Urban tree detection using mobile laser scanning data

Abstract

Nearly half of the world's population (47 per cent) lives in urban areas and is expected to grow by two per cent per year during 2000-15 (UN, 2001). Urban trees are essential data sets for studies on urban biomass, ecological function, water budgets and radiation transfer in an urban system. Furthermore, it is also useful for 3D cadastre and 3D city models used by planners and environmentalists for planning, modelling and ecological assessments of the city. Traditionally, optical remote sensing methods have been used for acquisition of urban data sets. These data sets have been extensively used for tree identification, 3D modelling, and biomass estimation. An advantage of these methods is that the data is being collected for a large area in a single attempt. Whereas the disadvantages of these methods include comparatively coarse resolutions and some shadow effect on all the optical images. Therefore, it is complicated to identify the single tree pixel and also two-dimensional measurements is not sufficient to determine canopy structure. Nowadays, laser scanning is a well-established technology to acquire the three dimensional data. Several experiments have been carried to detect the tree and estimation of tree geometry parameters such as canopy height, crown width and foliage for the forest monitoring applications and modelling of 3D tree using airborne as well as terrestrial laser scanning (TLS) data (Bucksch and Lindenbergh, 2008; Mallet and Bretar, 2009; Rosell et al., 2009; Vauhkonen et al., 2009).