Multimask about pymasker HOT 3 CLOSED

Hi Roman,
If I understand the BAQ file and pymasker properly, medium confidence mask should have high confidence clouds in it. I think it works this way:
1 High confidence mask will have only those clouds that statistically there is no doubt that they are clouds
2 Medium confidence mask should have high confidence clouds also with those clouds that statistically the propability that they are clouds is medium
3 Low confidence mask should have both high and medium clouds with those clouds that statistically the probalibity of them to be clouds is low.

from pymasker.

@roemelll If you are already successful to create medium confidence and high confidence cloud mask separately, I would recommend to do a or operation for two masks. The mask is just a two-dimension numpy array of 1 and 0. numpy.bitwise_or can be helpful in this case. The function get_multi_mask is not very well designed so I drop it at the current version of pymasker. If that function is still very useful to you, please let me know.

@lvhengani You are right when the cumulative option for the mask generation function is explicitly set as True. Currently, this options is False by default so the mask generated with no extra cumulative setting will be strictly representing the input confidence. The original masking value used in the BAQ band represent the state distinctly, or accumulative. See the BAQ band documentation for pre-collection and collection-1 for more details.

from pymasker.

@haoliangyu thank you for clarity, I have been running with cumulative = True. This makes sense to me than to do another process for merging the two.

from pymasker.