0% found this document useful (0 votes)
65 views20 pages

Work Flow: Mapping and Aerial Image Processing Software

This document provides instructions for using mapping and aerial image processing software. It outlines the key steps: 1) loading images, 2) importing telemetry data, 3) bundle adjustment to align images, 4) orientation using ground control points to georeference the data, 5) 3D reconstruction, and 6) generating outputs like digital surface models, orthophotos and volume calculations. The process involves aligning images, manually marking ground control points, checking accuracy and adjusting points as needed to georeference the aerial imagery.

Uploaded by

Ilyas Aissi
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
Download as pdf or txt
0% found this document useful (0 votes)
65 views20 pages

Work Flow: Mapping and Aerial Image Processing Software

This document provides instructions for using mapping and aerial image processing software. It outlines the key steps: 1) loading images, 2) importing telemetry data, 3) bundle adjustment to align images, 4) orientation using ground control points to georeference the data, 5) 3D reconstruction, and 6) generating outputs like digital surface models, orthophotos and volume calculations. The process involves aligning images, manually marking ground control points, checking accuracy and adjusting points as needed to georeference the aerial imagery.

Uploaded by

Ilyas Aissi
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
Download as pdf or txt
Download as pdf or txt
You are on page 1/ 20

 

Mapping and Aerial Image Processing Software 

WORK FLOW 
 
 
 

Version 2.9.0 

   
TABLE OF CONTENTS 

1. Load images 3 
2. TELEMETRY 3 
3. BUNDLE ADJUSTMENT 4 
4. Orientation (geo-referencing) 5 
5. Reconstruction 11 

5. CLASSIFICATION 12 

6. Digital surface model (DSM) manipulation   13 


7. Digital Orthophoto 14 
8. Draw volume boundary in cad 17 

8.1 Preparing layers 17 

8.2 Draw line 17 


9. Volume calculation 19 

9.1 Volume calculation based on one surface 19 

9.2 Volume calculation based on two consecutive measurements 20 


   
1. LOAD IMAGES 

Start  3Dsurvey,  select  ​New  project  ​button  and  select  project  images  in your folder. Same can be done 
through ​Menubar’s​ I​ mages​ - > L​ oad images​. 

 
When images are loaded, ​Telemetry importer​ window pops up.  

2. TELEMETRY 
 
Telemetry  data  is  attribute  data  of  your  images.  This  data  includes  information  about  the  drone’s  (or 
camera’s)  position  at the time of recording of images.  Adding telemetry data at this stage significantly 
speeds up the process of Bundle Adjustment/Image Alignment. 

 
 
For DJI, Yuneec, eBee and falcon drones drones use E​ xif data​ for Telemetry import.  
For drones with the separate log file use one of following options: 
- C-astral Bramor -> *.csv 
- MaVinci Sirius Pro -> *.csvExif data (DJI Phantom 4, Inspire 1, eBee, Falcon 8,..) 
- Other -> *.csv 
For Horizontal coordinate system use the coordinate system that you want your results to be in. 
For Vertical coordiante system -> If using non-RTK drone - Use EGM96 Global 15-minute Geoid 
If using a RTK drone - Use your appropriate regional Geoid model 

Telemetry  window  pops  up,  presenting  imported  data  and  its  transformation  to  project  coordinate 
system with their accuracy. Click c​ lose​.  
 

3. BUNDLE ADJUSTMENT 
Press B
​ undle Adjustment​ to start processing the dataset. 
 

Select  ​Global  mode  of  adjustment  and  click  ​OK  to start Bundle adjustment. The processing time could 


be  between  a  couple of minutes and up to several hours. It depends on the number of images and your 
CPU specifications (number of cores,…) 

Once  project is bundle adjusted, you can observe several function in the viewer. You can watch ​images​, 
or  click  ​telemetry​,  to  observe  location  and  orientation  (external orientation) of images. ​Toggle text​, to 
view or not image names. 

 
 

4. ORIENTATION (GEO-REFERENCING) 
To geo-reference a project, proceed with O
​ rientation​. Three options are available depending on your data 
type. 
 

To  proceed,  click  ​Orientate  with  GCP  ​and  follow  the  Orientation  pop-up  wizard:  First,  ​select  ​ground 
control points​ (*.txt) file and click ​next​.  

Further  select  GCP’s  coordinate  system  and  also project coordinate system in which computed results 


and data will be transformed. Select L​ ocal coordinate system​. 

 
 
 
 
 
On  the  left  of  the  pop-up  window,  ​select  image  with  GCP.  Scroll  mouse  to  zoom  in  or  out.  For 
automatic  orientation  a  minimum  of  3  GCPs  need  to  be  manually  marked.  Mark  GCP  with  a  ​right 
mouse click​. Marked position is coloured in orange and ascribed a ‘’?’’ as its name is not defined.  

 
Continue  by  ​selecting  additional  two  GCPs  from other segments of your project area. When you select 
the  third  ground  control  point  the  automatic  recognition  for  all  other  GCPs  will  start.  Three  marked 
GCPs  remain  orange  coloured  while  automatically  found  GCPs are now coloured green and the names 
of all GCPs are identified. Click C
​ onfirm unknown marks​ beneath minimap. 

 
 
 
 
 
 
Click  on  the  first  image  from  the  list  of  images  and  check  if  circle  is  in  top  of  gcp.  The  GCPs,  that 
weren’t selected in the process of automatic orientation are marked with green circle.  

 
 
If  the  green  circle  is  not  positioned  over  the  GCP,  select  the  gcp  on  the  minimap  and click on it. Then 
click  GCP’s  correct  position  on  image.  Color  of  its  mark  will  turn  blue.  When  you  are  satisfied,  click 
Next image, and check gcps on all other images. 
 

 
 
In  the  lower right corner appears the ​residuals of marked points. Residuals should be within 1m. Check 
if  the  automatic  positions  are  calculated  correctly  -  GCPs  need  to  be  coloured  green.  If  not,  correct 
them  with  manual  selection.  First,  select  the  name  of  GCP  on  GCP  map, then show the exact position 
of GCP map on an image. Use left mouse click. Click ​next​ to c​ ontinue with orientation​.  
 

In  the  next  step,  we  can  see  automatic  recognition  of  ground  control  points  (in  case  of  3Dsurvey 
targets).  If the GCP is not recognised on a specific image, it is eliminated and coloured red. If you want 
to  include  this  GCP  into  orientation,  just  double  click  the  image  and  the  red  overlay  will  disappear. 
Several  images  can  be  marked  or  unmarked  by  ticking  them  in  their  upper  left corner, where unticked 
square appears. Use Ctrl + a shortcut to select all target windows at once. You can also correct/realign 
the  automatic  measurement  -  put  the  mouse  in  the  centre  of  the  target  and  right  mouse  click  for 
marking  the  center  of  GCP.  Use  mouse  wheel  or  slide  bar below to zoom in or out, which works for all 
target  windows  to  realign  easier..  Same  approach  can  be  used  to  correct  the  positions  of  GCPs  that 
have  been  automatically  detected  but  not  measured  perfectly  because  of  the  bad  conditions  during 
data acquisition.  

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
If  you  do  a  lot  of  manual  centering,  back  up  your  work  by  ​exporting  GCP  image  positions  ​of  centred 
targets.  Click  ​tooth  wheel  button in the upper right corner and click ​export​. In case if you calculate the 
same  project  again  next  time  you  don’t  need  to select and correct the GCP targets once again but just 
import the saved file and go to the next step. 

 
 
Click ​next​, to observe orientation’s accuracy. 
 
Orientation  summary  displays  GCP  errors  –  difference,  between  where  we  measured  GCPs  with  Total 
Station/GPS,  and  where  these  points  are  in  our  photogrammetric  model..  ​Check  ​GCPs  accuracy  by 
observing  their  ​residuals​.  If  residuals  show  high  displacements  of  certain  GCPs  (over  4  cm),  go  and 
recheck  them.  Click  ​back  button  on  the  upper  left  corner and repeat the marking process. Click ​Finish 
when satisfied. 

 
Save  project.  Click  ​Project  ->  Save  AS​,  and  select  the  name  of  project.  Its  path  is  in  the  same  file  as 
imported images. 

5. RECONSTRUCTION 
When  you are finished with ​orientation wizard the main window will appear. Next step is Reconstruction 
of  dense  point  cloud.  On  image  you  can  find  the  default  values  for  Reconstruction  step which offers 
you the most effective calculation (time VS results).  

 
Click OK to start reconstruction. 
 
Continuing  each  reconstruction  parameter  is  described  for  more  information.  ​Reconstruction  level  – 
there are 4 levels of reconstruction:  
- Low, 
- Medium, 
- High, 
- Extreme. 

By default we use Medium, when main result will be ortho map, and High when we are interested in 3D 
data like volumes or 3D map. 

 
 

5. CLASSIFICATION  
After pointcloud calculation finishes, it is important to first classify ground points.

Click ​Classification​ button to classify terrain points. In the pop up window, you can adjust:  
- terrain elevation –maximal sudden elevation change in terrain 
- terrain slope –maximal angle between horizon and slope in degrees,  
- minimal height – minimal height of objects to be deleted,  
- DSM cell size - cell size of previously computed DSM, 
- smoothing factor – proportionate to smoothness of classified terrain 
- smoothing maximum size – maximal size of influential area. 

Use mouses scroll and drag, to zoom to desired area. Left mouse click on desired area, to select it. 
Selected areas appear coloured. Not all terrain areas need to be selected. Click n​ ext​. 
 

If all preferable areas are selected, click ​Finish​. Otherwise, select the missing area and click ​Next​. 
Click ​Finish​ to complete classification.  

Click ​Delete other​, to delete non-terrain point cloud points. 

6. Digital surface model (DSM) manipulation 


Select from dropdown list D
​ SM​ ​-> C
​ alculate new​ and set parameters. 
 

For Orthophoto production the default parameters are: 


● Grid cell size: 0.5m 
● Hole filling mode: All 
● Smoothing size: 0 cells 
 
The lowest grid cell to be set is 20 cm and gives the highest detail. 

 
 

7. DIGITAL ORTHOPHOTO 
 

Select from dropdown list D


​ OF​ -> ​Calculate new​. 

Set  parameters  in  a  ​Digital  Orthophoto  Calculation pop-up. First, set ​Desired resolution. Click ​more to 


check  the  ​split  DOF  box.  Default  ​Tile  maximum  size  is  set  to  20.000,  meaning  DOF  will  be  split  if  it 
exceeds 20.000 pixels. When set, click C
​ alculate​. 
 

8. DRAW VOLUME BOUNDARY IN CAD 

This section describes the definition of area for volume calculations. 

8.1 Preparing layers 

Click ​+​ to make new layer. To change name, double click it. Press enter to confirm. Double click on 
colour square, to change the colour. Change line weight by double clicking on number. Type new 
weight. Those layers that are visible, have a scheme in the second row, as opposed to non visible 

ones . At the time, only one layer can be selected, in the second row is visible, as oppose to all 

the others . Third row shows whether layer is l​ ocked o


​ r​ unlocked . To change any of 
layers attributes, click on unwanted scheme in desired row of layer. 
 

8.2 Draw line 


To draw set the desired layer and select ​draw line t​ ool . Second, click on ​snapper settings​ to 
select type of snapping and m
​ easurements display​. If you want to snap directly on point cloud point, 
select p​ oint snapper​. This option supports also snapping on endpoints, midpoints and on or near 
existing lines.  

Start drawing lines, using l​ eft mouse click​.​ If you misclicked use ​backspace to​ delete last point.​ Abort 
drawing at any point with e​ scape​ button. End drawing polygon of lines using r​ ight mouse click​. 
 

To  close  the  polygon,  make  sure  to  have  Point  snpper  selected,  and  snapping  to  endpoints  enabled, 
and  then  just  click  on  the  first  starting  point  of  polygon.  You  can  change  ​snapper  settings  while 
drawing CAD objects. 

Click  Data  and  select  ​Save  CAD  as…​,  to  save  drawn  polygon.  Select preferable folder and type name. 
Data will be saved as dxf file. 

9. VOLUME CALCULATION 
 

9.1 Volume calculation from one surface 


Calculate  the  volume  of  material  for  single  measurement  -  appropriate  for the calculation of dredged 
material. 

To  calculate  the volume on the area drawn with CAD function, click ​Calculate volume​. Click ​From DXF. 


An I​ mport DXF file​ window appears. Select path and import desired *.dxf file. 
 

Name  of  dxf  file  is  appointed  to  calculated  volume.  Displayed  results present values for both current 
volume selection and a total sum of all volumes, consisting of: 
- Fill – volume of needed mass to fill holes beneath mean surface. 
- Cut – volume of mass above mean surface 
- Total – sum of filled and cut volume 
- Are 3D – three-dimensional area of selected surface 
- Area 2D – horizontal area of selected surface  

Select  ​Volume  from  the  ​Working  panel ​to observe all calculated volumes. Areas of volume calculation 


are  presented  as  cuts  from  ​DSM,  ​each with its own title. If ​Contour lines are visible, their cuts are also 
presented. You can s​ elect​, deselect and r​ emove selected​ volumes. 

 
 
9.2 Volume calculation based on two consecutive measurements 
Ideal  for  monitoring  on  construction  site.  Measure  the  volume  difference  between  two  consecutive 
measurements. 

You need to have 2, or more point clouds. Click D


​ SM -> Calculate new​ and set parameters as before. 

Click ok, to compute DSMs. Select ​Calculate volume​ and further F​ rom DXF​, then ​open​ dxf file. 

Name  of  dxf  file  is  appointed  to  calculated  volume.  Selected  areas  of DSMs are coloured in different 
colours, so you can observe, height differences. 
Presented results contain: 
- Fill – volume of added masses 
- Cut – volume of removed masses 
- Total – sum of filled and cut volume 
- Area 3D – three-dimensional area of selected surfaces 
- Area 2D – horizontal area of selected surfaces. 

  

In the Working panel, go to Volume, to observe differences in volume. 

You might also like