Knowledge Based Lacunas Detection and Segmentation for Ancient Paintings
Lacunas are a common form of the damage that can occur to paintings and more often to murals. Taking Dunhuang murals as research background, a new algorithm to detect and segment the lacuna area from mural images is proposed, which consists of a training
- PDF / 1,280,663 Bytes
- 11 Pages / 430 x 660 pts Page_size
- 32 Downloads / 180 Views
1 College of Computer Science and Technology, Zhejiang University, State Key Lab. of CAD and CG, Zhejiang University, 310027, Hangzhou, Zhejiang, China liujianming, [email protected]
Abstract. Lacunas are a common form of the damage that can occur to paintings and more often to murals. Taking Dunhuang murals as research background, a new algorithm to detect and segment the lacuna area from mural images is proposed, which consists of a training phase and a runtime phase. In the training phase, a Bayesian classifier is trained. At runtime, the Bayesian classifier is first applied to perform the rough lacuna regions detection. Then, a graph representing the mural image is built with output of the Bayesian classifier. The domain knowledge of murals is incorporated into the graph in this step. At last, the image segmentation using graph cut is done based on the minimal cut/maximal flow algorithm. The outputs of the image segmentation are lacuna regions and background regions. About 250 high resolution Dunhuang mural images are collected to test the proposed method's performance. Experimental results have demonstrated its validity under certain variations. This research has the potential to provide a computer aided tool for mural protectors to restore damage mural paintings. Keywords: Image segmentation, concurrent detection and segmentation, deterioration murals, mean shift, Bayesian classification, graph cuts.
1 Introduction Many ancient paintings, especially the murals, suffer from serious deterioration. For Dunhuang murals, there are more than fifteen kinds of deterioration, such as1 Cracks, Crater eruption, Flaking, Disruption and so on. Lacunas are a common form of the damage caused by most of the deterioration. To protect these priceless murals, we need to find out all the regions where these deteriorations are located. Therefore, detecting the lacunas and labeling the lacuna area in the murals are very important. Fig. 1 shows an example of the mural deterioration distribution map, which was labeled by Dunhuang mural protectors. With advanced computing and image analysis technique, it provides an opportunity to label the deteriorated area in an automatic 1
Cracks: fissures in the painting. Crater eruption: bulges leading to losses (from 2mm to 1cm in diameter) of the paint and plaster. Flaking: lack of adhesion between paint layer and ground. Disruption: De-cohesion affecting any or all layers composing the wall painting. Lacunas: regions which missing paint layer.
T.G. Wyeld, S. Kenderdine, and M. Docherty (Eds.): VSMM 2007, LNCS 4820, pp. 121–131, 2008. © Springer-Verlag Berlin Heidelberg 2008
122
J. Liu and D. Lu
way. In this paper, we present a knowledge based concurrent lacunas detection and segmentation algorithm for ancient paintings. We follow the idea given in [1] of supervised texture detection, which provided a small template of a texture of interest and get the image being segmented into regions with similar properties and background regions. However, the method [1] was designed for natural textur
Data Loading...
