Estimating the geometric features of a 3D vascular structure

Methods and algorithms for estimating the geometric features of 2D and 3D tree-like structures are proposed. The methods have major application areas in biomedical problems associated with the analysis and measurement of vascular system peculiarities: 2D eye-retina structure and 3D cardiovascular system. The immunity of the estimation methods to different types of noise and the feasibility of the feature-based clustering of vessel samplings are experimentally studied.