10 cm Drone Orthophoto → 1 m RGB GeoTIFF → Raster Optimization → Civil 3D → DTM → 3D Drape
Software: ArcMap + Autodesk Civil 3D
Source imagery: 10 cm/pixel RGB orthophoto
Working imagery: 1 m/pixel RGB orthophoto
Coordinate System: Nepal Nagarkot TM 84
Final objective: Use the optimized orthophoto as a drape image over a Civil 3D DTM surface.
1. Introduction
High-resolution drone surveys provide extremely detailed orthophotos. For engineering projects, a 5 cm or 10 cm orthophoto can be very useful for detailed inspection, mapping and survey interpretation.
However, the highest possible resolution is not always the best choice for everyday CAD and 3D visualization. Large raster datasets can make Civil 3D slow when zooming, panning, querying and displaying the imagery together with large terrain surfaces.
During our Soti Drone Survey project, we therefore created a separate 1 m/pixel working orthophoto from the original 10 cm/pixel master orthophoto.
The original 10 cm orthophoto should always be preserved as the master survey dataset. The 1 m raster is a working/visualization copy and should not replace the original survey data.
2. Why Did We Convert 10 cm to 1 m?
The key reason is performance. A 10 cm pixel represents an area of only:
0.10 m × 0.10 m = 0.01 m²
Therefore, there can be approximately 100 pixels in every square metre. At 1 m resolution, there is only one pixel per square metre.
| Parameter | 10 cm Raster | 1 m Raster |
|---|---|---|
| Pixel size | 0.10 × 0.10 m | 1.00 × 1.00 m |
| Pixels per m² | 100 | 1 |
| Detail | Very high | Moderate |
| CAD/GIS performance | Heavy | Much lighter |
| Recommended use | Detailed survey | Engineering visualization |
3. Our Original Soti Orthophoto
The original Soti orthophoto used in this workflow had the following properties:
- Cell size: 0.1 m × 0.1 m
- Columns: 24,329
- Rows: 18,881
- Bands: 3 RGB bands
- Pixel depth: 8 bit
- Format: TIFF
- Uncompressed size: approximately 1.28 GB
- Coordinate system: Nepal Nagarkot TM 84
This is a very large raster for routine Civil 3D visualization.
Insert the screenshot showing the original raster Layer Source information here.
4. Step 1 — Open the Orthophoto in ArcMap
Open ArcMap and add the original drone orthophoto to the map. Before processing, check the raster properties carefully.
Verify:
- Cell size
- Number of bands
- Pixel depth
- Coordinate system
- Extent
- NoData value
5. Understanding Raster Bands
A standard RGB image normally contains three colour bands:
Band 2 → Green
Band 3 → Blue
Some drone products may contain a fourth band, normally related to transparency or alpha information. A fourth band is not necessarily wrong, but it is not required when our final objective is a normal RGB orthophoto for Civil 3D visualization.
6. Step 2 — Resample 10 cm to 1 m
In ArcMap open:
ArcToolbox → Data Management Tools → Raster → Raster Processing → Resample
Use the following settings:
| Setting | Value |
|---|---|
| Input Raster | Soti_MUTM84.tif |
| Output Cell Size X | 1 |
| Output Cell Size Y | 1 |
| Resampling Technique | BILINEAR |
Insert the screenshot showing Output Cell Size X/Y = 1 and Bilinear here.
7. Why Did We Choose Bilinear Resampling?
Bilinear resampling calculates the new pixel value from neighbouring pixels. For photographic RGB imagery this produces a smooth and visually natural result.
Therefore, for our drone orthophoto:
RGB Orthophoto → Bilinear
For categorical data such as land-use classes, nearest neighbour is generally more appropriate. The correct resampling method depends on the type of raster.
8. Step 3 — Configure Raster Storage
Before running the raster processing operation, open:
Environments → Raster Storage
Our working configuration was:
- Build pyramids: ON
- Pyramid resampling: Bilinear
- Pyramid compression: LZ77
- Raster compression: LZW
- Tile size: 256 × 256
- Raster resample: Bilinear
Insert the Raster Storage screenshot showing pyramids, LZW and 256 × 256 tile size.
9. What Are Raster Pyramids?
Raster pyramids are lower-resolution representations of the raster created mainly to improve display performance.
Conceptually:
Full Resolution
↓
Pyramid Level 1
↓
Pyramid Level 2
↓
Pyramid Level 3
↓
Lower Resolution Display
Pyramids do not replace or destroy the original full-resolution pixels. They simply allow GIS software to display large rasters more efficiently.
10. What Is LZW Compression?
LZW stands for Lempel-Ziv-Welch. It is a lossless raster compression method.
For engineering imagery, this is useful because we want to reduce file size without deliberately introducing the visual degradation associated with lossy compression.
LZW = Lossless Compression
11. Step 4 — Calculate Raster Statistics
We enabled:
Calculate Statistics
Our settings were:
- X skip factor = 4
- Y skip factor = 4
- Statistics ignore value = 0
Insert the screenshot showing Calculate Statistics, X=4, Y=4 and Ignore Value=0.
What Does Skip Factor 4 Mean?
The skip factor tells ArcMap that it does not need to inspect every single pixel while calculating statistics. Instead, it samples the raster at intervals.
With:
X Skip = 4 Y Skip = 4
ArcMap can calculate representative statistics much faster than analysing every pixel in a very large raster.
What Does "Ignore Value = 0" Mean?
Our orthophoto contains areas outside the actual survey footprint. Those background pixels can contain zero values.
We therefore told ArcMap:
Do not use pixel value 0 when calculating statistics.
This does not remove zero pixels from the raster. It only excludes them from the statistics calculation.
12. Step 5 — Check the Resampled Raster
After Resample completed, we checked the resulting raster properties. The 1 m raster had approximately:
- Columns: 2,433
- Rows: 1,889
- Cell size: 1 m × 1 m
- Pixel depth: 8 bit
- Coordinate system: Nepal Nagarkot TM 84
Insert the screenshot showing 2,433 × 1,889, 1 m cell size and raster properties.
13. Step 6 — Check the Number of Bands
The resampled raster initially contained four bands. We inspected the statistics of each band.
Band 4 showed:
Minimum = 255 Maximum = 255 Mean = 255 Standard Deviation = 0
This tells us that Band 4 contained only a constant value. It was therefore not carrying useful colour information for our RGB visualization.
14. Step 7 — Create a Clean 3-Band RGB Raster
We created a final working raster containing only:
Band 1 → Red Band 2 → Green Band 3 → Blue
In ArcMap this can be achieved using:
Composite Bands
The important point is to select only Bands 1, 2 and 3.
Insert the Composite Bands screenshot here.
15. Make Raster Layer vs Composite Bands
During this workflow we learned an important distinction.
| Tool | What it does |
|---|---|
| Make Raster Layer | Creates a raster layer/reference for use in the map. It does not necessarily create a new physical TIFF. |
| Resample | Creates a raster at a different cell size. |
| Composite Bands | Creates a raster by combining selected raster bands. |
16. Final 1 m RGB Raster
Our final raster was:
- Name: Soti_MUTM84_C3D_1m_RGB.tif
- Columns: 2,433
- Rows: 1,889
- Bands: 3
- Cell size: 1 × 1 m
- Pixel depth: 8 bit
- Format: TIFF
- Compression: LZW
- Pyramids: Built
- Coordinate system: Nepal Nagarkot TM 84
FIGURE 07 — Final 1 m RGB TIFF properties
Insert the final Layer Source screenshot showing 3 bands and 1 m cell size.
17. Step 8 — Connect the Raster to Civil 3D
After preparing the optimized raster, the next task was to bring it into Civil 3D.
Civil 3D provides more than one way to work with raster imagery. During our testing, we used the Data Connect method.
The workflow was:
Civil 3D ↓ Data Connect ↓ Add Raster Image or Surface Connection ↓ Select Soti_MUTM84_C3D_1m_RGB.tif ↓ Connect ↓ Add Data to Map
Insert the Data Connect screenshot showing the 1 m RGB TIFF selected.
18. Data Connect vs Map Image Insert
We tested both approaches and found an important practical difference.
Data Connect
Data Connect treats the raster as an external data source. It is particularly useful when working with GIS datasets and large raster files.
Map Image Insert
Map Image Insert is more directly related to inserting a raster image into the drawing/map environment. It can be convenient for individual images and certain CAD workflows.
For our large engineering raster workflow, Data Connect performed smoothly after the raster had been optimized to 1 m.
For large georeferenced engineering rasters, preparing an optimized working raster first and then using Civil 3D Data Connect gave us a smooth working experience.
19. Step 9 — Use the DTM as the Elevation Base
The RGB orthophoto only provides colour information. It does not create the terrain elevation by itself.
For the 3D terrain, we used our previously created:
EG-SOTI-DTM
This surface provides the actual elevation geometry. The RGB orthophoto provides the photographic texture.
DTM
+
RGB Orthophoto
↓
3D Textured Terrain
20. Step 10 — Use DRAPEIMAGE
Civil 3D's DRAPEIMAGE command can be used to apply an image as a render material over a surface.
In our case we selected:
- Image: Soti_MUTM84_C3D_1m_RGB
- Surface: EG-SOTI-DTM
Insert the screenshot showing the RGB image and EG-SOTI-DTM selected.
Civil 3D automatically created a render material similar to:
EG-SOTI-DTM_Soti_MUTM84_C3D_1m_RGB_30123_RenderMaterial
21. An Important Problem We Encountered
At first, the DRAPEIMAGE command appeared to fail. The command line showed:
Command: DRAPEIMAGE Command: *Cancel*
We initially suspected that the drape operation itself was not working. However, we checked the result using Civil 3D's Object Viewer.
22. Object Viewer Revealed the Truth
The Object Viewer displayed the terrain correctly with the orthophoto draped over the DTM.
This was a very important troubleshooting discovery. It proved that:
- The DTM was correct.
- The RGB image was correct.
- The render material was created.
- The DrapeImage operation was actually working.
Insert the Object Viewer screenshot showing the textured 3D Soti terrain here.
23. The Real Problem — 2D Wireframe
The actual problem was the Civil 3D viewport visual style. The drawing was being displayed using:
2D Wireframe
The render material was therefore not being displayed correctly in the normal drawing viewport.
When we changed the viewport visual style to:
REALISTIC
the draped orthophoto appeared correctly on the 3D DTM.
If DRAPEIMAGE appears to fail, do not immediately recreate the surface or raster. First check the Civil 3D Visual Style.
2D Wireframe → Realistic
Insert the final Soti 3D terrain screenshot here.
24. Final Working Workflow
10 cm Drone Orthophoto
↓
Check Raster Properties
↓
Resample
↓
1 m × 1 m
↓
Bilinear Resampling
↓
Build Pyramids
↓
LZW Compression
↓
Calculate Statistics
↓
X/Y Skip = 4/4
↓
Ignore Value = 0
↓
Check Bands
↓
Composite Bands
↓
Keep RGB Bands 1,2,3
↓
Final 1 m RGB GeoTIFF
↓
Civil 3D Data Connect
↓
EG-SOTI-DTM
↓
DRAPEIMAGE
↓
Realistic Visual Style
↓
3D Textured Terrain
25. Troubleshooting Guide
| Problem | Likely Solution |
|---|---|
| Civil 3D is very slow when displaying the raster | Create a lower-resolution working raster such as 1 m and use pyramids. |
| Image appears flat | Check that the image is being used as a drape/render material over the DTM. |
| DRAPEIMAGE appears to cancel | Check the surface, render material and especially the Civil 3D visual style. |
| Object Viewer shows the drape correctly but Model Space does not | Change the viewport visual style from 2D Wireframe to Realistic. |
| Raster contains an unnecessary fourth band | Create a new 3-band RGB raster using Bands 1, 2 and 3. |
| Black/NoData area appears around the orthophoto | Handle the NoData/background separately. Do not confuse statistics ignore value with removal of the pixels. |
26. Recommended Engineering File Structure
Drone Survey
│
├── MASTER
│ └── Soti_MUTM84_10cm_RGB.tif
│
├── GIS_WORKING
│ └── Soti_MUTM84_C3D_1m_RGB.tif
│
├── DTM
│ └── Soti_DTM_MUTM84.tif
│
└── CIVIL_3D
└── Soti 3D Visualization
This approach keeps the original survey information safe while allowing us to create optimized datasets for different engineering applications.
27. 10 cm or 1 m — Which Should We Use?
| Application | Recommended |
|---|---|
| Detailed survey interpretation | 10 cm |
| Detailed orthophoto mapping | 10 cm |
| General GIS viewing | 1 m |
| Civil 3D visualization | 1 m |
| Large 3D terrain visualization | 1 m |
28. Final Engineering Lessons
Lesson 1 — More resolution is not always better
The best raster is the one appropriate for the task. A 10 cm raster may be unnecessarily heavy when the objective is general 3D visualization.
Lesson 2 — Never overwrite master survey data
Always create a separate working copy. The original high-resolution orthophoto remains the authoritative survey dataset.
Lesson 3 — Pyramids improve display performance
Large rasters benefit significantly from properly built pyramids.
Lesson 4 — LZW provides lossless compression
LZW is an excellent choice when we want to reduce storage requirements without intentionally degrading image quality.
Lesson 5 — Statistics are not image editing
Statistics help GIS software understand the raster's pixel-value distribution. They do not remove or modify the underlying pixels.
Lesson 6 — Object Viewer is an excellent diagnostic tool
If Object Viewer displays the draped terrain correctly, it is strong evidence that the DTM, image and render material are working. The problem may simply be the current viewport display settings.
Lesson 7 — Check Visual Style before rebuilding anything
Our most important Civil 3D troubleshooting lesson was:
DRAPEIMAGE working + 2D Wireframe = misleading result
DRAPEIMAGE working + Realistic = correct 3D visualization
29. Quick Reference — Recommended Settings
| Parameter | Recommended Setting |
|---|---|
| Source GSD | 10 cm |
| Working GSD | 1 m |
| Resampling | Bilinear |
| Final bands | RGB 1,2,3 |
| Pixel depth | 8 bit |
| Format | TIFF |
| Compression | LZW |
| Pyramids | Build |
| Pyramid Resampling | Bilinear |
| Statistics | Calculate |
| Statistics Skip X/Y | 4 / 4 |
| Statistics Ignore | 0 |
| Civil 3D connection | Data Connect |
| Elevation base | EG-SOTI-DTM |
| Drape command | DRAPEIMAGE |
| Visual Style | Realistic |
30. Final Conclusion
A high-resolution drone orthophoto is an extremely valuable engineering dataset, but it should not always be used directly inside CAD software. Creating a dedicated working raster provides a much better balance between image quality, file size and software performance.
In this Soti project, we successfully converted the original 10 cm/pixel orthophoto into a 1 m/pixel, 3-band RGB GeoTIFF, optimized it using pyramids, statistics and LZW compression, connected it to Civil 3D and finally used it as a drape image over the EG-SOTI-DTM surface.
The final result was a realistic 3D terrain model containing both:
Real terrain elevation + Real drone imagery
Keep the original 10 cm orthophoto as the master survey dataset.
Create a 1 m RGB working raster for general GIS/Civil 3D visualization.
Use the DTM as the elevation surface.
Use DRAPEIMAGE to apply the orthophoto.
Use the Realistic visual style to display the render material correctly.
Next Article in the GIS Series
GIS-04 | Cleaning NoData / Black Background Areas and Preparing the Final Civil 3D 3D Terrain Presentation
In the next workflow, we will deal with the black background around the drone orthophoto, improve the final appearance, and prepare the 3D terrain for professional engineering presentation.
Technical Note: This workflow was developed and tested using the Soti Drone Survey project with ArcMap and Autodesk Civil 3D. Settings may require adjustment depending on project size, raster characteristics, hardware and intended engineering application.
Post a Comment
Please No spam link, No Hate Speech in the comment box.