Microscopy system and method for generating an HDR image

A method for generating an HDR image, comprising

• receiving a plurality of raw images showing a scene with different image brightnesses;
• setting coefficients for different regions in the raw images depending on objects depicted in the regions, wherein the coefficients define whether or to what extent pixels of the raw images are incorporated in the HDR image;
• and generating the HDR image by combining at least two of the raw images based on the set coefficients,
• wherein the setting coefficients for different regions includes:
• determining semantic regions in the raw images and assigning said regions respective semantic labels;
• and
• deriving coefficients for each of the raw images depending on at least the semantic labels of the included semantic regions;
• wherein determination of the semantic labels occurs based on one or more of the raw images or based on another image, and
• wherein the semantic labels of the semantic regions discriminate between at least two or more of the following predefined classes:
• sample carrier, cover slip, sample, holding frame, and background.

A method for generating an HDR image, comprising

• receiving a plurality of raw images showing a scene with different image brightnesses;
• setting coefficients for different regions in the raw images depending on objects depicted in the regions, wherein the coefficients define whether or to what extent pixels of the raw images are incorporated in the HDR image;
• and
• generating the HDR image by combining at least two of the raw images based on the set coefficients,
• wherein the setting coefficients for different regions includes:
• determining semantic regions in the raw images and assigning said regions respective semantic labels;
• and
• deriving coefficients for each of the raw images depending on at least the semantic labels of the included semantic regions;
• wherein determination of the semantic labels occurs based on one or more of the raw images or based on another image, and
• wherein the semantic labels of the semantic regions discriminate between predefined classes, at least two or more of which relate to microscopic samples, sample holders, or sample surroundings.