delta.pipeline.Position

class delta.pipeline.Position(position_nb: int, reader: xpreader, models: Dict, drift_correction: bool = True, crop_windows: bool = False)

Position processing object

__init__(position_nb: int, reader: xpreader, models: Dict, drift_correction: bool = True, crop_windows: bool = False)

Initialize Position

Parameters

position_nbint

Position index.

readerobject

utilities xpreader object.

modelsdict

U-Net models as loaded by utilities loadmodels().

drift_correctionbool, optional

Flag to perform drift correction. The default is True.

crop_windowsbool, optional

Flag to crop out windows. The default is False.

Returns

None.

Attributes

position_nb	Position index number in the experiment
reader	Experiment reader object
models	List of Tensorflow models
rois	List of ROI objects under position
drift_values	XY drift correction values over time
drift_correction	Flag to perform drift correction
crop_windows	Flag to crop overlapping windows for segmentation
verbose	Console output verbosity

Methods

__init__(position_nb, reader, models[, ...])	Initialize Position
clear()	Clear Position-specific variables from memory (can be loaded back with load())
detect_rois(reference)	Use U-Net to detect ROIs (chambers etc...)
features(frames[, features])	Extract features for all ROIs in frames
legacysave(res_file)	Save pipeline data in the legacy Matlab format
load(filename)	Load position from pickle file
preprocess([reference, rotation_correction])	Pre-process position (Rotation correction, identify ROIs, initialize drift correction)
results_movie([frames])	Generate movie illustrating segmentation and tracking
save([filename, frames, save_format])	Save to disk
segment(frames)	Segment cells in all ROIs in position
track(frames)	Track cells in all ROIs in frames