Using thematic layers of a classified points cloud in visibility analysis in urban space

Abstract

The city as a space is made up of many different components: buildings of variable heights and shapes, the multitude of trees’ species and many other objects (bench, street lights, banners etc.). This makes it necessary to take into account the particularities of the preparation of digital surface model used in the visibility analyses. Consideration of presentation of high vegetation (trees) is a particularly important issue in such studies. Although the occurrence of a number of algorithms that allow for the detection of crowns and axes of trees, their presentation allowing for an exact reflection of the perception of reality still creates a big problem. Therefore an approach using ALS data was drawn up for the modelling of trees based on the detection of their axes by searching for local maxima on the high vegetation thematic layer. Thematic layers formed on the basis of a classified points cloud were used to propose and compare two variants of a digital surface model (DSM): the classical DSM presenting trees as crowns and the DSM showing trees as trunks (axes). They were created in order to verify whether the inclusion of non-uniform presentation of trees at a given vertical viewing angle affects the possibility to achieve results closer to the actual mode of human perception. Cracow Błonia was selected as the study area. The main objective of the analysis was to present the complexity of the optimization problem of data analysis in terms of visibility and to demonstrate the potential that lies in the ALS technology, taking into account the urban areas. The results proved the superiority of analysis performed with the use of a combination of ranges of visibility compared to the classical DSM without consideration of diversified nature of trees.