Authors: Peter L. Guth and Filipe Vieira
The rise of handheld smart phones and tablets brings sufficient storage capacity and processing capabilities to perform significant terrain analysis on the local device. This work can be done as either native apps, or using the capabilities of HTML5. Digital elevation models (DEMs) come in a wide variety of formats, and it does not make sense for handheld device have translators for all of them. The handhelds need a simple format that compresses well, but retains accurate geospatial registration information. We created a JSON format with bzip2 compression, both of which have widespread support with JavaScript libraries. This format allows data only on the WGS84/NAD83 datum, in either geographic or UTM coordinates. A command line converter, based on MICRODEM supplemented with the GDAL libraries, performs DEM reprojection if necessary, converts to JSON format, and compresses for download to the handheld device. Lidar point clouds are now supplied almost exclusively in the LAS format. While the LAS format has a number of options, all include far too many fields for use on a handheld—for instance the 8 bytes for GPS time will not help someone in the field seeking to visualize terrain. We chose a similar JSON/bzip2 solution for MICRODEM to prepare point clouds, and achieve lossy compression roughly comparable to lossless LAZ compression. A GeoJSON format specification exists, but only applies to vector data. We propose GeoJSON-G for grids and GeoJSON-P for point clouds, which will retain full geospatial registration information while intelligently limiting the volume of data for the handheld devices to download and process. The handhelds will not edit modify the data, making elimination of unnecessary fields acceptable. We have adapted HTML5 and Javascript open source projects to display both DEMs and lidar point clouds.
You are reviewing the
Download.
Your machine is locate at 3.128.255.103.
