Statistical feature based craquelure classification

Abstract

We demonstrate an automatic procedure for extracting features such as directionality of crack patterns, distribution of island sizes, and crack segment lengths to aid in automatic classification of painting cracks or craquelures. To test our classifier, we make use of four distinct craquelure patterns, designated by names based on its country of origin, namely: Dutch, Flemish, French or Italian. We report that using the above statistical measures, linear discriminant analysis can classify the craquelure patterns with an 85% accuracy. This accuracy is three times better than mere chance classification, or the proportional chance criterion, computed to be 26%. The article hopes to serve as basis in which the appropriateness of models for craquelure formation can be tested.