Rasterizing Point-Cloud data

[AhmadObeid]

Hello everyone. 

I'm trying to figure out if there's any standard, go-to procedure that people in the field of remote sensing use in order to convert point-cloud data to a matrix.

I am a signal processing masters student, specializing in remote sensing and data fusion. Lately I've been going through the literature to see what people usually use, but authors do not, usually, explicitly mention what they do.

I have some Lidar point-cloud data, and the method I used was to simply interpolate the data so as to fill in all the gaps. As you can imagine, this would create a great source of variability as the quaity of the results might depend greatly on the quality of the interpolation.

What do you usually do/know that people do?

Any input would be greatly appreciated.

Ahmad

[jorrarro]

There is no "standard" workflow, as there are too many variables involved - source of data, quality of data, terrain differences, vegetation, building and of course the goals of analyses are also different.

This is what I have done recently:

1. Filter the dataset with LASTools - extract terrain, vegetation and buildings.

2. Convert each separate dataset to grid using SAGA. Cell size depends on area size and resolution.

3. Fill in the gaps using B-spline interpolation (SAGA). The specific parameters and masks used depend on the dataset. For example, terrain should have full coverage, but building need only cover their footprints (from OSM for example).

[AhmadObeid]

Thanks for the input jorrarro.

How would you rate such a process in terms of autonomy and quickness?

Do you think it could be used if you have a big data-set of different las files?