Comparison of length-intensity estimators for segment processes

Several unbiased length-intensity estimators for stationary segment processes in the dd-dimensional Euclidean space are considered. The variances of the estimators are compared. The asymptotic behaviour is investigated if the sampling window increases unboundedly in all directions. The segments are assumed to be independent, identically distributed and independent of the locations. Special attention is devoted to the particular case of stationary Poisson segment processes. For the planar case and rectangular sampling windows, Neyman–Scott segment processes are studied in more detail.