Answers are grounded in this paper and cite their sources.
An integrated methodology for the detection and removal of cracks on digitized is presented in this paper. Crack-like pattern detection have been a matter of high among researchers mostly for its useful contribution to a variety of applications. The presented here regard the craquelure of old paintings, however, the same methodology be used for a much wider set of application. Many images contain similar patterns: crack in coating for polymers and other surfaces; fatigue crack in MEMS/NEMS; crack in used for underfill and encapsulation microelectronics components; etc. In this paper cracks are detected by thresholding the output of the morphological top-hat transform. , the thin dark brush strokes which have been misidentified as cracks are removed automatic procedure. Finally, crack filling using texture synthesis algorithms. The has been shown to perform very well on digitized images suffering from cracks.
No takes yet. Share an insight, caveat, or question.
Spagnolo et al. (2010) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: