-
Notifications
You must be signed in to change notification settings - Fork 1
Image_sort_by_statistics_v31
Sorts the input images for identifying junk particles The program associates to each image four vectors. One vector is composed by descriptors that encodes the particle shape. Another two vectors give information about the SNR of the objects. Finally, the last vector provides information of the particle histogram
These vector are then scored according to a multivariate Gaussian distribution
You can reject erroneous particles choosing a threshold, you must take into account that it is a zscore. For univariate and mulivariate Gaussian distributions, 99.
Parameters
$: Selfile with input images $`-o <rootname`> $`--train <selfile`> $`--zcut <float`-1> $`--percent <float`0> $: Add columns also to inputMetaData