Semantic thresholding

The use of thresholding to separate objects in an image is analyzed in the light of a notion that ‘objects are in the eye of the beholder’, i.e. the existence of objects in an image is subjective rather than objective. Interpretation of the term ‘semantic’ as an association of an observed entity with an (other) entity which can be used for communication, is combined with Whorf's hypothesis on influence of language on reasoning, to yield a thresholding method. The method is based on minimizing the ambiguity which results from the different contexts (local and global) in which grey levels are classified as dark or bright. Examples are supplied and compared to thresholds which are optimal in the L₂ norm.