Below is an incomplete visual description of the process I explore in my MA thesis. Written on Jupyter Notebooks in Python/Numpy, a photomosaic process is implemented for 2D pixel data and then extended into 3 dimensions to process volumetric (here: X-ray tomography) data.
Blocks of voxels between a dataset and a set of modular tiles are compared for similarity, and new datasets and 3D models are created based on the findings.
The first batch of images visualizes parts of the process in 2D. In the second batch, voxel data is analyzed and reorganized into typographic elements constructed from digital seahorse bones. Finally, Python scripts written for Modo and Blender make it possible to construct digital sculptures from predefined 3D modules based on features found in the datasets.
Seahorse dataset courtesy of Ghent University:
