I have one paper scanned into a PDF, which has a size of about 700KB and 300dpi. But if I take a snapshot of the PDF file using DPI settings of 300 as well, the image file (in PNG, JPG, or TIFF) is more than 2.5mb, even if after I tried several optimization tools. If I extract the image using pdfimages, the result is even bigger, about two 50mb PPM files, which can be converted into two 1mb PNG files.
My question is, how does the original PDF file achieve so high compression level? The metadata says it is created by Adobe Acrobat 7.0, PDFScanLib 1.2.2. And is it possible to achieve such compression level with other software? I need to manipulate the original PDF, but any attempts now substantially increases the file size or decreases its quality.