This research work highlights the potential of airborne LiDAR data to assess differences in Vegetation and Terrain
This study delves into the capabilities of point clouds, exploring how the spatial relationships between points (XYZ-coordinates) can be leveraged in focal operations to derive attributes for evaluating terrain characteristics, such as flatness or roughness.
Additionally, the study utilizes red, green, and blue attributes obtained from aerial photography together with Intensity attributes (sourced from public LiDaR data at 1064nm wavelength) to evaluate variations in vegetation by computing indices like the Green Leaf Index and the Normalized Difference Vegetation Index (NDVI)
Read more on the paper below:
Computation of Vegetation Indices on Point Clouds in SagaGIS
Vegetation indices offer a straightforward approach to gauge plant responses across different seasons and ecosystems. Typically, these indices involve establishing a connection between reflected light in the visible electromagnetic spectrum (specifically, Red, Green, and Blue Color bands) and the unseen electromagnetic spectrum; specifically in the Near-Infrared (NIR), since photosynthetic activity reflects most in this wavelength.
It’s worth noting that for scientific and precise applications, a multi-spectral imaging camera is commonly employed to acquire reflectance values of vegetation in the Near-Infrared. However, in the context of this study, we will explore the use of intensity values from LiDAR sensors that operate in the NIR.
Intensity (reflectance) attributes at wavelengths in the Near-Infrared (NIR)
Intensity (reflectance) attributes that are recorded using LiDaR sensors which operate at a wavelength in the Near-infrared NIR (from 780 nm to 2500 nm) are here used to assess plant health and differentiate plant types. In this case, the Sensor operated at a wavelength of 1064 nanometers so that the recorded intensity- /reflectance values are here used as approximated NIR values to compute the Normalized Difference Vegetation Index (NDVI) and highlight Vegetation through Color Infra-red Imaging (CIR) by altering the RGB band composition to IRG.
Vegetation Indices using Near Infrared (NIR) and RGB
| RATIO | Ratio Vegetation Index | NIR / R |
| NDVI | Normalized Difference Vegetation Index | (NIR – R) / (NIR + R) |
| RGBVI | Red-Green-Blue Vegetation Index | [G-(B*R)] / [G*G + (B*R)] |
| EGI | Excess Green Index | (2*G – R – B) |
| GLI | Green Leaf Index | (2G – R – B) / (2G + R + B) |
| VARI | Visible Atmospheric Resistant Index | (G – R) / (G + R – B) |
| NGRDI | Normalized Green Red Difference Index | (G – R) / (G + R) |
NDVI
NDVI is used to quantify vegetational “greenness” and is useful in understanding vegetation density and assessing changes in plant health and ecological compostion. NDVI is calculated as a ratio between the red (R) and near-infrared (NIR) values in traditional fashion: (NIR – R) / (NIR + R). (source: U.S. Geological Survey)
CIR
Colour Infra-Red (CIR) Images can be used to highlight and differentiate Vegetation. Vegetation with higher photosynthetic productivity will reflect in more saturated red than Vegetation with lower photosynthetic productivity. Areas that are not Vegetation, such as roads, buildings, etc. will not reflect in Red.
Computation of geometric feature attributes in CloudCompare
Highlighting differences in Point Clouds | Geometric Feature Attribute: Planarity (1.5) | data source: Norge Digitalt – Kartverket | Image: Marc Ihle 
Highlighting differences in Point Clouds | Geometric Feature Attribute: Normal change rate (1.5) | data source: Norge Digitalt – Kartverket | Image: Marc Ihle 
Highlighting differences in Point Clouds | Geometric Feature Attribute: Planarity (1.5) | data source: Norge Digitalt – Kartverket | Image: Marc Ihle 
Highlighting differences in Point Clouds | Geometric Feature Attribute: Sphericity (1.5) | data source: Norge Digitalt – Kartverket | Image: Marc Ihle 
Highlighting differences in Point Clouds | Geometric Feature Attribute: Roughness (1.5) | data source: Norge Digitalt – Kartverket | Image: Marc Ihle 
Highlighting differences in Point Clouds | Geometric Feature Attribute: Eigenentropy (1.5) | data source: Norge Digitalt – Kartverket | Image: Marc Ihle 
Highlighting differences in Point Clouds | Geometric Feature Attribute: 1st order moment (1.5) | data source: Norge Digitalt – Kartverket | Image: Marc Ihle 
Highlighting differences in Point Clouds | Green Channel Color | data source: Norge Digitalt – Kartverket | Image: Marc Ihle 
Highlighting differences in Point Clouds | Geometric Feature Attribute: Sphericity (1.5) | data source: Norge Digitalt – Kartverket | Image: Marc Ihle 
Highlighting differences in Point Clouds | Geometric Feature Attribute: Omnivariance (1.5) | data source: Norge Digitalt – Kartverket | Image: Marc Ihle 
Highlighting differences in Point Clouds | LIDAR Attribute: Intensity | data source: Norge Digitalt – Kartverket | Image: Marc Ihle 
Highlighting differences in Point Clouds | Geometric Feature Attribute: Omnivariance (1.5) | data source: Norge Digitalt – Kartverket | Image: Marc Ihle 
Highlighting differences in Point Clouds | Geometric Feature Attribute: PCA2 (1.5) | data source: Norge Digitalt – Kartverket | Image: Marc Ihle 
Highlighting differences in Point Clouds | Geometric Feature Attribute: Z-elevation | data source: Norge Digitalt – Kartverket | Image: Marc Ihle 
Highlighting differences in Point Clouds | Geometric Feature Attribute: Verticality (1.5) | data source: Norge Digitalt – Kartverket | Image: Marc Ihle 
Highlighting differences in Point Clouds | Geometric Feature Attribute: Anisotropy (1.5) | data source: Norge Digitalt – Kartverket | Image: Marc Ihle 
Highlighting differences in Point Clouds | Red Channel Color | data source: Norge Digitalt – Kartverket | Image: Marc Ihle 
Highlighting differences in Point Clouds | Geometric Feature Attribute: Planarity (1.5) | data source: Norge Digitalt – Kartverket | Image: Marc Ihle 
Highlighting differences in Point Clouds | Blue Channel Color | data source: Norge Digitalt – Kartverket | Image: Marc Ihle
Besides the Computation of Vegetation Indices, various other attributes can be computed using focal operations and evaluating neighbourhood relationships between points within a point cloud. This can help to classify point clouds and for example, extract points based on shared geometric features.
Data used
Source: Kartverket
Sensor
| Laser sensor: | RIEGL_VQ-1560i-DW |
| IMU: | Trimble Applanix IMU-57 |
| Gyromount: | SOMAG GSM4000 |
| GNSS: | Trimble Applanix AV-610 / 6 Loggrate: 5 Hz |
| Fly: | Piper PA-31-350 Chieftain |
liDaR Recording
Tromsø 8pkt 2019 | recorded by Terratec AS – Field Group / Flight date: 09.07.2019
RGB Data
Sampled by Ortho Images recorded by Blom Norway AS
Software used
- SagaGIS Conrad, O., Bechtel, B., Bock, M., Dietrich, H., Fischer, E., Gerlitz, L., Wehberg, J., Wichmann, V., and Böhner, J. (2015): System for Automated Geoscientific Analyses (SAGA) v. 2.1.4, Geosci. Model Dev., 8, 1991-2007, doi:10.5194/gmd-8-1991-2015.
- CloudCompare (version 2.12.4) [GPL software]. (2023). Retrieved from http://www.cloudcompare.org/ | Author: Daniel Girardeu-Monteau
- Blender (version 4.0) Blender Foundation | Founder & Chairman: Ton Roosendaal // Blender Spreadsheet Importer Add-on | Retrieved from: github.com | Author: Simon Broggi
