Data mining approach to image feature extraction in old painting restoration
2013
scientific article
english
- data mining application
- image processing
- k-means clustering
- decision tree based image segmentation
- virtual restoration of paintings
EN In this paper a new approach to image segmentation was discussed. Amodel based on a data mining algorithm set on a pixel level of an image was intro-duced and implemented to solve the task of identification of craquelure and retouchtraces in digital images of artworks. Both craquelure and retouch identification areimportant steps in art restoration process. Since the main goal is to classify and un-derstand the cause of damage, as well as to forecast its further enlargement, a propertool for a precise detection of the damaged area is needed. However, the complexnature of the pattern is a reason why a simple, universal detection algorithm is notalways possible to be implemented. Algorithms presented in this work apply miningstructures which depend of expandable set of attributes forming a feature vector, andthus offer an elastic structure for analysis. The result obtained by our method incraquelure segmentation was improved comparing to the results achieved by mathe-matical morphology methods, which was confirmed by a qualitative analysis.
159 - 174
CC BY-NC-ND (attribution - noncommercial - no derivatives)
public
15