Digitalization and data structuring of historical documentary files using computer vision

In this data recovery project for the first Quito Architecture Biennials, we converted typed documents into structured digital resources. These historical records contained valuable information about Latin American architectural development but were inaccessible in their original format and deteriorated over time. Our approach combined computer vision and Optical Character Recognition (OCR) to digitize the physical documents. This approach allowed us to extract text from damaged and faded pages, transforming unstructured content into searchable digital text.

The digitized content was organized into a database that captured key information including details of participants, thematic focus, award recipients and countries of origin. This structured format allowed for systematic analysis over time. We implemented exploratory data analysis techniques to identify patterns and relationships within the information. The resulting visualizations revealed changing architectural trends, evolving international participation, and the transformation of thematic priorities across different editions of the event.