|
OmniPage Web 1.0 Review
|
|
by Hewie Poplock
|
|
http://www.hewie.net
|
|
I often like to use utilities that make my work easier. One chore that has plagued me has been converting a printed document to a Web
page. I wanted to take a 17-page printed version of our Homeowner's Association By-laws and place it on a Web site. This document was probably created on a typewriter and is not saved on a computer, so
to type it would be extremely time-consuming. Of course, there are ways to capture the document. However the only printed versions are copies of copies from either a typewriter or a dot matrix printer.
|
|
|
I tried using Adobe's Acrobat, but my printed version was so bad that Acrobat couldn't capture it to text and saved it to extremely large
.pdf files. I could capture the each page as a .jpeg graphic file, but that not only looked terrible, but the files were also large and took too long to load.
|
|
You can develop your own templates, but the documentation to do this is very limited. It was frustrating setting up my own template based on
an existing one, but once I figured it out, it was easy. Some of the paragraphs formatted differently than others, but it is simple to edit the pages to fix any inconsistencies.
|
|
|
There is a built-in wizard which eases the entire process. There are five steps to OmniPage Web:. 1.Scan or load image, 2.Set up OCR
zones, 3.OCR & proofing, 4.Outline the document, and finally 5.Save and launch. Each step can be done manually, but the wizard does provide some choices and works rather well.
|
|
|
My main fault with the wizard was the proofing process, which doesn't automatically start for each page as it completes its scan of that
page. The proofing can be a page or two behind the actual scanning, which sometimes got confusing as to what page I was at in the scanning process. However, the wizard did an exceptionally good OCR job and
the proofing allowed me to accept items that it thought might be wrong, but were fine. I could fix errors. The main OCR problems were due to copier specs that it saw as periods or commas. These errors were
easily corrected.
|
|
As the pages are scanned, the program takes headings and creates an outline which can be manually manipulated and apply filters to it. This
outline is the basis for creating the formatting of the text on the Web pages.
|
|
|
|
There is no HTML editor included. If HTML changes are desired, you will need to use Notepad or a program similar to HomeSite. The HTML code
is quite clean, which is quite the opposite of code produced by Microsoft Word.
|
|
There are several views of information available within the program. There are individual page scans, the outline, the image view, and the
text view. There are some choices for type settings, but they are limited to HTML, such as H1, H2, Headline, Header, Footer, etc. These can be preset. Style sheets can be automatically
developed. Choices for HTML are set up in an Options box prior to creating pages.
|
|
|
|
I also used some magazine pages with pictures to create Web pages. Pages with pictures need to be handled manually, instead of using the
wizard. The results need some fixing, but certainly it works as expected.
|
|
OmniPage Web comes with an 88-page manual. The manual has a lot of pictures, but it is more of a listing of all of the choices, settings, and
screens, rather than a "how-to" type of book. Unfortunately, the documentation is not on the CD. But it is available for download from the Scansoft's Web site. The internal help files do help with
enough information to complete a task.
|
|
OmniPage Web is product of Caere Corporation, which was purchased by Scansoft, Inc. OmniPage Web version 1.0 is a 1999 product (based on version 9
of OmniPage) and the ScanSoft is currently offering version 11. I am not sure if this will be a continuing product or will be a victim of the merger. It is a good product, and with some fixes would be a great
product. It is a bit over-priced, however, at $499.99. Scansoft is located at http://www.scansoft.com.
|
|
|
