You are working on a project that requires extracting text from images of scientific papers. The images are of varying quality and contain complex layouts, including tables and equations. Describe how you would use Azure AI Vision to extract text from these images and what additional features or configurations you would need to consider to handle the complexity of the layouts.