One step clustering based on a-contrario framework for detection of alterations in historical violins

Analysis of Multi-temporal Image Series for the Preventive Conservation of Varnished Wooden Surfaces

“…Attempts have been made to detect areas with established wear, that had peculiar chromatic characteristics [13]. More recently, Rezaei et al [36,37] proposed a method for detecting changed pixels between two multi-modal images and then grouping them together, while Dondi et al [14] presented a segmentation approach based on histogram quantization and genetic algorithm. However, all these works study the evolution of alterations using no more than two frames at a time.…”

“…In the field of preventive conservation, a-contrario framework has been used for comparison of two single images to detect candidate points (seed detection) and then group them together (clustering) [36]. Later, Rezaei et al [37] introduced a noise model to detect in one step very unlikely (under noise model assumption) arrangements of pixels based both on their location and their gray-level value. They assume the graylevel values to be a third axis and introduce a custom 3D distance to relate all three axes together.…”

“…Let {I t } n t=0 be a series of n multi-temporal RGB images taken from a varnished wooden sample; I 0 being the reference image. The size of each I t is w × h. We perform the change detection in each I t (t > 0) image with respect to I 0 using the method described in [37]. This results in a set of binary images denoted {B t } n t=1 .…”

“…We specifically work with varnished wooden surfaces subjected to mechanical wear over time. Our proposition is based on the a-contrario framework [10,9] and it improves the method presented in [37] which deals with detection of changes between two single frames: a reference image and an up-to-date acquisition. In this study, we extend the detection framework in order to process a temporal sequence of changes and to account thus for their significance with respect to the spatial plus time dimensions.…”

Structural feature investigation of wooden artifacts through imaging techniques: a step forward in the preservation of historical musical instruments

“…Over the centuries, some of these masterpieces changed many owners and have been intensively played. Therefore, the intense concert activity together with the natural ageing of the materials [1] could be responsible for the presence of mechanical stresses and physical-chemical degradations of the varnish [2], [3] that may also induce cracks and structural deformations [4]. For these reasons, in most cases, these instruments underwent maintenance and restoration processes to remove heaps of dirt, replace the altered or worn out varnishes [5], [6], and to repair cracks and other defects [7].…”