Abstract
With the increasing local market demands of LIDAR technology and equipment,
the post processing software of LIDAR point cloud is always used corresponding
commercial software. The surveying and mapping products of these softwares are
incomplete in according with the corresponding criterion of China. Not only on the
respect of drawing efficiency but also on the respect of drawing ideas, the processing
of LIDAR point cloud is very different from the conventional measuring data.
Therefore, combining with commercial point cloud processing software and domestic
surveying and mapping criterions and developing a point cloud data processing
software that is suitable for domestic surveying and mapping products is a totally
necessary and urgent.
The researched processing software, PCEdit, is based on the rich information of
point cloud and Digital Orthophoto Map (DOM), used AutoCAD as development
platform and objectARX as development tool in virtue of visual C++8.0. Combining
point cloud with DOM, we can get lots of surveying and mapping products, such as
Digital Line Graphs (DLG), Digital Raster Graphic (DRG), Digital surface model
(DSM) and so on. This paper is mainly focuses on the key points as follows:
The input and display of point cloud is the first step of data processing.
Considering the low efficiency of general point cloud input and edit, this paper
derives a custom point entity using Object DBX. In this way, each point cloud file can
be imported into database as one entity along with one handle, which makes the
browsing and roaming of this file more convenient.
In filtering, interpolation and compression of point cloud, it is very frequent to
accessing some points, such as searching, insertion and deleting. In order to
improving the accessing efficiency, the idea of dynamic window is put forward. By
pulling the vertex of the minimum rectangle (cuboid) of poind cloud, we can find the
interesting area to edit, which can avoid accessing the whole point cloud data but
focus only on the certain point cloud data in rectangle (cuboid). Therefore, a high
efficiency in editing and displaying the point cloud can be acquired.
Since different densities of point cloud are needed in different objects, the idea of
sampling, erase and interpolating point cloud based on closed polygon and strip area
is necessary, which accelerate the speed of producing surveying and mapping.