Using the HP/Axis scanner on the Birmingham system

Using the HP/Axis scanner on the Birmingham system

Author: L.S.Lowe. File: hpaxis. This update: 20071106. Part of Guide to the Local System.

We have an HP 7650 scanner, attached to an Axis 70U Network Document Server, which can be used to scan a photograph, drawing, or other image into bitmap form. It's located in room W330, on the left hand side of the door.

It can be used for a single page or multiple pages, single-sided or double-sided, by inserting pages into the feeder on the top of the scanner. Alternernatively it can be used as a flat-bed scanner, by lifting the feeder unit lid, placing your document on the glass, and closing the lid while the scan takes place. Remember to remove the document from the glass afterwards, because the same scanner light source is used for feeder unit scans too.

There are basically three ways of starting a scan:

  • On the Axis attached keyboard, you can type in your userid, choose your options, and press Start, so that the scan is sent to you as an email attachment.
  • From the Axis control panel, you can select a destination of /home/scanner, choose your options, and press Start, so that the scanned file is stored in that local directory. Note that the directory is cleared every morning at 6am. Don't use this method for private documents.
  • Using the Axis epscan web page, you control the scanner using your web browser, and the scanned image is sent to your web browser, either as an inline image, or via a browser helper application.
There are a number of different profiles which can be used to select the type of scan. Some of these are built-in, and some of them have been provided locally. One important criterion is the resolution, which is by convention given in dots per inch (dpi).  So 75 dpi is good for an image to be incorporated in a web document at roughly its original size, and 150 dpi is usually enough for other circumstances. Remember that the size of resulting file will vary roughly as the square of the resolution. You can try:
  • Color 150 jpeg: for a 150dpi scan producing a jpeg file.
  • Color 150 PDF: for a 150 dpi scan to produce a PDF file from one or more single-sided pages.
  • Color 150 PDF ds: to produce a PDF file from one or more double-sided pages. Note: double-sided scanning can be relatively slow.
  • Text: for a 300dpi scan to produce a compressed TIFF file of black and white text. See below for converting to editable text.
  • Text PDF: for a 300 dpi scan to produce a PDF file of black and white images. See below for converting to searchable text PDF.
Note that using the Axis epscan web page gives you the most control over the format, size, resolution, brightness, contrast, and quality, as you can choose Custom to modify each of these parameters before you start the scan. You then have the option of starting a Scan immediately, as a one off, or Adding that modified profile as a temporary addition to the choices that the Axis Document Server will offer you when you are standing in front of it. If you regularly use a profile of your own, ask LSL about making it permanent.

Selecting parts of an image

The Axis document scanner interface doesn't make it easy to select part of a picture to scan. The best you can do is to change the page size to (say) A5 if the part you want is in the upper part of the glass.

However, it is easy enough to edit the part of the picture you want afterwards:

  • If you have scanned to a destination of /home/scanner or as an email attachment, then you can use a utility like xv or xpaint or gimp to crop the image in the way you want afterwards;
  • If you have used the web browser epscan interface to the scanner, then after the web browser has drawn the scanned image, you can capture that part of the screen to a file. You'll find a screen capture utility in KDE Start -> Birmingham Tools -> Capture or you can use the import command directly. Bear in mind that when a browser like Firefox views an image with more width or height than your browser window, you will need to click the image to see it in its full scanned resolution.

Converting images of text to recognised text

  • Text profiles produce images, just like other profiles, not text. But the images are in a form which Optical Character Recognition (OCR) software works well with. 
  • To convert a Text PDF from image form to a searchable-text PDF, you can use Adobe Acrobat Professional on the Windows Terminal Server. Use menu item Document -> Recognise Text Using OCR. Then save the updated PDF.
  • To convert a Text image into raw editable text, you can use the gocr command in Linux (on our SL4 systems, or check with LSL).

Troubleshooting

  • If you used the document feeder and get a scanned image which just consists of vertical lines, then remove any document that the last user left on the flat-bed glass.
  • If your scanned image has poor brightness or contrast, then from the web interface you can try tweaking the brightness and contrast controls under Custom for the chosen Profile. If you find this problem for a locally-provided Profile (these always contain the resolution in the Profile name) then let LSL know. You may alternatively want to tweak the appearance using xv or gimp, for most formats except PDF.
  • Avoid scanning at an unnecessarily high resolution, as the time taken increases quadratically and it's rarely needed in practice.
  • Double-sided scans take more than twice the time of single-sided because of the extra feeds required to flip each page and then flip it back.



L.S.Lowe