProcessorMetaData object>, 'threshold': <cvkit.pose_estimation.processors.processor_interface.ProcessorMetaData object>, 'threshold_velocity': <cvkit.pose_estimation.processors.processor_interface.ProcessorMetaData object>}

Processor parameters’ metadata. UI generator uses this information to generate a form.

PROCESSOR_ID = 'cvkit_velocity_filter'

Unique Processor Identifier.

PROCESSOR_NAME = 'Velocity Filter'

Name of the processor. Used as label by the UI generator.

PROCESSOR_SUMMARY = 'Filters body parts with velocity higher than the threshold.'

Processor summary.

REQUIRES_STATS = True

Indicates whether this processor requires data statistics generated by py:class:.ClusterAnalysis

Module contents