Semilandmarks: A method for quantifying curves and surfaces

Quantitative shape analysis using geometric morphometries is based on the statistical analysis of landmark coordinates. Many structures, however, cannot be quantified using traditional landmarks. Semilandmarks make it possible to quantify two or three-dimensional homologous curves and surfaces, and analyse them together with traditional landmarks. Here we first introduce the concept of sliding semilandmarks and discuss applications and limitations of this method. In a second part we show how the sliding semilandmark algorithm can be used to estimate missing data in incomplete specimens.