Abstract
This paper presents an innovative computer graphic method for viewshed generation from big point clouds. The proposed approach consists in simplification of typical methods for viewshed formation that are based on sorting and binary trees. The proposed method is based on the k-d tree concept optimized with radial segmentation and a dedicated mathematical algorithm for subtree rejection. The final visualization of the viewshed is designed with a graphic method using triangulated irregular network (TIN) surfaces from the accepted subtrees.
Highlights
The following figures show a comparison of times for generating one radial slice with one degree angle (Figure 12) and time for generating ninety slices based on one degree angle (Figure 13)
The presented charts show very accurately that for a single segment, the algorithm does not accelerate the generation of the slice. This is due to a large amount of time spent on generating a k-d tree
The determination of the angle between the point of observation and the corner resulting from the intersection of median lines in k-d tree was assumed as the leading operation
Summary
Emotions arising from the surrounding landscape greatly affect human wellbeing. More and more people would like to prevent the destruction of natural landscape and its monuments in an ill-conceived pursuit for financial profit. Information about natural resources is increasingly collected by virtue of a variety of terrain scanning methods. More and more data are created, which include point clouds. Due to the increasing technological possibilities, the obtained point clouds have an enhanced density, and that makes them more detailed [1,2]. The development of acquisition methods must be accompanied by the development of software for their rapid processing. The amount of data is so large that only a well-prepared environment enables their practical use
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have