Yet another PDF to LRF converter

[cacapee]

Zip (or rar) up the png files.

Also I reuploaded new builds of dos and linux versions.

[godel10]

Originally Posted by cacapee

godel10 - editlrfmetadata works for me in both .9 and .99 Can you pm me a pdf file which has that problem.

For instance you can use a pdf file I created myself. When I use your linux .99 version my output is this, and the output of editlrfmeta with this file is

title:
author:
bookid: PDFLRF046fff061
publisher:
description:
date: 2007-09-30
error displaying meta information: unexpected end of file. Current depth is 3 Line 17, position 14.

Note: Anyway, if I try your windows version then the output lrf file is neither working with Sony's Software for Librie ("Librie for Windows")

[kovidgoyal]

cacapee there is a bug in creation of the metadata block by pdflrf. I had to work around it in the latest libprs500. Basically there's content after the closing </Info> tag, which probably indicates that the size of the emtadata block is not being correctly set in the output LRF file.

[ereszet]

Originally Posted by cacapee

Zip (or rar) up the png files.

Also I reuploaded new builds of dos and linux versions.

Some tests:

djvu to lrf - 200 seconds
png (converted from djvu to tiff by DjVuDecode and from tiff to png by Irfanview - size over 20 Mb each) - 375 seconds (output size about 50 Kb for each png)
png (converted from djvu to tiff by DjVuDecode and from tiff to png by tiff2png - size less than 100 Kb each) - 340 seconds (output size as above)

It looks like unzipping and zipping takes it toll.

Zips created by an old pkzip version 5.1.2600 do not seem to work (I cancelled processing after more than 10 minutes).
Zips created by Total Commander are ok.

Pictures attached are from an old Polish heraldic book available in djvu format from an academic network of dLibra.

[cacapee]

I've uploaded a new build up that fixes a couple of metadata bugs.

[ereszet]

The pdflrf processed png picture I attached to my previous post was taken from Sony's small view. It is in practice the only view available for pdfs or image based lrfs. But the pngs can be seen (and panned) in medium and large view as well.

What a pity that Sony does not allow collections/folders of pictures. Why their developers have crippled the Linux implementation so much is beyond any reasoning. The Archos allows folders, so each collection of png book pages can be placed in its own folder and read page after page. In Sony, it is just a long list of png pages from the pictures menu. Fortunately they can be viewed in sequence by using the page buttons.

[godel10]

Originally Posted by cacapee

I've uploaded a new build up that fixes a couple of metadata bugs.

It seems that the Librie compatibility issue in the previous release was in that small bug.

Thus, release 0.99 is the first one that have Librie compatibility and a Linux version. Thank you very much cacapee

[ereszet]

Cacapee, below is a list of refinements for you to consider to make your program a comprehensive tool for "paper books to image" processing:

1. Deskew (based on lines rather than borders - like ClearImage)
2. Clean Black Borders before wiping out White Borders (again like ClearImage that removes the borders, and sets border area pixels to white). That option is reliable for B&W pictures only.
3. Straighten the curved text lines (the only program I know that does it properly is BookRestorer)
4. Remove the perspective/trapezoidal effect (as above)
5. Despeckle/ Remove noise
5. Remove black objects (lines and blobs) with a specified number and density of black pixels. No program that I know can do it properly.Some programs allow to remove the punch holes (ClearImage) and some allow to remove lines or blobs of certains dimensions (Scanfix) but every so often they remove parts of text as well. That is why I suggest to take into account the number of black pixels that are close together rather than vertical and horiontal dimensions of black objects. The line of text is a mix of black and white pixels, while a black blob will have much higher proportion of black to white.

I do not expect you to do what professional/commercial teams of developers have not achieved yet, but it looks from what you have already done that you can improve on them. The cleaning of black borders would be the most welcome refinement for "picture perfect" photos of book pages (camera pictures are of different proportions than paper documents/books and that leaves the black bacground at the picture borders). The other refinements are for poorly taken photos/microfilms from digital libraties.

Once again let me tell you how much I admire what you have done, and my suggestions are just a dream for the future.

[sszekacs]

Hi all,

I've read through the whole discussion. Cacapee, you did a wonderful job!

Just one question:

How can one output the generated pages in an image format?

I tried (using the command prompt):
pdflrf.exe --outputimages -vrs -f1 -l3 -i xxx.pdf -o xxx.lrf

The lrf file was generated without any error, however I could not find any generated images in the folder. What am I doing wrong?

Thanks guys,

Szabi

[ereszet]

In Windows interface you just give your output file the zip extention (in the setting window). Same for images as the input.

[godel10]

Originally Posted by sszekacs

Hi all,

I've read through the whole discussion. Cacapee, you did a wonderful job!

Just one question:

How can one output the generated pages in an image format?

I tried (using the command prompt):
pdflrf.exe --outputimages -vrs -f1 -l3 -i xxx.pdf -o xxx.lrf

The lrf file was generated without any error, however I could not find any generated images in the folder. What am I doing wrong?

Thanks guys,

Szabi

Try
pdflrf.exe -vrs -f1 -l3 -i xxx.pdf -o xxx.zip

[MosFet]

I think that the PdfLrf is stable and mature enough to reach the 1.00 Version.

Many thanks to Cacapee !

[ereszet]

The first picture shows from left to right:
original image,
image after automatic black background removal with ClearImage demo,
image after black background removal, processed with pdflrf (default portrait mode, thicken=3).

The black line and the black spot at the left margin prevent pdflrf from automatic trimming that part as a white space. That line comes from taking the picture with the opposite panel of the v-cradle getting into the view.

The second picture shows the same sequence but with the original image shot with greater care in order to avoid the view of the opposite panel of the v-cradle.

The lighting for both original images was not perfect, as one can see well in the decorative frame part of the images. I can do better than that but not in provisional conditions when I am far away from home now.

The pdflrf output of both images was set at the original input size rather than adjusted for Sony in order to better see the resulting details. For the same reason the pictures I attach are quite large (still less than 1 Mb each) although I saved them at 25% size of the originals at 50% compression.

[sszekacs]

Originally Posted by ereszet

In Windows interface you just give your output file the zip extention (in the setting window). Same for images as the input.

thank you, ereszet and godel10

[daiphile]

Thank you for wonderful program.
BUt I have a trouble in using your program.
When I let ghostscript option on, I always got error message.(Windows XP, Ghostscript 8.54) Is there anyone who doesn't have the problem?