Artwork Data Extraction Pipeline

Problem:

Art galleries and collectors often have large collections of artwork stored in PDF documents. Manually extracting data such as artwork images, titles, years, and other metadata is time-consuming and error-prone. There is a need for an automated solution to streamline this process and ensure that only the most relevant and detailed information is extracted.

Solutions:

We developed an end-to-end pipeline that automates the extraction of artwork data from PDF documents. The process begins with users uploading a PDF file, which may contain a single artwork per page or multiple artworks on a single page. The system then parses the PDF to extract text and images in the order they appear. It detects artwork images within the PDF and performs a similarity check to identify and retain only the largest image with the most detailed metadata, such as title, year, dimensions, and cost. Extracted data is cleaned, formatted, and validated to ensure accuracy. Finally, the processed data is stored in a database for easy retrieval and management, and users can download the extracted data in a structured format (e.g., CSV, JSON) along with the artwork images.