Quantifying the complexity of black-and-white images.

ABSTRACT: We propose a complexity measure for black-and-white (B/W) digital images, based on the detection of typical length scales in the depicted motifs. Complexity is associated with diversity in those length scales. In this sense, the proposed measure penalizes images where typical scales are limited to small lengths, of a few pixels -as in an image where gray levels are distributed at random- or to lengths similar to the image size -as when gray levels are ordered into a simple, broad pattern. We introduce a complexity index which captures the structural richness of images with a wide range of typical scales, and compare several images with each other on the basis of this index. Since the index provides an objective quantification of image complexity, it could be used as the counterpart of subj

SUBMITTER: Zanette DH 

PROVIDER: S-EPMC6258117 | biostudies-literature | 2018

REPOSITORIES: biostudies-literature