a41c1ed49bcf7554eaf113d9302bca05342bbeb0,imgaug/augmenters/arithmetic.py,,CoarseSalt,#Any#Any#Any#Any#Any#Any#Any#Any#,1079

Before Change

    

    if ia.is_single_number(p):
        mask = Binomial(p)
    elif isinstance(p, tuple):
        ia.do_assert(len(p) == 2)
        ia.do_assert(0 <= p[0] <= 1.0)
        ia.do_assert(0 <= p[1] <= 1.0)
        mask = Binomial(Uniform(p[0], p[1]))
    elif ia.is_iterable(p):
        ia.do_assert(all([0 <= pi <= 1.0 for pi in p]))
        mask = Binomial(iap.Choice(p))
    elif isinstance(p, StochasticParameter):
        mask = p
    else:
        raise Exception("Expected p to be number or tuple of two number or list of number or StochasticParameter, got %s." % (type(p),))

    if size_px is not None:
        mask_low = FromLowerResolution(other_param=mask, size_px=size_px, min_size=min_size)
    elif size_percent is not None:
        mask_low = FromLowerResolution(other_param=mask, size_percent=size_percent, min_size=min_size)

After Change

    to the input image size, leading to large rectangular areas being replaced.

    
    mask = iap.handle_probability_param(p, "p", tuple_to_uniform=True, list_to_choice=True)

    if size_px is not None:
        mask_low = FromLowerResolution(other_param=mask, size_px=size_px, min_size=min_size)
    elif size_percent is not None:

Instances