Use Acrobat to export the pdf file to single page tiff image.

Let's have a look at the output images:
Right click→ properties

Most of them are already in perfect condition: 300 dpi or higher, binary image with only 1 bit depth, which is good for the final pdf's small volume.
Of course, there would also be some exceptions, such as this one, with a significantly larger file size.

We just need to throw them all into Comic Enhancer Pro

Press on the button to process scanned books.

The best effect comes from curve.

X axis is the grayscale of original, while y is the projected new grayscale intensity. We need to suppress the tint gray that was the picture on the back side, while maintaining proper intensities on the characters.

Most of the gray suppressed to close to white, and the dark parts, text included, are also enhanced.
The adjusting method is like adjusting a Bezier curve, intuitive enough.

In pursuit of extremely tiny file size, we only need binary image.
Tick on the shaking
check box to achieve decent (though
limited) gray effect.

Let the batch begin. Some images might require an individual profile.

Launch FreePic2Pdf



Click Go, and be patient before it finishes.

Recognition results are quite impressive, we can hardly detect errors in the copied text. (Sumatra PDF, strongly recommended, btw)

Not perfect on English or Equations, but I'll say it has exceed my anticipation. And the quality of pictures are unbelievable.


Searching or annotating is more than usable, which is the purpose of our effort. (Drawboard PDF, also strongly recommended, btw)

File size dropped from 49MB to 35MB, the saving is not remarkable, because the original file was already doing a good job by making most of its pictures binary.
One last thing, the index or bookmarks. it can be readily solved by PdgCntEditor.

Still very easy drag-and-drop, the original file has included a good index, we'll make just make use of it.

Copy and paste — and done, yeah, it's that simple.

Sometimes you'll need to specify the page number of the 1st page, but basically, that's pretty much it.

Maybe we'll open up a page to talk about PdgCntEditor
some day.