Six cases where reshaping intensity sets up the next step:
Flat-field correction. An illumination gradient across the field shifts the apparent intensity of identical biology from one position to the next. Uneven illumination multiplies the signal, so acquire or estimate a flat-field reference image and divide the sample by it; subtract a background estimate only where the background adds a constant offset. With the per-position bias gone, a single threshold reads the same biology the same way across the field.
Ratiometric measurements. Sometimes the biology is in the ratio of two channels imaged on the same field: a calcium indicator, a FRET donor/acceptor pair, a pH-sensitive dye. Divide images gives a per-pixel ratio map. That map reads out the property the ratio encodes, independent of absolute intensity.
Dynamic-range compression. Microscopy images often span four orders of magnitude, from dim cytoplasm to bright nuclei, and Linear thresholdsLoading... work poorly across that range. Logarithm compresses the high end and expands the low end. Afterward dim and bright sit on comparable footing, and one threshold can address both.
Difference imaging. Image the same field at two time points and subtract images. The static structure cancels, and what changed surfaces.
Inversion for downstream methods that require a bright foreground. Some detection engines require the foreground to be bright. Where it is dark, as in transmitted-light brightfield and certain stains, complement swaps the convention and keeps relative intensity differences intact.
Channel weighting and composition. Add weighted images combines two channels with weights you choose. Use it to synthesize a composite signal, derive a channel from a linear combination, or build a single-channel projection from multichannel data.