4digitalbooks
Products Contact Reference Services Press Home
Digitizing Line Options Book Cradle Questions /
Answers
 
Cameras Software Miscellaneous  

 


Page Improver® is a registered trademark of 4DigitalBooks - ASSY SA.

Introduction :

Page Improver® is a software for automatic image processing of scanned pages and a fundamental production tool for digitization of books and bound documents. It is the ideal complement to any book scanner.

It is designed for fast processing of sequential images of scanned pages from books, bound documents and lose sheets. It produces a set of typical treatments that are necessary to prepare pages for OCR and for online publishing.

Key features of Page Improver® include ability to automatically follow evolution of pages as they shift and rotate under the camera over time of scanning. Page Improver® detects precisely position of gutter and dimensions of pages to split pages and crop outer borders with an extraordinary precision.

Page Improver® reduces substantially operator interventions on image processing and greatly improves production capacity.

Source images :

While a single sheet document is usually scanned by a single camera scanner, a bound document may be scanned with a single camera or a dual camera scanner. Following source images are accepted :

single pages
dual pages on single images
left and right pages on two images

Source images are usually raw image scans of pages presenting artifacts and defects.

Typical artifacts: - a dark border around the page
- a view of the gutter at binding of pages
- shadows of page curvature close to the binding
Typical defects: - skewed text (text not being horizontal, usually due to rotation of page)
- bleed through (print that appears by transparency from the back side of the page)
- dark page background


Page Improver® removes these artifacts and defects, to enhance aspect and readability of document. Processing is optimized for speed without compromise on quality. A book may be processed faster than any automatic book scanner may produce raw images. This enables a book scanner operator to run Page Improver® in parallel to the book scanning task.

Improvements on pages:

Page Improver® produces the following essential treatments:

Page Split : Separates left and right pages in two images, taking into account the presence of spine on the image.
Page Crop : Crops to individual pages, by taking away black borders and artifacts associated to presence of spine.
Page Deskew : Corrects rotation of grayscale and color text on pages, that delivers a page with horizontal text lines when the source image presents an angle following to imperfect turning of pages or following to prints with incorrect angle. Our deskew method works also on pages with very small amounts of text, pages with graphics and tables.

Page Unbleed : Suppresses text that appears by transparence from the back side of page, corrects density of text, corrects contrast of photos, clarifies background of page by removing darkness of aged paper for example.

Page Resize : Resizes pixel matrix to a different size and corrects DPI factor. This is very useful when one wants to scan at fixed resolution but provide images on different resolutions. This feature does over-sampling and provides better aspect of text when converting images to bitonal (threshold).
Page Threshold : Converts images to black and white, that is useful for all monographs or books that include only text and line art images or gravures.
Page Join : Assembles two images of left and right pages in order to make a dual page image.
Page Canvas : Applies a selected canvas size to all images of the document, resulting in uniform page format output.
Page Center : Centers horizontally content of pages on canvas. This allows left pages and right pages to appear in a similar way. Usually on source images left pages present larger left margin and right pages, larger right margin. Browsing sequentially left and right pages on the screen of a computer results on a continuous right and left shifting of page content that is undesirable for online presentation of documents.

Image Formats : INPUT
File format Compression
TIFF None, LZW, CCITT G4
JPG Any
BMP None
  OUTPUT
TIFF None, LZW, CCITT G4
JPG Any

Image Colors : Input and Output images may be bitonal B&W, 256 gray scales or color RGB.
Page Sizes : Page Improver is able to handle very large source images, such as color scans of newspapers for example (file size of a single A2 format page in 300 dpi RGB is about 100 Mb).

 

User Interface

User interface is composed of a page navigator, an image treatment configurator and an output preview pane and a lens.

Other Features

Step by Step / Batch mode: Page Improver® may work in page by page or in batch mode. Page by page allows fine tuning image treatment on sequential pages while batch mode allows to run settings on a whole set of pages.
Use of Presets : Presets may be stored in a profile with all configured image treatment settings for a book that represents a typical collection. This profile may be used to process all the books of this collection, thus saving significantly operator time.
Multi-Processor Licensing : On a single multi-processor system, our license allows to run as many instances of Page Improver® to simultaneously treat several books.
Installation Requirements :

CPU : Pentium IV (Core2Duo recommended)
RAM : 1GB
Hard Disk space used by the application: about 5MB
Screen resolution 1280x1024 or better.


Why is Page Improver better than other solutions ?

Page Improver implements a set of image treatment methods dedicated for scanned pages that are unequalled per their efficiency and speed performance, resulting in higher throughput and less operator interventions.

- Page-splitting and page-cropping - independent of page position and angle

  It's ability to automatically follow evolution of pages as they shift and rotate under the camera over time of scanning allows Page Improver® to detect precisely position of gutter and dimensions of pages to split. This is essential to reduce subsequent interventions on scanned images to remove artifacts of page borders.

- Polyvalent deskew - independent of page content

  Many deskew implementations in commercial products are limited to pages with large text blocks. When graphics or tables appear in these documents, or when pages have only few words, these deskew methods produce unexpected results resulting in wrong rotated pages. Working with such tools requires inspection of results that is an expensive and time consuming task.