Altadena Scans
Working with LiDAR and Photogrammetry in Post-Fire Landscapes
2025
A fire burned through homes in Altadena in January 2025. Homes which are more than built structures; they are a part of their inhabitant's lives, memories, and places of dreams. Books collected, furniture curated, spaces lived in.
There was little time before the plots were cleared in order to rebuild. Satellite images after the fires show entire neighborhoods cleared.
The LiDAR and photogrammetric scans of the homes were initially documentation processes; to create an immutable record of these homes and the lives lived within them. Through the processing period, we became familiar with details and moments in each scan.
After the initial processing, however, it was apparent that the raw LiDAR point clouds had an aesthetic which could not be recreated; they captured reality and represented themselves in a powerful, sublime way.
Working with pointclouds as a visual medium gives us the opportunity to present the ephemerality of scenes we scanned.
Most important of all was creating a balance of light and dark, which, to us, the raw LiDAR captured beautifully.
Photogrammetry-based 3D pointcloud, from drone capture of subjects. (Created using Premiere Pro, Lightroom, Metashape and Autodesk ReCap)
Developing point clouds aesthetically is particularly challenging, as we're not working with solid objects, but with layers of points, which render reality as a translucent, spectral entity. Scanned objects are, at once, invisible, and formally defined.
Photogrammetry-based 3D pointcloud, from drone capture of subjects. (Created using Premiere Pro, Lightroom, Metashape and Autodesk ReCap)
Processing
There were 3 major LiDAR capture methods used, including a handheld scanner and a stationary scanner, and a "Wearable" SLAM scanner which the homeowners could fit into and walk through their homes as they did before the fires. The last, in particular, created a LiDAR scan which was far more personal to each homeowner than aerial or technical scans.
Each homeowner was provided all the raw scans, processed scans, 3D models, raw drone footage, processed footage, and images from the site documentation.
LiDAR pointcloud (left) and processed 3D model derived from the pointcloud (right) (Created using Autodesk Maya and Leica Cyclone360: Natalie Rubio, and Autodesk ReCap, Blender, Rhino: Kahin Vasi)
In addition to the LiDAR scanning on of the site, we built multiple workflows around photogrammetry and the creation of 3D models using a combination of drone footage, DSLR images, images from the homeowners.
In order for the homeowners to be able to use these, we developed an entirely open-source workflow to create, export and edit 3D models.
Using RealityScan to align camera poses and build a pointcloud and textured 3D model from drone footage: a single flight path
Photogrammetrically derived 3D model from DSLR images (ultra-high resolution textures and features)
Photogrammetry was limited to certain homes, and certain elements of all homes; primarily the hearth.
After the initial processing sprint, we also tested out high-fidelity photogrammetry with over 2000 images per model, largely extracted from drone footage. While these were not standard or could not be created for every home, testing this method gave me more experience with professional photogrammetry workflows.
Using Agisoft Metashape to create hyper-dense pointclouds with drone footage: (Top) drone flight path and image alignment; (Middle) point cloud derived from photo alignment; (lower) high resolution 3D model of the subject
This project was a student-driven initiative. We are ever-grateful for the participation of the homeowners.
Created in collaboration with Altadena residents: Kelly Akashi, The Syms Family, Alison Amegatcher, Diana Thater, Erik Ghenoiu, and Sameer Elayyan — and a team of producers - Joel Feree, Sara Simon, Ade Ayoade, Natalie Rubio, Kahin Vasi, Marti Vera, Carlos Bonachea, Kai Johnson, Jillian Leedy, Sophie Pennetier, Pierce Meyers, and Matt Shaw - with the support of the SCI-Arc Resilient Futures Task Force, LACMA Art & Technology Lab, ScanLAB Projects, PT Capture, and The Little Things AI.
