Crack detection in paintings
Being able to accurately detect cracks can be very relevant to
painting conservation, since cracking is one of the most common
forms of deterioration. Fluctuations in humidity, causing
the wooden support to shrink or expand, is the main reason for
crack formation. Because the way in which cracks develop and
spread partly depends upon the choice of materials and methods
used by the artist, assessing cracks is useful for judging authenticity
[4]. Cracks can also assist conservators by providing clues
to the causes of degradation of the paint surface. An in-depth
study of the factors contributing to their formation can support
preventive measures [1]. Furthermore, the analysis of crack patterns
provides noninvasive means of identifying the structural
components of paintings [4].
Visually, cracks can be categorized into bright cracks on a dark
background or dark cracks on a bright background. One can further
distinguish between different types of cracks such as ageing
cracks, premature cracking (generally due to drying defects
related to the painting materials or their application), or cracks
formed only in the varnish layer when it becomes brittle through
oxidation. The literature discusses mainly dark cracks; they are
typically considered as having low luminance and being local (gray
scale) intensity minima with elongated structure [14]. Different
crack-detection techniques include simple thresholding, line
detectors, and various morphological filters (see [1] for an overview).
The method in [7] operates on a single image modality (visible
image) and combines by means of a voting scheme three-crack
detection techniques: oriented elongated filters, a multiscale
extension of the morphological top-hat transformation, and a
detection method based on dictionary learning [13].