What is the most effective method to accurately extract text and figures from PDF documents?