A couple questions for you. I’m curious how your tool compares to the existing Cesium Terrain Builder conversion utility ( https://github.com/geo-data/cesium-terrain-builder ). For reference, CTB is written in C++, supports multi-threading, and uses the GDAL geodata library to be able to read a variety of input formats.
What geodata file formats does your tool support? GeoTIFF, DTED, … ?
Does your tool do multi-threading? Do you have some rough benchmarks on performance and time needed to generate datasets?
What additional dependencies, if any, does your tool require?
All GDAL formats are accepted, however currently not tested.
Currently there is multi processing support but it is also in development. I can not give a benchmark since trying to make algorithm robust atm. However, we can say it is pretty slow at the moment.
Detecting bugs in tile consistency, and fixing them. But it is almost done.
I am planning to publish new quantized datasets this week if all goes well
All GDAL formats are accepted, however currently not tested.
Currently there is multi processing support but it is also in development. I can not give a benchmark since trying to make algorithm robust atm. However, we can say it is pretty slow at the moment.
Detecting bugs in tile consistency, and fixing them. But it is almost done.
I am planning to publish new quantized datasets this week if all goes well
9 Ağustos 2017 Çarşamba 11:45:39 UTC+3 tarihinde Muhammed Kalkan yazdı:
I was on a break, will continue next month.
8 Ağustos 2017 Salı 09:38:21 UTC+3 tarihinde soniu...@gmail.com yazdı:在 2017年2月11日星期六 UTC+8下午4:25:13,Muhammed Kalkan写道:
> Hi,
> Congratulations for your project,
I'm working on same lesson at university.I have to find a solution for finding route in terrain using quantized mesh.But ,i don't know how can i integreted this two theme? Thanks for texts...
I’m currently working on a (node.js) tool Here
that re-draw input DEM with height values sampled from “STK-World” terrain data, and I was wondering if your source-code is available now , i t might help me fix the slowness in performance for my tool.
I have focused more on terrain generation and serving thus i am afraid i cant be of any help about that.
One idea is to make your algorithm to work with regular heightmaps and move on to the quantized format. Since it has regular distances, you can proceed in your algorithm more easily.
Route finding is a kind of constrained optimization problem and can be applied any kind of sampled data.
@Mussab
Code is not available but i believe it wouldnt be any use since sampling which is what you are trying to do is much more calculation free than generation part. Generation and re-sampling focus completely different parts of the data.