Refining the computational method for the evaluation of visual complexity in architectural images: significant lines in the early architecture of Le Corbusier

Past research over the last two decades has demonstrated that fractal analytical methods can be used to evaluate the visual complexity of architectural designs. Applying a computer program that has been developed specifically for the quantitative evaluation of architecture, the authors previously tested the fractal analysis method on several well-known architects’ works including those of Le Corbusier. However, while undertaking this research the computational method revealed unexpected results for the visual complexity of Le Corbusier’s early, highly ornamented, house designs. This paper examines these anomalous results and the proposition that they are a by-product of the data chosen for analysis. The fractal method relies on architectural drawings as raw “data” but little research has been undertaken into which elements should be chosen to represent the buildings or alternatively, which lines are “significant”. The present paper uses the results for ten of Le Corbusier’s house designs to examine the sensitivity of the method.