Measuring the Entropy of Sinan’s Muqarnas Patterns

This paper presents a novel computational model to measure the entropy value of 2-dimensional geometric patterns. The model takes an image file as input, it automatically identifies shapes and calculates entropy values for the given patterns based on the adjacency relationships of the detected shape tokens and the repetition frequencies of the shapes in the whole. The model utilizes Shannon’s definition of entropy to calculate shape entropy (SE), horizontal relationship entropy (HRE), and vertical relationship entropy (VRE) of 2-dimensional muqarnas patterns. In this study, the proposed model was tested in the muqarnas patterns in the portal of 12 mosques by architect Sinan which were built in the Classical Ottoman Period (sixteenth century). Initial findings of the implementation of the model show that the proposed automatized calculations are applicable to the selected sample group and have the potential to be adapted into different contexts.