clustertracking.find_link

clustertracking.find_link(reader, search_range, separation, diameter=None, memory=0, minmass=0, noise_size=1, smoothing_size=None, threshold=None, percentile=64, before_link=None, after_link=None, refine=False)

Find and link features in an image sequence

Parameters:

reader : pims.FramesSequence

search_range : number or tuple

maximum displacement of features between subsequent frames

separation : number or tuple

minimum separation distance between features

diameter : number or tuple, optional

feature diameter, used for characterization only. Default: separation.

memory : number, optional

number of frames that features are allowed to dissappear. Experimental. Default 0.

minmass : number, optional

minimum integrated intensity (in masked image). Default 0.

noise_size : number or tuple, optional

Size of Gaussian kernel with whith the image is convoluted for noise reduction. Default 1.

smoothing_size : number or tuple, optional

Size of rolling average box for background subtraction. By default, equals separation. This may introduce bias when refined on the background subtracted image!

threshold : number, optional

Threshold value for image. Default None.

percentile : number, optional

The upper percentile of intensities in the image are considered as feature locations. Default 64.

before_link : function, optional

This function is executed after the initial find of each frame, but but before the linking and relocating. It should take the following arguments (or **kwargs):

coords, reader, image, image_proc, diameter, separation, search_range, margin, minmass.

And it should return coords. coords is an ndarray containing the initially found feature coordinates. image and reader are unprocessed. image_proc is the processed image. Default None.

after_link : function, optional

This function is executed after the find and link of each frame. It should not change the number of features. It should take the following arguments (or **kwargs):

features, reader, image, image_proc, diameter, separation, search_range, margin, minmass.

And it should return features. features is a DataFrame containing the feature coordinates and characterization. image and reader are unprocessed. image_proc is the processed image. Default None.

refine : boolean, optional

Convenience parameter to do center-of-mass refinement. Cannot be used combined with an after_link function. Default False.