For an all-Java approach I think the best option is to use a library like Apache POI for reading
DOC files. You would then write code that uses the POI API for getting at the content and style information of the document, and use that to create an in-memory image (like a BufferedImage). The image can then be stored as a TIFF file. This would be a decent-sized project for even moderately complex DOC files, though - probably at least a week or two.