GIS/MEA 584: Mapping and Analytics Using UAS

GIS based analysis of UAS derived data processing outputs

In this assignment we analyze the DSMs derived from the UAS imagery in terms of their accuracy and type of distortions based on the use of GCPs and different processing software. Then we map the terrain change (due to vegetation growth, erosion, and other impacts using two different uas surveys).

Including GCPs in geoprocessing improves the results. Does it have an effect on phenomenon that you are investigating?
Agisoft is only one example of the software solutions available for the UAS data processing. In the GRASS location there are available DSM generated in different environments: Agisoft, Pix4D and Trimble Business Center (just for one example flight). Compare the differences between these DSMs.
In Imagery processing assignment you’ve used arbitrary given parameter values in each processing step. You can change various modes, quality, values and options and investigate if it does improve the results. You can find descriptions of the options for each processing step on lecture slides and in the Agisoft Manual.

Data

Lake Wheeler data set (formatted as GRASS location): Lake_Wheeler_NCspm (you should already have it from previous assignments)

DSMs comparison - influence of GCPs

In order to detect the bowl effect, additional DSM needs to be computed. This time GCPs should not be used in the processing. You can repeat all the steps in Agisoft from Imagery processing assignment Batch processing commands should be in this order:

Align Photos
Optimize Alignment (this is only based on camera coordinates)
Build Dense Point cloud (quality: medium)
Generate DEM

Now you should have two DSMs:

Packs are provided for those working in DEMO mode, save them in your working directory.

Calculate differences between processed sample data

We will use GRASS GIS to compare the DSMs.
Open GRASS with the "Lake_Wheeler_NCspm" location and create a new mapset for this assignment (you can call it "analysis"). Then change your working directory
Settings > GRASS working environment > Change working directory to where your DSMs files are.

r.unpack DSMagi_2016_09_22_sample_noGCPs.pack -o
r.unpack DSMagi_2016_09_22_sample.pack -o

* If the code doesn't run. Type r.unpack in the console and click enter twice. A dialog window will pop up, where you can indicate the location of your pack by pressing the Browse button.

Calculate the differences in elevation in the DSM generated with GCPs and the new one without GCPs in GRASS GIS using raster map algebra:

g.region rast=DSMagi_2016_09_22_sample
r.mapcalc expression="GCP_noGCP_class = DSMagi_2016_09_22_sample - DSMagi_2016_09_22_sample_noGCPs"

Change the color map to better see the differences. You can experiment with your own color table or enter following values directly in the dialog r.colors dialog (second tab Define)

0 red
35 orange
36 yellow
37 cyan
38 aqua
39 blue

Calculate differences between DSMs from larger area

Your sample data has very limited extend. To fully understand the phenomenon that is cause by the absence of GCPs, compare the sample DSMs from the Lake_Wheeler_NCspm_course location.
There were 4 GCPs used for generating sample_DSM. You can see localization by adding the GCP_12 vector layer to the display.

g.region rast=sample_DSM
r.mapcalc expression="GCP_noGCP = sample_DSM - sample_DSM_noGCPs"

Change the color map to better see the differences. You can experiment with your own color table or enter following values directly in the dialog r.colors dialog (second tab Define)

0 red
35 orange
36 yellow
37 cyan
38 aqua
39 blue

Why are these patterns different? Consider the distribution of the GCPs in both datasets and the shape of the area.

Compare DSMs generated in different software

For the flight from 2015-06-20, in the Lake_Wheeler_NCspm_course there are DSM with GCPs generated in 3 different software environments (Trimble Business Center, Agisoft and Pix4D). Additionally there are DSMs generated without GCPs in Agisoft and Pix4D.

Check the bowl effect for the Agisoft generated products:

g.region rast=2015_06_20_DSM_agi_11GCP
r.mapcalc expression="agi_GCP_agi_noGCP = 2015_06_20_DSM_agi_11GCP - 2015_06_20_DSM_agi_noGCP"

Change the color table to better see the differences.

0 red
35 orange
36 yellow
37 cyan
38 aqua
42 blue

Check the bowl effect for the Pix4D generated products:

r.mapcalc expression="p4d_GCP_p4d_noGCP = 2015_06_20_pix4d_11GCP_dsm - 2015_06_20_DSM_pix4d_NoGCP"

Change the color map to better see the differences. You can experiment with your own color table or enter following values directly in the dialog r.colors dialog (second tab Define)

50 red
69 orange
70 yellow
71 cyan
72 aqua
75 blue

Which software does generate larger bowl effect?

r.mapcalc expression="agi_trimble = 2015_06_20_DSM_agi_11GCP - 2015_06_20_DSM_Trimble_11GCP"
r.mapcalc expression="p4d_agi = 2015_06_20_pix4d_11GCP_dsm - 2015_06_20_DSM_agi_11GCP"
r.mapcalc expression="p4d_trimble = 2015_06_20_pix4d_11GCP_dsm - 2015_06_20_DSM_Trimble_11GCP"

Apply the color table that varies from -1m to 1m to visualize results of the comparison using r.colors module. In the Define tab, type the following rules in the enter values directly option.

 -40 red
 -1 orange
 -0.5 yellow
 -0.1 grey
 0 white
 0.1 grey
 0.5 cyan
 1 aqua
 35 blue

Compare the maps. How do they relate to each other? What patterns (artifacts) can you recognize?

Detect potential processing artifacts using terrain analysis

It is also useful to a perform terrain analysis using GIS tools to see the differences even more clearly.

Run the r.slope.aspect and r.relief command for chosen DSMs, for example:

g.region rast=sample_DSM
r.slope.aspect elevation=sample_DSM slope=sample_DSM_slope aspect=sample_DSM_aspect pcurvature=sample_DSM_curv
r.relief input=sample_DSM output=sample_DSM_relief
r.slope.aspect elevation=sample_DSM_noGCPs slope=sample_DSM_noGCP_slope aspect=sample_DSM_noGCP_aspect pcurvature=sample_DSM_noGCP_curv
r.relief input=sample_DSM output=sample_DSM_noGCP_relief

You can also install GRASS GIS Addons (via g.extension):

and apply this modules to sample data or your data.

g.region rast=sample_DSM

g.extension extension=r.local.relief operation=add 
r.local.relief input=sample_DSM output=local_relief_sample_DSM
r.local.relief input=sample_DSM_noGCPs output=local_relief_sample_DSM_noGCPs

g.extension extension=r.shaded.pca operation=add 
r.shaded.pca input=sample_DSM output=shaded_pca_sample_DSM
r.shaded.pca input=sample_DSM_noGCPs output=shaded_pca_sample_DSM_noGCPs

g.extension extension=r.skyview operation=add
r.skyview input=sample_DSM output=skyview_sample_DSM
r.skyview input=sample_DSM_noGCPs output=skyview_sample_DSM_noGCPs

Detect terrain change using two UAS surveys

g.region rast=2015_06_20_DSM_agi_11GCP
r.mapcalc "diff_jun_oct_agis = 2015_06_20_DSM_agi_11GCP - 2015_10_06_DSM_agi_8GCPs"

Apply the color table that varies from -1m to 1m to visualize results of the comparison using r.colors module. In the Define tab, type the following rules in the enter values directly option.

 -40 red
 -1 orange
 -0.5 yellow
 -0.1 grey
 0 white
 0.1 grey
 0.5 cyan
 1 aqua
 35 blue

Note the accuracy of the DSM along the roads, interpret the observed gain and loss in elevation. Use orthophotos from Nov 2015, March 2017 and/or June 2016 to assist with interpretation (use r.unpack -o).