|
Getting your Trinity Audio player ready...
|
In this technology era, 3D modelling has emerged as a crucial innovation across various fields, including mapping. Integrating 3D models in mapping processes provides a deeper understanding of the mapped objects. Light Detection and Ranging (LiDAR ) is one of the technologies employed for acquiring 3D data, generating precise Point Cloud data.
LiDAR is capable of capturing millions of high-precision coordinate points. By leveraging artificial intelligence (Deep Learning) through accurate data processing, 3D models can be generated to obtain a more comprehensive representation of the mapped object’s structure. It can yield richer information for making informed decisions across various domains such as urban planning, environmental monitoring, natural resource management, and more.
To address these matters, recently, the National Research and Innovation Agency (BRIN), through the Geospatial Research Centre (PRG), organised the Geomodelling Webinar Series #1 titled “3D Geomodelling: Past, Present, and Future”.
Atriyon Julzarika, the Head of the Geomodelling Research Group at PRG-BRIN, underscored that Geomodelling involves modelling the Earth from its subsurface or surface. He mentioned that this modelling approach is still relatively uncommon in Indonesia and has yet to be widely known by the public. However, its benefits will become increasingly evident in the future.
In his presentation titled “Developments of 3D Modeling in Indonesia,” Fahmi Amhar, a researcher at PRG BRIN, explained the need for 3D data in various mapping applications. It includes modelling topography, elevation, and contours for mapping purposes.
Additionally, 3D data is essential for image correction or orthorectification tasks in land cover mapping. It allows for precise aerial or satellite imagery alignment with the Earth’s surface, compensating for distortions caused by terrain variations, sensor geometry, and viewing angles. By incorporating 3D information, the resulting orthorectified images provide:
- A more accurate representation of the Earth’s surface.
- Enabling better analysis and interpretation in various applications such as urban planning.
- Environmental monitoring.
- Resource management.
Fahmi further elaborated on the wide range of applications that rely on 3D data. These include urban planning, visual analysis, environmental impact simulations, hydrological analysis, water flow modelling, flood management, construction site mapping, deformation analysis, interactive maps, and virtual reality. Understanding the need for 3D data makes it apparent how crucial it is in various domains.
He also emphasised the importance of studying data structures as a fundamental aspect of understanding modelling. For example, Point Clouds consist of randomly recorded sets of 3D coordinates.
“Grid refers to a matrix that contains z-values, with regularly spaced xy positions, known as a 2.5D model. This model is typically generated through interpolation from a Point Cloud on the topography. Mesh, on the other hand, is a data structure consisting of a collection of triangles or polygons that form the surface of a 3D object, used for graphic rendering and visualisation,” he explained.
He further discussed Voxel, representing a 3D object as a collection of volume elements. Voxel is commonly used in medical visualisation, fluid simulation, and volumetric analysis.
Fahmi also presented Octree as a 3D spatial tree structure divided into eight smaller subspaces. Additionally, Bounding Volume Hierarchy (BVH) is a hierarchical data structure that accelerates collision detection operations in graphics simulations.
“The sources of 3D data include LiDAR, which uses laser pulses to measure the sensor’s distance to objects and generates a 3D Point Cloud. Photogrammetry involves capturing photos from various angles to reconstruct a 3D object. Automatic Image Correlation now also produces 3D Point Cloud,” he explained.
Fahmi added that remote sensing or satellite imagery could be processed using photogrammetry or Synthetic Aperture Radar (SAR) using Radargrammetry/Interferometry.
“There are also position and motion sensor data, such as inertia sensors, GPS, or IoT. 2D maps can be converted to 3D using certain assumptions. Furthermore, there are engineering 3D models created using 3D software,” he concluded.
