Masking

The KBMOD algorithm uses a data mask to represent invalid pixel values that should be ignored during the search. Masking is applied by:

• applying the per-image mask

• applying a computed a global mask

• growing the computed mask footprint

As stated previously (see Input Files), KBMOD expects Vera C. Rubin Science Pipelines calexp-style FITS files. Therefore each science image has an associated mask. Per-image masks are applied by checking whether the associated mask image contains one or more of the specified flags, see default_flag_keys attribute of run_search. These values can be overwritten by providing the flag_keys attribute of run_search. Futhermore, pixels with values that exceed a threshold value will be masked if a flux threshold is provided, see mask_threshold key in Search Parameters.

Global mask is computed for all images by counting how many individual masks have flagged a pixel for any reason in the list of allowed global masking flags, and masking it out if that count surpasses a given threshold. The number of times a pixel has to be flagged to be masked can be set by mask_num_images parameter (see Search Parameters) and the list of flags to use in global masking can be modified via the repeated_flag_keys parameter and is by default set to default_repeated_flag_keys. Global masking is not applied when the list of allowed masking flags is empty.

After the per-image and global masks are applied to every image, KBMOD grows the mask to nearby pixels. The parameters mask_grow (see Search Parameters) determines the ammount of the growth.

The provided pixel bitmask uses the following mapping between flag and bitmask values, which corresponds to the Rubin Science Pipelines mask values:

Key	Value
BAD	0
CLIPPED	9
CR	3
CROSSTALK	10
DETECTED	5
DETECTED_NEGATIVE	6
EDGE	4
INEXACT_PSF	11
INTRP	2
NOT_DEBLENDED	12
NO_DATA	8
REJECTED	13
SAT	1
SENSOR_EDGE	14
SUSPECT	7
UNMASKEDNAN	15

To overwrite the default mapping set mask_bits_dict parameter, see Search Parameters.